Stable Photocatalytic Coupling of Methane to Ethane with Water Vapor Using TiO2 Supported Ultralow Loading AuPd Nanoparticles

Jun Xie ,Yuheng Jiang ,Siyang Li ,Peng Xu ,Qiang Zheng ,Xiaoyu Fan ,Hailin Peng ,Zhiyong Tang ,＊

1 Chinese Academy of Science(CAS)Key Laboratory of Nanosystem and Hierarchy Fabrication,CAS Center for Excellence in Nanoscience,National Center for Nanoscience and Technology,Beijing 100190,China.

2 University of Chinese Academy of Sciences,Beijing 100049,China.

3 Academy for Advanced Interdisciplinary Studies,Peking University,Beijing 100871,China.

4 CAS Key Laboratory of Standardization and Measurement for Nanotechnology,National Center for Nanoscience and Technology,Beijing 100190,China.

5 Center for Nanochemistry,Beijing Science and Engineering Center for Nanocarbons,Beijing National Laboratory for Molecular Sciences,College of Chemistry and Molecular Engineering,Peking University,Beijing 100871,China.

Abstract: The selective conversion of methane to C2 hydrocarbons offers a sustainable approach to utilize natural gas efficiently and reduce reliance on conventional fossil fuels.Unlike the conventional thermal catalytic conversion that requires high temperatures and pressures,the photocatalytic pathway enables methane activation and selective conversion under mild conditions,holding great promise as a sustainable method.However,achieving the efficient generation of C2 compounds under flowing conditions using cost-effective photocatalysts remains great challenge.In this work,we synthesized an ultralow loading AuPd alloy nanoparticlesupported on TiO2(Au0.05-Pd0.05/TiO2)photocatalyst via simple chemical reduction.Characterization using X-ray diffraction(XRD),aberration corrected high-angle annular dark field scanning transmission electron microscopy(AC-HAADF-STEM)and in situ COdiffuse reflectance infrared Fourier transform spectroscopy(DRIFTS)confirmed its composition and structure.The performance of the Au0.05-Pd0.05/TiO2 photocatalyst in methane conversion was evaluated under flow-reaction conditions.Remarkably,the photocatalyst efficiently converted methane containing water vapor into C2 compounds,including ethane and ethylene,with a remarkable C2 production rate of up to 10092 μmol·g-1·h-1 and a selectivity of 77%.While water vapor was not essential for methane conversion,its presence enhanced the production of ethane and ethylene while suppressing overoxidation to CO2.The photocatalyst demonstrated excellent stability,maintaining its catalytic activity even after continuous reaction for 32 h,surpassing previously reported results.With the assistant of transient photocurrent response test,in situ X-ray photoelectron spectroscopy spectra and in situ DRIFTS,we uncovered that the exceptional catalytic activity of Au0.05-Pd0.05/TiO2 originates from the synergistic effect of Au and Pd,which promotes the separation of photogenerated carriers and facilitates the C—C bond coupling of·CH3 to produce C2 compounds.Furthermore,XPS characterization revealed that the introduction of water vapor replenished consumed lattice oxygen during the methane activation process,thus contributing to the catalyst’s stability.This study not only offers a cost-effective and efficient photocatalyst for methane conversion but also provides insights into the fundamental mechanism of photocatalytic methane conversion.We believe that our work will inspire the exploration of inexpensive catalysts with simple preparation methods,driving advancements in efficient methane to C2 compound conversion and contributing to sustainable photocatalytic pathways for the future.

Key Words: Photocatalysis;Methane;Flow-reaction;Water vapor;AuPd alloy nanoparticle;C2 compounds

1 Introduction

Given the limited availability of coal and oil resources,the direct and selective conversion of methane(CH4),the main component of methane hydrate and shale gas reserves,into highvalue hydrocarbons represents a highly desirable strategy for chemical industry to reduce its dependence on coal and crude oil1,2.However,due to the inert nature of CH4such as high C—H bond dissociation energy(439 kJ·mol-1)and low polarizability,its conversion often necessitates harsh reaction conditions and/or strong oxidants,making the process energyintensive3.The direct conversion of CH4to C2hydrocarbon compounds including ethane(C2H6)and ethylene(C2H4)can be implemented through non-oxidative coupling(NOCM)or oxidative coupling(OCM)pathways.However,these thermocatalytic processes require high reaction temperatures(＞600 °C),resulting in undesirable side reactions like carbon deposition in NOCM or over-oxidation in OCM as well as significant energy wastage.Therefore,the development of efficient and sustainable strategies for the selective conversion of CH4to high value-added multi-carbon compounds holds immense scientific value and practical significance.

Photocatalysis offers a sustainable approach to selectively convert CH4by activating its C—H bond under mild conditions4.In the closed reactors,various effective catalysts,such as Pd1/TiO2,ZnO-AuPd2.7%,GaN :ZnO solid solutions,have been developed for photocatalytic NOCM reactions5-7.Typically,anaerobic reaction conditions contribute to the coupling reaction and the formation of C2compounds.However,the continuous lattice oxygen consumption during the reaction leads to poor catalyst durability.The decline in photocatalytic NOCM performance over the extended reaction time has been widely observed,which may be alleviated by treating the catalyst in an oxygenated environment.For instance,a decrease in C2H6yield was observed over Pd-modified ZnO-Au hybrid catalysts after 8 h of reaction,and the consumed lattice oxygen was regenerated through H2O exposure and drying treatment6.Similarly,the performance of TiO2-loaded Pd single-atom photocatalysts declined after 6 h of reaction following by being recovered by heating in air5.The reaction also slowed down on Ag-HPW/TiO2catalyst after 7 h,which could be regenerated by exposure to air under irradiation8.Alternatively,the development of photocatalytic OCM in flow systems has garnered significant attention due its good stability and scalability.For example,Pt and CuOx-loaded TiO2photocatalyst endowed a C2product yield of 68 µmol·g-1·h-1in a flow reactor.However,the addition of O2gave rise to the generation of more overoxidation product CO2and low C2selectivity of 60%.Subsequently,an impressive C2H6production rate of 5000 μmol·g-1·h-1with 90% selectivity was achieved but with precise control of O2concentration at very low level(0.3%).Therefore,it is highly desired to develop scalable flow reaction systems for efficient and stable conversion of CH4to C2hydrocarbons products under anaerobic conditions.

The design of scalable photocatalytic systems for coupling CH4to C2hydrocarbon necessitates the use of inexpensive catalysts with simple preparation methods.Currently,most employed photocatalysts consist of metal oxide-supported precious metals like Au and Pd,due to their unique properties that enhance carrier separation and stabilize CH3intermediates.Nevertheless,the majority of recent studies employed precious metal loadings exceeding 0.5 wt%,limiting the application of these catalysts owing to their high cost and scarcity.Typically,reducing the noble metal content leads to a decrease in catalyst activity per unit mass,presenting a significant challenge to minimize noble metal loading without sacrificing catalytic activity.

Herein,we present TiO2supported ultra-low loading AuPd alloy nanoparticle catalysts(Au0.05-Pd0.05/TiO2)for the efficient and stable conversion of CH4to C2compounds in a flow reaction system with H2O vapor.Under full spectrum irradiation,we achieve a remarkable C2production rate of 10092 μmol·g-1·h-1and a selectivity of 77%.No noticeable degradation is observed over a continuous reaction period of 32 h.A series of controlled experiments andin situcharacterization reveal that AuPd alloy nanoparticles not only promote photogenerated carrier separation but also facilitate the crucial step of C—C bond coupling of·CH3to produce C2hydrocarbons.Furthermore,the presence of H2O vapor allows for timely regeneration of catalysts,ensuring its stability.

2 Experimental and computational section

2.1 Chemicals

Commercial titanium dioxide(TiO2,P25-type)was purchased from Shanghai King Chi.Palladium dichloride(PdCl2)was bought from Macklin.Sodium borohydride(NaBH4)was obtained from Thermofisher.Chloroauric acid(HAuCl4)was achieved from Innochem.All materials were used as received without further purification.Deionized water with a resistivity of 18.2 MΩ·cm was utilized in all experiments.

2.2 Photocatalyst preparation

The AuPd alloy nanoparticle loaded TiO2photocatalysts were synthesized using an ice-assisted NaBH4reduction method.Briefly,500 mg TiO2and 100 mL deionized water were added into a 250 mL flask.Ultrasonic treatment was conducted for 10 min until a uniform suspension was obtained.Subsequently,specific amounts of 50 mmol·L-1HAuCl4and 100 mmol·L-1H2PdCl4solutions were added dropwise to the suspension and continuously stirred for 3 h.Afterward,the flask was placed in an ice bath,and a NaBH4solution(56.75 mg NaBH4dissolved in 5 mL deionized water)was slowly added into the suspension while continuously stirring for 1 h.The resulting sample was collected by centrifugation and thoroughly washed with deionized water.Finally,the cleaned sample was dried at 80 °C overnight.A similar method was used to synthesize the Au(or Pd)modified TiO2catalyst,with the addition of specific amounts of 50 mmol·L-1HAuCl4(or 100 mmol·L-1H2PdCl4)solutions,while keeping the other steps same as those in the aforementioned method.As-prepared samples were labeled as Aux-Pdy/TiO2and Aux/TiO2or Pdy/TiO2,wherexandyrepresented the weight ratio of Au or Pd to TiO2,respectively.

2.3 Characterization

Inductively coupled plasma optical emission spectrometer(ICP-OES)measurements were performed using a Varian ICPOES 720.Powder X-ray diffraction(XRD)patterns were obtained on a D/MAX-TTRIII(CBO)instrument.In situX-ray photoelectron spectroscopy(XPS)spectra were acquired using an escalab 20 Xi XPS system.The data were collected both before light irradiation and after being irradiated for 120 min.The binding energy was referenced to the C 1sline at 284.8 eV,which corresponds to the binding energy of adventitious carbon.Aberration-corrected high-angle annular dark-field scanning transmission electron microscope(AC-HAADF-STEM)images and energy dispersive X-ray spectroscopy(EDS)data were acquired using a Spectra300 microscope operating at 300 kV.In situdiffuse reflectance Fourier transform infrared spectroscopy(DRIFTS)was carried out using the Thermo Scientific Nicolet IS50.Photocurrent test plots were recorded using the CHI-760E electrochemical workstation,where samples in 1 mol·L-1Na2SO4were tested.

2.4 Photocatalytic performance tests

The photocatalytic methane conversion tests were conducted using a custom-made flow reactor(Fig.S1,Supporting Information)consisting of five components:a gas wash bottle,a quartz reactor,a 300 W Xenon lamp,a drying tube and a gas chromatography(GC).Initially,a 3.3 cm×3.3 cm FTO slide coated with the photocatalyst was placed at the center of the quartz reactor.Subsequently,CH4gas was passed through a gas wash bottle containing deionized water to generate a moist CH4gas,which was then introduced into the quartz reactor.After the light reaction,the exhaust gas passed through a drying tube filled with silica gel to remove H2O vapor.Finally,the reacted gas components were quantitatively analyzed using GC(Shimadzu GC-2024C)equipped with a methane converter.A 300 W Xenon lamp(PerfectLight)served as the light source(Fig.S2),equipped with a 780 nm short-pass filter that transmitted light within the range of 300 to 780 nm and light intensity was 1000 mW·cm-2).In our experiments,the surface temperature of catalyst was maintained around 49 °C detected by an infrared thermometer(Fig.S3).

In a typical test,10 mg as-synthesized catalyst was dispersed in 500 μL deionized water.The resulting dispersion was uniformly applied onto a 3.3 cm×3.3 cm FTO glass substrate(with a thickness of 0.2 cm).The CH4flow rate was controlled at 70 mL·min-1using a mass flow meter.The reaction was performed at a pressure of 1 bar without additional heating,and the reaction time was 4 h.

Controlled experiments in Fig.5 were carried out using moist Ar(Ar + H2O),CH4gas(only CH4),moist CH4gas(CH4+ H2O),CH4and O2(CH4+ O2)as reactants.The same test procedure as described above was used for these experiments.

3 Results and discussion

3.1 Catalyst synthesis and characterization

To enable the scalability of photocatalytic system,we deliberately chose P25,the most commonly used commercial TiO2,as the support material.Subsequently,gram-scale synthesis of monometallic Au,monometallic Pd and bimetallic AuPd nanoparticles loaded onto TiO2was achieved through a simple NaBH4ice-assisted reduction method as previously reported9,10.The actual amounts of Au and Pd were determined through ICP-OES and the results were in line with the expected quantities(Table 1).

The crystal structure of as-prepared photocatalysts was analyzed using XRD(Fig.1).The XRD patterns of Au0.05-Pd0.05/TiO2,Pd0.05/TiO2,Au0.05/TiO2,and TiO2exhibit wellindexed peaks corresponding to anatase TiO2(PDF#21-1272)and rutile TiO2(PDF#21-1276),indicating the coexistence of both anatase and rutile crystal phases.Furthermore,there are no additional peaks observed in the XRD pattern after loading with Au,Pd or Au-Pd nanoparticles,which should be attributed to their low amount and/or high dispersion.

Fig.1 XRD patterns of Au0.05-Pd0.05/TiO2,Pd0.05/TiO2,Au0.05/TiO2,and TiO2.

The morphology and dispersion state of AuPd nanoparticles supported on TiO2was investigated by aberration corrected high-angle annular dark field scanning transmission electron microscopy(AC-HAADF-STEM).The obtained results are presented in Fig.2,where numerous nanoparticles are randomly dispersed on TiO2surface,exhibiting an average size of approximately 1.7 nm(Fig.2a,b).AC-HAADF-STEM images provide further insight into the crystal structure of photocatalysts,revealing a lattice stripe spacing of 0.352 nm corresponding to the(101)crystal plane of anatase phase TiO2(Fig.2c).Elemental mapping through energy-dispersive X-ray spectroscopy(EDS)confirms the uniform distribution of Au and Pd elements within the nanoparticles,thus demonstrating the successful synthesis of AuPd alloy nanoparticles(Fig.2d).In addition,control samples of TiO2supported with an average size of 4.63 nm Au nanoparticles(Au0.05/TiO2)and 1.82 nm Pd nanoparticles(Au0.05/TiO2)are obtained.Representative ACHAADF-STEM images and nanoparticle size distributions are given in Fig.S4-S5.

Fig.2 (a,c)AC-HAADF-STEM images of Au0.05-Pd0.05/TiO2;(b)Size distribution of Au-Pd nanoparticle;(d)HAADF-STEM image of the Au0.05-Pd0.05/TiO2 and elemental mapping of nanoparticle to show Pd(red)and Au(green)distribution within the nanoparticle.

Due to the ultra-low metal loading,we employed surfacesensitivein situCO-diffuse reflectance infrared Fourier transform spectroscopy(DRIFTS)to investigate the surface structure of AuPd nanoparticles supported on TiO211.The Au0.05-Pd0.05/TiO2,Pd0.05/TiO2and Au0.05/TiO2catalysts were initially exposed to a CO atmosphere for adsorption and saturation,followed by desorption using an Ar purge.Fig.3 illustrates the CO-DRIFTS spectra obtained during the desorption process,where the two adsorption bands at 2173 and 2109 cm-1correspond to the R and P branches of gaseous CO,respectively.In the case of Au0.05/TiO2,a shoulder at 2057 cm-1emerges and gradually decreases in intensity during the CO desorption process,indicating the presence of weakly interacting CO adsorbed on Au atop sites(Fig.3a)12.On the Pd0.05/TiO2surface,a sharp peak at 2057 cm-1gradually disappears,which is attributed to the desorption of weakly bound CO on the atop sites of Pd nanoparticles.Additionally,the broad peak ranging from 1980 to 1840 cm-1corresponds to CO adsorbed on bridge and 3-fold hollow sites(Fig.3b).Upon the exposure of Au0.05-Pd0.05/TiO2to CO,the shape of the peak at 2109 cm-1deviates from the typical P branch of gaseous CO due to its overlap with CO adsorbed on Au atop sites.Noteworthily,the apparent frequency shift from 2057 cm-1for Au0.05/TiO2to 2085 cm-1for Au0.05-Pd0.05/TiO2reveals the interaction between Au and Pd,confirming the formation of AuPd alloy nanoparticles13.The sharp peak at 2057 cm-1and the broad peak from 1980-1840 cm-1correspond to CO bound to isolated Pd atop site and bridge or 3-fold hollow sites,respectively(Fig.3c).

Fig.3 CO-probe DRIFT spectra of Au0.05/TiO2(a),Pd0.05/TiO2(b)and Au0.05-Pd0.05/TiO2(c).Note that all the gaseous CO and physically adsorbed CO are removed by argon flow purging at 298 K.

3.2 Photocatalytic performance

Once the successful synthesis of ultralow loading AuPd nanoparticle supported on TiO2was confirmed,we proceeded to evaluate the photocatalytic performance of Au0.05-Pd0.05/TiO2for CH4conversion.The photocatalytic experiments were conducted in a gas-solid phase flow reactor under Xenon arc lamp irradiation with CH4and H2O vapor.Gaseous products including CO2,C2H6and C2H4are quantitatively analyzed by GC(Fig.S6).As depicted in Fig.4a,pure TiO2exhibits a low CO2production rate.Upon metal loading,Au0.05/TiO2displays a modest C2H6production rate of 2263 μmol·g-1·h-1,while Pd/TiO2shows the formation of C2H6and C2H4,with the production rate of C2compounds gradually increasing as the Pd loading increases from 0.025 wt% to 0.05 wt%.However,when the Pd loading further increases to 0.5 wt%,the activity gradually decreases and CO2becomes the dominant product,indicating that 0.05 wt% is the optimum Pd amount for C2compounds formation.The production rate of C2compoundsreaches 6398 μmol·g-1·h-1with a selectivity of 60.82%.Notably,the formation of C2H4is ascribed to the strong dehydrogenation ability of Pd6.Importantly,the addition of only 0.05 wt% Au to form AuPd alloy nanoparticles results in higher catalytic activity and less overoxidation products.Significantly,Au0.05-Pd0.05/TiO2exhibits a C2H6production rate 3.85 times higher than that of Au0.05/TiO2and a C2H6production rate 1.46 times higher than that of Pd0.05/TiO2,while the CO2production rate is only 75% of that of Pd0.05/TiO2(Fig.4a).These results highlight that the synergistic effect of AuPd alloy loading efficiently promotes the conversion of CH4to C2compounds while suppresses the overoxidation towards CO2.

Fig.4 Photocatalytic CH4 conversion performance of various catalysts.(a)C2 compounds production rate for TiO2,Au0.05/TiO2,Pd/TiO2 with different Pd loading amount and Au0.05-Pd0.05/TiO2.(b)Aux-Pdy/TiO2 with different ratios of Au and Pd.Reaction conditions:10 mg photocatalyst,1 atm CH4,CH4 flow rate:70 mL·min-1,saturated water vapor,4 h,300 W Xenon lamp.

Subsequently,the influence of Au and Pd ratio in AuPd alloys on photocatalytic CH4conversion was investigated,keeping the Pd content fixed at 0.05 wt%(Fig.4b).When the mass ratio of Au to Pd is 1 :2,a higher amount of CO2is generated compared to Au0.05-Pd0.05/TiO2,leading to a decrease in the selectivity of C2compounds.This suggests that Pd surface sites are more prone to CH4overoxidation than Au sites.Conversely,when the ratio of Au to Pd is 2 :1,more Au surface is exposed,resulting in a decrease of CH4conversion rate compared to Au0.05-Pd0.05/TiO2,while maintaining the selectivity of C2compounds.Therefore,the optimal loading amount of Au and Pd in Aux-Pdy/TiO2is both 0.05 wt%.Notably,the Au0.05-Pd0.05/TiO2achieves a C2compounds production rate of 10092 μmol·g-1·h-1with a selectivity of 77.00% under light irradiation in flow system,surpassing the previous results of photocatalytic NOCM and even exceeding the reported performances of photocatalytic OCM(Table 2).

Table 2 Comparison of the state-of-the-art catalytic performance in NOCM and OCM.

After identifying the optimal catalyst,we then investigated the influence of reaction conditions.To begin with,control experiments were conducted to eliminate the possible interfere of contaminants on photocatalytic CH4conversion reaction.As shown in the leftmost panel in Fig.5,when CH4is substituted with Ar,no carbon-containing products are detected,confirming that all the C2compounds originate from CH4that serves as the sole carbon source.While only CH4is used as reactant(the second panel on the left),C2H6and C2H4are produced at relatively high rate.Upon the introduction of H2O vapor(the third panel on the left),the production rate of C2compounds slightly increases accompanying with decrease in CO2yield.This result indicates that H2O vapor is not essential for the conversion of CH4to C2compounds,but can enhance the production of C2H6and C2H4while suppressing the overoxidation reaction to CO2,aligning with reported results14.However,when O2is introduced instead of H2O vapor,a significant amount of CO2is generated and the production rate of C2compounds considerably decreases(the rightmost panel),which is assigned to the promotion of over-oxidation of CH4to CO2by O2.The aforementioned controlled experiments clearly highlight the importance of H2O vapor in the selective photocatalytic conversion of CH4to C2compounds.

Fig.5 Photocatalytic CH4 conversion performance under various conditions.C2 compounds production rate in the absent of CH4,in the present of only CH4,CH4 and H2O vapor or CH4 and O2.Reaction conditions:10 mg photocatalyst,1 atm CH4,CH4 flow rate:70 mL·min-1,4 h,300 W Xenon lamp.

To further assess the durability of photocatalytic CH4conversion systems,we carried out the time-dependent catalytic tests.Remarkably,when CH4and H2O vapor are supplied to Au0.05-Pd0.05/TiO2,a remarkably stable production rate and selectivity of C2compounds are discerned under continuous light irradiation for up to 32 h(Fig.6a).Impressively,this stability surpasses that reported in other systems(Table 2).As comparison,in the absence of H2O vapor,the production rate C2compounds gradually decreases with prolonged reaction time.After 32 h,the rate of C2H6production decreases by 30%(Fig.6c).Therefore,in the absence of H2O vapor,the catalyst gradually experiences deactivation.This phenomenon is commonly observed in photocatalytic NOCM and primarily attributed to the participation of lattice oxygen as an oxidant,which becomes consumed during the activation of CH4through a process similar to the Mars-van Krevelen mechanism17.In the absence of O2,the catalyst gradually deactivates as the consumed lattice oxygen cannot be replenished;whereas,in the presence of CH4and H2O vapor,the catalyst exhibits excellent stability.As for pristine catalyst,two characteristic O 1speaks at 529.2 and 530.8 eV are distinguished,corresponding to the lattice oxygen(Ti-O)and the surface oxygen vacancies(Ov),respectively.After 32 h of continuous reaction,the lattice oxygen slightly decreases from 79.6% to 76.9% accompanied with the oxygen vacancies increasing from 20.4% to 23.1%,which demonstrates that no significant consumption of lattice oxygen(Fig.6b).At the same time,no obvious change is observed in the binding energy of Ti before and after the reaction(Fig.6d).All the above results indicate that the H2O vapor can replenish the consumed lattice oxygenin situand sustain the stability of catalysts under irradiation8.

Fig.6 Stability tests of photocatalytic CH4 conversion over Au0.05-Pd0.05-TiO2 in presence of(a)CH4 and H2O vapor,(c)only CH4.Reaction conditions:10 mg photocatalyst,1 atm CH4,CH4 flow rate:70 mL·min-1,saturated water vapor,300 W Xenon lamp.XPS spectra of fresh and used catalysts:(b)O 1s,(d)Ti 2p.

3.3 Mechanistic investigation

The next question is why Au0.05-Pd0.05/TiO2possesses the excellent performance of photocatalytic CH4conversion to C2compounds under the flow condition of CH4and H2O vapor.The photocatalytic efficiency is known to be largely affected by the separation efficiency of photogenerated electrons and holes as well as the surface redox reaction efficiency.The carrier separation behavior was then investigated by photocurrent tests.As shown in Fig.7a,the AuPd nanoparticles loaded TiO2display higher photocurrent intensity than both Au0.05/TiO2and Pd0.05/TiO2,revealing that TiO2loaded with AuPd alloy has higher carrier separation efficiency compared with TiO2loaded with Au or Pd nanoparticles.The higher charge separation efficiency contributes to the improvement of photocatalytic CH4conversion rate.To further explore the carrier transfer behavior,in situXPS was conducted.However,due to its limit detection sensitivity,in situXPS measurement under light irradiation for Au0.05-Pd0.05/TiO2was featureless.Therefore,Au0.5-Pd0.5/TiO2that was prepared under the identical condition but with a higher content of Au(0.5 wt%)and Pd(0.5 wt%)was employed to study the function of Au and Pd.Fig.7b outlines the highresolution Au 4fspectra.The characteristic double peaks indicate that Au is in metallic state.After irradiation,the binding energy of Au 4fXPS shifts negatively by 0.1 eV,indicating the accumulation of photogenerated electrons at the Au site.The high-resolution Pd 3dspectra(Fig.7c)reveal that two representative peaks at 334.8 and 340.2 eV are attributed to the metallic state of Pd,while those at 336.5 and 342 eV are assigned to the Pd2+species.When Au0.05-Pd0.05/TiO2is exposed to light,the content of Pd2+significantly decreases while the content of Pd0increases,uncovering that Pd also acts as an electron acceptor.At the same time,the binding energy of O 1smoves to higher value(Fig.7d),indicating that holes are aggregated at the oxygen site of TiO2for CH4activation18.

Fig.7 (a)Transient photocurrent response test.In situ XPS under light irradiation of(b)Au 4f,(c)Pd 3d and(d)O 1s.(e)Schematic diagram of the energy band level position of TiO2,Au,Pd and Au-Pd nanoparticles.

A schematic diagram of the energy band level position of TiO2,Au,Pd and Au-Pd nanoparticles is presented in Fig.7e.The band structure of TiO2is adopted from our previous data19.The work function of Au(-4.70 eV),Pd(-5.12 eV)and Au-Pd(-5.10 and -5.12 eV)nanoparticles20,21lie within the band gap of TiO2,specifically closer to the conduction band edge.Pd has a larger work function than Au,making Pd/TiO2a more efficient electron acceptor.The work function of AuPd alloy is between-5.10 and -5.12 eV,which is lower than Au but equivalent to Pd.In comparison to Au or Pd nanoparticles,the energetically favorable state to receive the electrons from higher energy levels is responsible for AuPd nanoalloy.Altogether,charge transfer from TiO2to AuPd alloy is more effective than that to elemental nanoparticles,resulting in the efficient photogenerated electron and hole separation.

In situDRIFTS was employed to investigate the surface reaction intermediates during the photocatalytic CH4conversion over Au0.05/TiO2,Pd0.05/TiO2and Au0.05-Pd0.05/TiO2.Initially,the catalysts were exposed to CH4and H2O vapor under dark condition for adsorption.Upon irradiation,characteristic vibrational modes of CH2/CH3deformation at 1475 and 1438 cm-1emerge immediately along with peaks at 1360 cm-1attributed to *CH3(Fig.8),indicating the rapid dissociation of CH4on catalyst surfaces by photogenerated carriers.Additionally,peaks at 1100-1220 cm-1are ascribed to the characteristic mode of adsorbed *HCHO species,while the peak at 1054 cm-1is attributed to the C—O stretching of *CH3O.Note that both *HCHO and *CH3O have been recognized as the important intermediates in the photocatalytic conversion of CH4to CO2.Evidently,a *COO intermediate at 1527 cm-1is observed in the DRIFTS spectra of Pd0.05/TiO2(Fig.8b),indicating that Pd promotes the overoxidation of CH4to CO2.Most notably,a prominent peak corresponding to C—C stretching at 874 cm-1is observed on the surface of Au0.05-Pd0.05/TiO2(Fig.8c),confirming the occurrence C—C coupling on catalyst surface.In contrast,Au0.05/TiO2(Fig.8a)and Pd0.05/TiO2exhibit a weak signal of C—C stretching,suggesting that the presence of AuPd alloy could boost the C—C coupling reaction.

Fig.8 In situ DRIFTS spectra for photocatalytic methane conversion over(a)Au0.05/TiO2,(b)Pd0.05/TiO2 and(c)Au0.05-Pd0.05-TiO2.

Armed with the aforementioned observations,we have established the mechanism underlying the photocatalytic CH4conversion to C2compounds.When exposed to light,the carriers generated within TiO2are efficiently separated,facilitated by the suitable work function of AuPd alloy.The holes trapped in the lattice oxygen of TiO2play a crucial role in activating CH4,leading to the formation of·CH3.Subsequently,a portion of·CH3migrates to the surface of AuPd alloy,where it undergoes self-coupling reaction to generate C2H6.C2H6can further react with holes and dehydrogenate to form C2H4on the Pd surface.Meanwhile,the CH3group adsorbed on the TiO2surface(*OCH3)undergoes continuous oxidation,proved by the intermediates *HCHO and *COO,to form CO2.Importantly,the lattice oxygen consumed in this process is replenished by the presence of H2O vapor,ensuring that the stability of catalyst is maintained.

4 Conclusions

In conclusion,the ultralow loading AuPd alloy nanoparticle supported on TiO2(Au0.05-Pd0.05/TiO2)photocatalyst is successfully prepared through a simple chemical reduction method.Unlike the previous photocatalytic NOCM system,this catalyst exhibits exceptional activity and stability for photocatalytic conversion of CH4to C2compounds in a flow reaction system with H2O vapor.The achieved C2yield of 10092 μmol·g-1·h-1and a selectivity of 77% remain almost constant over a continuous reaction period of 32 h without noticeable degradation.The outstanding performance is attributed to the ability of AuPd alloy nanoparticles to enhance the separation of photogenerated carriers and provide a suitable C—C coupling site for *CH3intermediate.Furthermore,the introduction of H2O vapor effectively replenishes the lattice oxygen consumed during CH4activation,thereby guaranteeing the stability of the catalyst.This work offers an avenue for achieving sustainable and efficient conversion of CH4to multi-carbon hydrocarbons through the utilization of cost-effective photocatalysts and design suitable reaction systems.The findings provide valuable insights into the design and optimization of photocatalytic systems for the selective conversion of methane,contributing to the advancement of environmentally friendly and economically viable technologies for methane utilization.

Author Contributions:Conceptualization,Yuheng Jiang,Xiaoyu Fan and Hailin Peng;Sotftware:Siyang Li;Validation,Jun Xie,Yuheng Jiang,Peng Xu,Qiang Zheng and Xiaoyu Fan;Investigation,Jun Xie,Yuheng Jiang and Xiaoyu Fan;Writing -Original Draft Preparation,Yuheng Jiang,Jun Xie;Writing -Review &Editing,Jun Xie,Yuheng Jiang,Xiaoyu Fan and Zhiyong Tang;Supervision,Xiaoyu Fan and Zhiyong Tang;Funding Acquisition,Zhiyong Tang.

Supporting Information:available free of chargeviathe internet at http://www.whxb.pku.edu.cn.