Selective light-driven methane oxidation to ethanol Authors: Fei Xue, Chunyang Zhang, Cheng Cheng, Xueli Yan, Feng Liu, Xiaozhi Liu, Biao Jiang, Qiuyue Zhang, Lin Sun, Huiping Peng, Wei-Hsiang Huang, Chih-Wen Pao, Zhiwei Hu, Mingshu Chen, Dong Su, Maochang Liu, Xiaoqing Huang, Yong Xu Published: 2024 ABSTRACT Methane (CH4) photocatalytic upgrading to value-added chemicals, especially C2 products, is significant yet challenging due to sluggish energy/mass transfer and insufficient chemical driven-force in single photochemical process. Herein, we realize solar-driven CH4 oxidation to ethanol (C2H5OH) on crystalline carbon nitride (CCN) modified with Cu9S5 and Cu single atoms (Cu9S5/Cu-CCN). The integration of photothermal effect and photocatalysis overcomes CH4-to-C2H5OH conversion bottlenecks, with Cu9S5 as a hotspot to convert solar-energy to heat. In-situ characterizations demonstrate that Cu single atoms play as electron acceptor for O2 reduction to ·OOH/ · OH, while Cu9S5 acts as hole acceptor and site for CH4 adsorption, C − H activation, and C − C coupling. Theoretical calculations demonstrate that Cu9S5/Cu-CCN reduces C − C coupling energy barrier by stabilizing ·CH3 and ·CH2O. Impressively, C2H5OH productivity reaches 549.7 μmol g–1 h–1, with selectivity of 94.8% and apparent quantum efficiency of 0.9% (420 nm). This work provides a sustainable avenue for CH4 conversion to value-added chemcials. Introduction The selective conversion of methane (CH4), a primary component of natural gas and a major greenhouse gas, into value-added chemicals, has attracted increasing attention of researchers to alleviate the current energy crisis and environmental concern1–7. Generally, the conventional strategy for CH4 conversion requires high temperature and pressure, which is energy-intensive4–8. In this regard, the selective conversion of CH4 in a sustainable way, such as solar-driven conversion, has been emerged as the new frontier in chemical industry1,6–11. Over the past decade, great efforts have been devoted to CH4 photocatalytic conversion, and many products including methanol (CH3OH), formic acid (HCOOH), formaldehyde (HCHO) and methyl hydroperoxide (CH3OOH) have been successfully produced from CH4 photocatalytic oxidation12–16. Nevertheless, great challenges still remain for CH4 photocatalytic conversion due to its chemical inertness and the recombination of photogenerated electron-hole pairs. Moreover, the reported photocatalysts are able to trigger CH4 oxidation to produce C1 products, while the generation of C2 products is formidably challenging6,9,17. In principle, ideal photocatalysts for driving CH4 oxidation to C2 products should be capable of excellent ability for C − H activation and subsequent C − C coupling. Over the past decades, substantial efforts have been devoted to designing functional catalysts with strong absorption to the full spectrum of sunlight, enhanced separation of photogenerated carriers, and strengthened stability under light irradiation9,11–17. Nevertheless, the photocatalytic CH4 oxidation to target products, especially C2 chemicals, is still difficult. It has been recently reported that in addition to the classic photocatalytic process, the photocatalyst can absorb light and convert it into heat, which is known as photothermal effect, and can further promote the catalytic performance18–23. Moreover, the photothermal effect of catalyst will rapidly heat the local temperature around the active site, which produces hot electrons to enhance the catalytic performance19–24. However, the precise cooperation of photocatalytic process and photothermal process is greatly challenging due to the complicated mechanism. Typical for CH4 oxidation, either thermal catalysis or photocatalysis faces difficulties to selectively produce C2 products from CH425. Under the circumstances, the integration of photothermal effect and photocatalytic process for triggering CH4 oxidation to C2 products is seductive. In this work, we successfully realize the selective CH4 oxidation to C2H5OH over crystalline carbon nitride modified with Cu9S5 nanoparticles (NPs) and Cu single atoms (Cu9S5/Cu-CCN) by synergistically integrating photothermal effect and photocatalytic process. In particular, Cu9S5 can efficiently absorb light and convert it into heat (i.e., photothermal effect), resulting in the formation of local hot spot for CH4 activation. Owing to the synergy of photothermal effect and photocatalysis, the productivity of C2H5OH reaches 549.7 μmol g−1 h−1 at a selectivity of 94.8%, which represents as the state-of-the-art selectivity of CH4-to-C2H5OH. Impressively, the apparent quantum efficiency (AQE) reaches 0.9% under the irradiation with a light wavelength of 420 nm. Moreover, in situ characterizations suggest that Cu single atoms play as the electron acceptor for O2 reduction to generate reactive ·OOH/ · OH radicals, while Cu9S5 NPs act as the hole acceptor and the site for CH4 adsorption, C − H activation, and C − C coupling to C2H5OH via the intermediates of ·CH3/ · CH2O. This work not only provides a sustainable avenue for CH4 conversion to C2 products, but also contributes to the reduction of greenhouse gas and carbon neutrality, which will attract rapid interest in chemistry, energy, material science and beyond. Results Synthesis and structural characterizations Cu9S5/Cu-CCN with Cu9S5 NPs and Cu single atoms was prepared with NaCl/KCl molten salt-assisted calcination (Fig. 1a, see Methods for details). Transmission electron microscopy (TEM) image and atomic force microscopy (AFM) profile suggest that the morphology of CCN nanosheets is largely reserved with an approximate thickness of 3 nm (Fig. 1b, c and Supplementary Fig. 1). Moreover, obvious NPs were observed in TEM image of Cu9S5/Cu-CCN (Fig. 1d), which was further validated by the elemental mapping (Fig. 1e). The lattice distances of 0.33 nm and 0.20 nm in the high-resolution TEM (HRTEM) image of Cu9S5/Cu-CCN are ascribed to (002) plane of crystalline carbon nitride and (0120) plane of Cu9S5, respectively (Fig. 1f)26,27. The atomic percentages of Cu and S were measured to be ~9.1% and 4.6%, respectively, based on energy dispersive spectroscopy (EDS), with a Cu/S ratio of ~2 (Supplementary Fig. 2), which was closed to that from X-ray fluorescence spectroscopy (XRF, Supplementary Table 1). Furthermore, aberration corrected high-angle annular dark-field scanning TEM (AC-HAADF-STEM) image was collected to investigate the state of Cu in Cu9S5/Cu-CCN. As shown in Fig. 1g, the bright dots in AC-HAADF-STEM image labeled with white cycles are ascribed to the isolated Cu atoms, which is further confirmed by the intensity profile with line scan (Fig. 1h). The above results imply Cu in Cu9S5/Cu-CCN consists of Cu9S5 NPs and Cu single atoms. To further reveal the composite structure, Cu9S5/Cu-CCN was etched by acid (named as Cu-CCN). Note that the morphology and structure of CCN were largely reserved (Supplementary Fig. 3). In contrast, small Cu atoms remain in Cu-CCN after acid treatment (Supplementary Fig. 3 and Supplementary Table 2), indicating Cu atom are chemically bonded within CCN framework. The X-ray diffraction (XRD) pattern of Cu-CCN displays similar feature but with a slight negative shift of peak comparing with CCN28, which is attributed to the single state of Cu in Cu-CCN (Supplementary Fig. 4). In addition to the peaks of CCN, the appearance of Cu9S5 peaks in the XRD pattern of Cu9S5/Cu-CCN confirms its composite structure with Cu9S5 NPs and Cu single atoms. Besides, the negative shift of peak at 810 cm−1 in Fourier transform infrared spectra (FTIR) after Cu introduction is attributed to the weakening vibration of tri-s-triazine structure (Supplementary Fig. 5a)29. Moreover, the appearance of Cu−N band in the Raman spectra of Cu9S5/Cu-CCN and Cu-CCN suggests that Cu single atoms are anchored with Cu−N coordination (Supplementary Fig. 5b)30.Fig. 1Synthesis and electron microscopy characterization of Cu9S5/Cu-CCN.a Schematic illustrating the preparation of Cu9S5/Cu-CCN. b–g TEM image (b), AFM profile (c), TEM image with high magnification (d), STEM image with elemental mappings (e), HRTEM (f) and aberration corrected HAADF-STEM images (g) of Cu9S5/Cu-CCN. h Intensity profiles along the x → y at positions 1 and 2 in (g). Synthesis and electron microscopy characterization of Cu9S5/Cu-CCN. a Schematic illustrating the preparation of Cu9S5/Cu-CCN. b–g TEM image (b), AFM profile (c), TEM image with high magnification (d), STEM image with elemental mappings (e), HRTEM (f) and aberration corrected HAADF-STEM images (g) of Cu9S5/Cu-CCN. h Intensity profiles along the x → y at positions 1 and 2 in (g). Afterwards, the X-ray photoelectron spectra (XPS) of Cu9S5/Cu-CCN, Cu-CCN and CCN were collected to identify the electronic properties (Supplementary Fig. 6). In N 1 s and C 1 s XPS spectra (Supplementary Fig. 7), Cu9S5/Cu-CCN and Cu-CCN display similar nitrogen and carbon species with CCN. Compare to CCN, a negative shift of peak in the N 1 s XPS spectra of Cu-CCN and Cu9S5/Cu-CCN indicates that Cu single atoms are anchored with Cu−N coordination31,32. Analysis on the Cu 2p and Cu LMM Auger electron spectra (AES) spectra demonstrates that Cu in Cu-CCN primarily presents as Cu2+ (Fig. 2a and Supplementary Fig. 8), while Cu in Cu9S5/Cu-CCN presents as a mixture of Cu+ and Cu2+33,34. Correspondingly, the characteristic peaks in the S 2p XPS spectrum of Cu9S5/Cu-CCN are ascribed to S2−, Sn2− and SOx (Fig. 2b)34. Given that the peaks are absent for Cu+, Cu0, and Cu9S5 in electron paramagnetic resonance (EPR) spectra (Supplementary Fig. 9)12,35, the peaks in the EPR spectra of Cu9S5/Cu-CCN and Cu-CCN can be indexed as Cu2+ species (Fig. 2c). Furthermore, X-ray absorption near-edge structure (XANES) and extended X-ray absorption fine structure (EXAFS) spectroscopy were performed to further study the electronic structure of Cu. Compared to Cu foil, Cu2O and CuO, Cu in Cu-CCN mainly displays as Cu2+, while Cu in Cu9S5/Cu-CCN is a mixture of Cu+ and Cu2+, which is in line with XPS results (Fig. 2d)28,29,31,35. Besides, the pre-edge transitions at 8982.2 eV and 8986.6 eV in the XANES spectrum of Cu9S5/Cu-CCN are ascribed to the 1 s → 4p transition of Cu+ and Cu2+, respectively (Fig. 2d)36. In the corresponding EXAFS spectrum of Cu-CCN, the feature at 1.50 Å is ascribed to Cu−N coordination28,29,31, suggesting that Cu single atoms are coordinated with unsaturated N sites in tri-s-triazine units. For Cu9S5/Cu-CCN, the presence of Cu−N/Cu−S coordination at 1.55/2.02 Å and the absence of Cu−Cu coordination imply that Cu9S5/Cu-CCN consists of Cu9S5 and Cu single atoms (Fig. 2e)37. Detailed fitting of EXAFS spectrum of Cu-CCN suggests that Cu single atoms are anchored with N in the form of Cu−N2 (Fig. 2f, Supplementary Figs. 10, 11 and Supplementary Table 3). Additionally, we simulated the coordination structure of Cu single atoms with the assistance of theoretical calculations (Supplementary Fig. 11 and inset of Fig. 2f). Note that the Cu−N and Cu−S coordination have been further confirmed by the wavelet transform profiles of corresponding EXAFS spectra (Fig. 2g).Fig. 2Structural characterizations.a High-resolution Cu 2p spectra of Cu9S5/Cu-CCN and Cu-CCN. b S 2p spectrum of Cu9S5/Cu-CCN. c EPR spectra over Cu9S5/Cu-CCN and Cu-CCN. d, e XANES (d) and EXAFS (e) spectra at Cu K-edge of Cu9S5/Cu-CCN and Cu-CCN. f The corresponding EXAFS fitting curves of Cu-CCN at R space. Inset is the structure of Cu-CCN. g Wavelet transform EXAFS spectra of Cu9S5/Cu-CCN, Cu-CCN, Cu foil, Cu2O and CuO at Cu K-edge. Structural characterizations. a High-resolution Cu 2p spectra of Cu9S5/Cu-CCN and Cu-CCN. b S 2p spectrum of Cu9S5/Cu-CCN. c EPR spectra over Cu9S5/Cu-CCN and Cu-CCN. d, e XANES (d) and EXAFS (e) spectra at Cu K-edge of Cu9S5/Cu-CCN and Cu-CCN. f The corresponding EXAFS fitting curves of Cu-CCN at R space. Inset is the structure of Cu-CCN. g Wavelet transform EXAFS spectra of Cu9S5/Cu-CCN, Cu-CCN, Cu foil, Cu2O and CuO at Cu K-edge. Performance of light-driven CH4 oxidation Subsequently, Cu9S5/Cu-CCN was used as the catalyst for CH4 photooxidation with CH4 (19 bar) and O2 (1 bar) in 100 mL H2O in a customized reactor (Supplementary Fig. 13), while Cu-CCN and CCN were used as references. Screening tests were performed to determine the optimal reaction parameters. We first investigated the influence of catalyst weight (Supplementary Fig. 14) and the ratio of Cu9S5 and Cu-CCN in Cu9S5/Cu-CCN on performance (Supplementary Fig. 15a). The optimal weight of catalyst and ratio of Cu9S5 were determined to be 10 mg and 13.8 wt%, respectively. Moreover, with a total pressure of 2.0 MPa, the optimal pressure of CH4 and O2 was determined as 1.9 and 0.1 MPa, respectively (Supplementary Fig. 15b). The selectivity and productivity of CH3CH2OH was positively depended on the pressure of CH4 and O2, implying that high pressure favors the oxidation of CH4 to CH3CH2OH (Supplementary Fig. 15c). Besides, the introduction of water could promote the oxidation of CH4 to CH3CH2OH, however, the excessive water would weaken the light absorption, as a result of the decay of yield and selectivity (Supplementary Fig. 15d). With the optimal parameters, the productivities of C2H5OH and CH3OH were only 58.6 μmol g−1 h−1 and 35.4 μmol g−1 h−1, respectively, when CCN was used as the catalyst (Fig. 3a and Supplementary Fig. 16). when Cu-CCN was used as photocatalyst for CH4 oxidation, the productivity of C2H5OH increased to 119.7 μmol g–1 h–1 with a selectivity of 70.3% (Fig. 3a and Supplementary Fig. 16). Impressively, the productivity of C2H5OH jumped to 549.7 μmol g−1 h−1 with a selectivity of 94.8% when Cu9S5/Cu-CCN was used as catalyst for CH4 photooxidation, suggesting that the synergy between Cu9S5 NPs and Cu single atoms could strongly promote the photocatalytic oxidation of CH4 to C2H5OH (Fig. 3a and Supplementary Fig. 16). The calculated apparent quantum efficiency (AQE) is strongly relevant to the absorption spectrum of Cu-CCN but deviates from that of Cu9S5/Cu-CCN, suggesting that light irradiation is not the sole factor for CH4 oxidation over Cu9S5/Cu-CCN (Fig. 3b). Considering the photothermal effect may contribute to CH4 activation, we further calculated the AQE values with different external temperatures, where the higher AQE with external heating verified that the photothermal was favored for CH4 oxidation (Supplementary Fig. 17). Note that the AQE of Cu9S5/Cu-CCN reaches 0.9% under the irradiation light with wavelength of 420 nm (Fig. 3b). Note that the C2H5OH yield and C2H5OH selectivity, as well as the AQE, of Cu9S5/Cu-CCN under various total pressure have largely exceeded the reported photocatalysts for CH4 oxidation to C2H5OH (Fig. 3c and Supplementary Table 4). Furthermore, the promising performance of Cu9S5/Cu-CCN has been confirmed by control experiments without light irradiation, CH4, or photocatalyst (Fig. 3d). To validate the CH4-to-C2H5OH conversion, isotope labeling experiment using 13CH4 was performed to track the carbon footprint. In 13C nuclear magnetic resonance (NMR) spectrum, the presence of peaks of CH313CH2OH (17.5 ppm) and 13CH3CH2OH (56.3 ppm) suggests that C2H5OH is indeed produced from CH4 oxidation (Fig. 3e)15. Both the formation of C2H5OH and CH3OH has been also confirmed by gas chromatography (Supplementary Fig. 18). The steady C2H5OH yield and selectivity (Fig. 3f), as well as the reserved structure and morphology of Cu9S5/Cu-CCN within 10 consecutive cycles demonstrated that Cu9S5/Cu-CCN could serve as active and stable catalyst for CH4-to-C2H5OH under light irradiation (Supplementary Figs. 19, 20). Additionally, the nearly-linear increase of C2H5OH in 15 h (Supplementary Fig. 21) and negligible S leaching (Supplementary Table 5) further confirmed the superior performance of Cu9S5/Cu-CCN for CH4-to-C2H5OH.Fig. 3Performance towards CH4 oxidation.a Yield and C2H5OH selectivity of Cu9S5/Cu-CCN, Cu-CCN and CCN. Reaction conditions: Wcat. = 10 mg, VH2O = 100 mL, PCH4 = 1.9 MPa, PO2 = 0.1 MPa, t = 1 h, light source: 300 W Xe lamp (full spectrum), light intensity: 400 mW cm−2. b AQEs for C2H5OH production of Cu9S5/Cu-CCN on various incident light wavelengths. Reaction conditions: Wcat. = 10 mg, VH2O = 100 mL, PCH4 = 1.9 MPa, PO2 = 0.1 MPa, t = 1 h, light source: 300 W Xe lamp. c Comparison between Cu9S5/Cu-CCN and other catalysts for CH4 light-driven oxidation to C2H5OH. d CH4 oxidation under various reaction conditions. Reaction conditions: Wcat. = 10 mg, VH2O = 100 mL, PCH4 = 1.9 MPa, PO2 = 0.1 MPa, t = 1 h, light source: 300 W Xe lamp (full spectrum or visible light λ ≥ 400 nm), light intensity: 400 mW cm−2 in conditions 1 − 4, 320 mW cm−2 in condition 5. e, f 13C and 12C NMR spectra of liquid products from 13CH4 and 12CH4 light-driven oxidation (e), and cyclic test (f) over Cu9S5/Cu-CCN. Performance towards CH4 oxidation. a Yield and C2H5OH selectivity of Cu9S5/Cu-CCN, Cu-CCN and CCN. Reaction conditions: Wcat. = 10 mg, VH2O = 100 mL, PCH4 = 1.9 MPa, PO2 = 0.1 MPa, t = 1 h, light source: 300 W Xe lamp (full spectrum), light intensity: 400 mW cm−2. b AQEs for C2H5OH production of Cu9S5/Cu-CCN on various incident light wavelengths. Reaction conditions: Wcat. = 10 mg, VH2O = 100 mL, PCH4 = 1.9 MPa, PO2 = 0.1 MPa, t = 1 h, light source: 300 W Xe lamp. c Comparison between Cu9S5/Cu-CCN and other catalysts for CH4 light-driven oxidation to C2H5OH. d CH4 oxidation under various reaction conditions. Reaction conditions: Wcat. = 10 mg, VH2O = 100 mL, PCH4 = 1.9 MPa, PO2 = 0.1 MPa, t = 1 h, light source: 300 W Xe lamp (full spectrum or visible light λ ≥ 400 nm), light intensity: 400 mW cm−2 in conditions 1 − 4, 320 mW cm−2 in condition 5. e, f 13C and 12C NMR spectra of liquid products from 13CH4 and 12CH4 light-driven oxidation (e), and cyclic test (f) over Cu9S5/Cu-CCN. Mechanistic studies of light-driven CH4 oxidation to C2H5OH To reveal the critical role of Cu single atom and Cu9S5 in photocatalytic reaction, the in situ XANES spectra of Cu-CCN at Cu K-edge were collected with and without light irradiation. As shown in Fig. 4a, the negative shift of K-edge in the XANES spectrum of Cu-CCN demonstrates that light irradiation leads to the decrease of Cu valence in Cu-CCN, that is, Cu single atom can accept the photogenerated electron (Fig. 4a)35. Besides, the transfer of photogenerated electron to Cu single site has been further validated by the negative shift in near ambient pressure XPS (NAP-XPS, Supplementary Fig. 22) and the attenuated Cu2+ peak in in situ EPR spectra (Supplementary Fig. 23)12–15,35. For Cu9S5/Cu-CCN, the slight decrease of Cu2+ peak (Cu single site) in in situ EPR spectra was attributed to the photoreduction of Cu2+ in Cu-CCN, further confirming that Cu single atoms could accept the photogenerated electrons (Fig. 4b). In this case, the increase of Cu2+ in Cu9S5 NP from 28.4% to 55.5% in Cu9S5/Cu-CCN with light irradiation was ascribed to the photooxidation of Cu9S5 NPs (Fig. 4c), demonstrating that Cu9S5 could serve as the hole acceptor38, which was further confirmed by the positive surface photovoltage in open circuit voltage decay (OCVD) measurement (Supplementary Fig. 24) and the different simulated charge density around Cu-CCN to Cu9S5 NPs (Supplementary Fig. 25)39,40. Besides, the weakened PL intensity and enhanced photocurrent density further confirm that the incorporation of Cu9S5 NPs and Cu single atoms can promote the spatial separation of photogenerated electron-hole pairs (Supplementary Fig. 26). Compared to CCN, the introduction of Cu single atom narrows the intrinsic bandgap and induces the formation of a midgap state (Et = 2.11 eV) (Supplementary Fig. 27)32,41. Combining these values with Mott−Schottky plots, the conduction band (CB)/valence band (VB) (vs. RHE) of CCN and Cu-CCN were determined as −0.92/1.72 V and −0.56/1.76 V, respectively (Supplementary Fig. 28)32. Analysis on the corresponding energy bands of these samples displayed that Cu9S5/Cu-CCN was capable of activating CH4 to ·CH3 and reducing O2 to ·OOH and ·OH, which further triggered the oxidation of CH4 to C2H5OH (Supplementary Fig. 29). Given that light irradiation is not the sole factor for CH4 oxidation over Cu9S5/Cu-CCN, we suppose that the photothermal effect may also contribute to the enhanced performance towards CH4 oxidation (Supplementary Fig. 17). We first collected the UV-vis-NIR spectrum of Cu9S5 NPs, where the strong broadband absorption in 200–1350 nm implied that Cu9S5 could harvest light and covert into heat (Supplementary Fig. 30)42. Consequently, we measured the surface temperatures of Cu9S5/Cu-CCN, Cu-CCN, and CCN under the irradiation of full-spectrum. The temperature increase (ΔT) of Cu9S5/Cu-CCN reaches 86.9 °C, which is significantly higher than that of Cu-CCN (45.9 °C) and CCN (37.4 °C) (Fig. 4d, e), suggesting that the incorporation of Cu9S5 NPs can induce the photothermal effect, which has been further confirmed by finite element simulation (Fig. 4f)43. Besides, the significance of photothermal effect on CH4 oxidation is validated by the decrease of overpotential (Fig. 4g) and Nyquist semicircle (Fig. 4h and Supplementary Fig. 31) with IR irradiation. In addition, CH4 oxidation was evaluated over Cu9S5/Cu-CCN under light irradiation with different wavelength to further reveal the synergy of photocatalysis and photothermal effect (Supplementary Fig. 32). The absence of C2H5OH even at the external temperature of 200 °C in dark suggested that the CH4-to-C2H5OH process (e.g., C − C coupling) was triggered by light irradiation, despite photothermal effect could promote CH4 activation (Supplementary Fig. 33). Besides, the disappearance of C2H5OH under IR irradiation (λ ≥ 800 nm) but the appearance of C2H5OH under the visible light irradiation (λ ≥ 420 nm) and full spectrum suggests that CH4 oxidation is triggered by photocatalysis, while photothermal effect further promotes CH4 oxidation to C2H5OH. Correspondingly, the increase of light intensity will both contribute to photocatalysis and photothermal effect (Supplementary Fig. 34), which further leads to the increase of C2H5OH productivity (Fig. 4i, j and Supplementary Fig. 35). Different from the gradual increase of C2H5OH selectivity with light intensity for Cu-CCN (Fig. 4j and Supplementary Fig. 35b), Cu9S5/Cu-CCN displays a volcano-shape of C2H5OH selectivity vs. light intensity, which can be attributed to the over-oxidation of C2H5OH due to its strong photothermal effect especially at high light intensity, as evidenced by the increase of CO2 selectivity with the increased light intensity. Note that the significance of the integration of photothermal effect and photocatalysis for CH4-to-C2H5OH was further confirmed by the lower activation energies (Ea) of Cu9S5/Cu-CCN than that of Cu-CCN (Supplementary Fig. 36)22.Fig. 4Mechanism investigation for light-driven CH4 oxidation.a In situ XANES spectra at Cu K-edge of Cu-CCN with and without light irradiation. b, c Solid-state in situ EPR spectra (b) and in situ Cu 2p NAP-XPS spectra (c) of Cu9S5/Cu-CCN with and without light irradiation. d, e Photothermal temperature curves (d) and corresponding IR thermal images (e) over Cu9S5/Cu-CCN, Cu-CCN and CCN. f Electric field distribution of Cu9S5 and Cu-CCN in Cu9S5/Cu-CCN irradiated at different wavelengths. g, h Polarization curves (g) and EIS Nyquist plots (h) of CH4 electrooxidation over Cu9S5/Cu-CCN in the dark and under IR irradiation. i, j Performance towards CH4 oxidation over Cu9S5/Cu-CCN (i) and Cu-CCN (j) at different incident light intensities. Reaction conditions in (i) and (j): Wcat. = 10 mg, VH2O = 100 mL, PCH4 = 1.9 MPa, PO2 = 0.1 MPa, t = 1 h, light source: 300 W Xe lamp (full spectrum), light intensity:200–500 mW cm−2. Mechanism investigation for light-driven CH4 oxidation. a In situ XANES spectra at Cu K-edge of Cu-CCN with and without light irradiation. b, c Solid-state in situ EPR spectra (b) and in situ Cu 2p NAP-XPS spectra (c) of Cu9S5/Cu-CCN with and without light irradiation. d, e Photothermal temperature curves (d) and corresponding IR thermal images (e) over Cu9S5/Cu-CCN, Cu-CCN and CCN. f Electric field distribution of Cu9S5 and Cu-CCN in Cu9S5/Cu-CCN irradiated at different wavelengths. g, h Polarization curves (g) and EIS Nyquist plots (h) of CH4 electrooxidation over Cu9S5/Cu-CCN in the dark and under IR irradiation. i, j Performance towards CH4 oxidation over Cu9S5/Cu-CCN (i) and Cu-CCN (j) at different incident light intensities. Reaction conditions in (i) and (j): Wcat. = 10 mg, VH2O = 100 mL, PCH4 = 1.9 MPa, PO2 = 0.1 MPa, t = 1 h, light source: 300 W Xe lamp (full spectrum), light intensity:200–500 mW cm−2. To further reveal the mechanism of light-driven CH4-to-C2H5OH, in situ EPR and attenuated total reflection infrared spectrum (ATR-IR) was conducted to track the intermediates during CH4 oxidation over Cu9S5/Cu-CCN. In in situ EPR spectra, the appearance of characteristic peaks of ·OOH and ·OH suggests that CH4 photooxidation experiences a radical process (Fig. 5a, b). The stronger peak intensities of ·OOH and ·OH for Cu9S5/Cu-CCN than those of Cu-CCN is consistent with the observed superior performance of Cu9S5/Cu-CCN to Cu-CCN (Fig. 5a, b). Moreover, the signal of DMPO- · CH3 of Cu9S5/Cu-CCN is stronger than that of Cu-CCN, suggesting that the synergy between Cu9S5 NPs and Cu single atoms can significantly promote the dissociation of CH4 to ·CH3 (Fig. 5b). In in situ ATR-IR spectra, several peaks appear at 1316/1340/1519/1540 cm−144–48, 670-770/1362/1376/1420 cm−115,16,49,50, and 1436/1490 cm−19,48, in the in situ ATR-IR spectra as increasing the light irradiation time to 110 min, which correspond to CH4, ·CH3 and ·CH2, respectively (Fig. 5c). Moreover, additional vibrational peaks corresponding to ·CH3OH/ · CH3OOH and methoxy ( · CH3O) appear at 1559 cm−115, and 1037/1092 /1458/1734/1749 cm−144,45,51–53, respectively. The peaks at 1475/1773 cm−1 are ascribed to ·CH2O54,55, which is attributed to the oxidation of ·CH3OH/ · CH3OOH and ·CH3O. The presence of ·CH3CH2O at 1398 cm−1 in ATR-IR spectra of Cu9S5/Cu-CCN and Cu-CCN under light irradiation suggests the formation of C2H5OH (Fig. 5c and Supplementary Fig. 37)55. The above results suggest that the synergy between Cu9S5 and Cu single site can promote CH4 activation to ·CH3, which is further oxidized into C2H5OH with the assistance of ·OOH and ·OH.Fig. 5Mechanism investigation and DFT calculation.a, b In situ EPR spectra of ·OOH (a) and ·OH (b) radicals trapped by DMPO over Cu9S5/Cu-CCN, Cu-CCN and CCN under light irradiation. c In situ ATR-FTIR spectra of photocatalytic CH4 oxidation over Cu9S5/Cu-CCN during light irradiation from 0 to 110 min. d Free energy diagrams for reaction pathways of CH4 oxidation to C2H5OH and corresponding energy profiles on Cu9S5/Cu-CCN and Cu-CCN. e CH4 adsorption energy on Cu9S5/Cu-CCN and Cu-CCN. f Differential charge difference of *CH2O (top) and *CH3 (bottom) absorbed on Cu9S5/Cu-CCN (red) and Cu-CCN (blue). The balls in pink, blue, aurantia, dark green and white represent C, N, Cu, S and H atoms, respectively. The cyan and yellow regions represent electron depletion and electron accumulation, respectively. The isosurface value is 0.004 e Å−3. g Desorption energy of C2H5OH on Cu9S5/Cu-CCN at different temperatures. h Schematic of the proposed reaction mechanism for photocatalytic CH4 oxidation to C2H5OH. Mechanism investigation and DFT calculation. a, b In situ EPR spectra of ·OOH (a) and ·OH (b) radicals trapped by DMPO over Cu9S5/Cu-CCN, Cu-CCN and CCN under light irradiation. c In situ ATR-FTIR spectra of photocatalytic CH4 oxidation over Cu9S5/Cu-CCN during light irradiation from 0 to 110 min. d Free energy diagrams for reaction pathways of CH4 oxidation to C2H5OH and corresponding energy profiles on Cu9S5/Cu-CCN and Cu-CCN. e CH4 adsorption energy on Cu9S5/Cu-CCN and Cu-CCN. f Differential charge difference of *CH2O (top) and *CH3 (bottom) absorbed on Cu9S5/Cu-CCN (red) and Cu-CCN (blue). The balls in pink, blue, aurantia, dark green and white represent C, N, Cu, S and H atoms, respectively. The cyan and yellow regions represent electron depletion and electron accumulation, respectively. The isosurface value is 0.004 e Å−3. g Desorption energy of C2H5OH on Cu9S5/Cu-CCN at different temperatures. h Schematic of the proposed reaction mechanism for photocatalytic CH4 oxidation to C2H5OH. Theoretical calculations Finally, density functional theory (DFT) calculations were conducted to gain the insight of CH4 photooxidation processes over Cu9S5/Cu-CCN and Cu-CCN, based on the optimized absorption structure of intermediates (Supplementary Figs. 38, 39). The energy profiles display that the cleavage barrier of C − H in CH4 to *CH3 on Cu9S5/Cu-CCN (0.40 eV) is lower than that of Cu-CCN (1.53 eV) (Fig. 5d). Further DFT calculations on the free energy imply that Cu9S5/Cu-CCN displays much stronger CH4 adsorption than Cu-CCN (Fig. 5e)15. Subsequently, the generated *CH3 will combine with ·OOH and ·OH to form CH3OOH and CH3OH, respectively. Compared to the direct desorption of *CH3OH (ΔG = −1.27 eV) over Cu9S5/Cu-CCN (Supplementary Fig. 40), the conversion of *CH3OH to *CH3O (ΔG = −3.32 eV) is thermodynamically favorable, which is conducive to C − C coupling to C2H5OH. Moreover, the Bader charge analysis and differential charge density demonstrate a total of 0.04 |e| and 0.59 |e| transfer from Cu9S5/Cu-CCN to *CH2O and *CH3, respectively (Fig. 5f), which are significantly larger than those of Cu-CCN (0.01 |e| and 0.37 |e| ). The above results imply that Cu9S5/Cu-CCN can stabilize the key intermediates of *CH2O and *CH356. Consequently, the energy barriers for CH4 activation to *CH3, *CH3O dehydrogenation to *CH2O, and subsequent coupling between *CH2O and *CH3 on Cu9S5/Cu-CCN were dramatically decreased in comparison to Cu-CCN (Fig. 5d). Additionally, DFT calculation was also used to reveal the thermal effect on CH4 photooxidation. The desorption energy of C2H5OH over Cu9S5/Cu-CCN decreases with the local temperature (Fig. 5g), suggesting that high local temperature can promote the desorption of C2H5OH. Based on the above results, we summarized the pathways for CH4 photooxidation to C2H5OH (Fig. 5h). First, the photoinduced electrons migrate to Cu single atom and lead to the reduction of Cu2+, and the adsorbed O2 reacts with the electrons to form ·OOH and ·OH radicals. The photoinduced holes migrate to Cu9S5 NPs, on which the adsorbed CH4 will be oxidized into ·CH3. Subsequently, the further interaction between ·OOH/ · OH radical and ·CH3 leads to the production of ·CH3OH/ · CH3OOH intermediate, which can be further oxidized and dehydrated into ·CH2O. Note that the significance of ·CH3OH for C2H5OH formation was further confirmed by the enhancement on C2H5OH yield after adding CH3OH into the reaction system (Supplementary Table 6). Finally, the combination of ·CH3 and ·CH2O generates C2H5OH (Supplementary Table 7). Discussion In summary, we have successfully integrated photocatalysis with photothermal effect for the conversion of CH4 to C2H5OH over Cu9S5/Cu-CCN with a composite structure of Cu9S5 NPs and Cu single atoms. Experimental observations and mechanism studies demonstrate that Cu9S5 NPs and Cu single atoms act as hole and electron acceptor, respectively, as a result of significant enhancement on the separation of photogenerated electron-hole pairs. Besides, Cu9S5 NPs can convert the absorbed light into heat and strongly raise the surface temperature of catalyst, namely, photothermal effect, which significantly promote the conversion of CH4 to C2H5OH. As a result, Cu9S5/Cu-CCN displays a C2H5OH productivity of 549.7 μmol g−1 h−1 at a selectivity of 94.8%, with an AQE of 0.9% at 420 nm, which has already surpassed all the reported photocatalysts. This work highlights the significance of rational photothermal enhancement, and provides new insights for selective C2+ oxygenates preparation from methane under mild conditions. Methods Chemicals and materials Biuret was purchased from Energy Chemical Co. Ltd. Cuprous thiocyanate (CuSCN) was obtained from Macklin Reagent Co. Ltd. Sodium chloride (NaCl), potassium chloride (KCl), hydrochloric acid (HCl, 36%–38%), hydrogen peroxide (H2O2, 30%) and dimethyl sulfoxide (DMSO, 99.99%) were supplied by Sinopharm Chemical Reagent Co., Ltd. All these chemicals were in analytical pure grade and used without further purification. The water in all the experiments is de-ionized (DI) water in resistivity of 18.2 MΩ·cm. Preparation of Cu9S5 decorated Cu single atom incorporated crystalline graphitic carbon nitride (Cu9S5/Cu-CCN) Typically, biuret (2 g) and CuSCN (137 mg) were ground together, then put into a covered corundum boat, followed by calcining at 500 °C (ramping rate, 5 °C min−1) for 2 h in Ar atmosphere. After cooling naturally to room temperature, the intermediate melon product was collected. Then, as-prepared intermediate melon product (120 mg), NaCl (860 mg) and KCl (170 mg) were finely ground together in a mortar. Subsequently, the mixture was further calcined at 550 °C for another 2 h in Ar flow and then cooled down to room temperature naturally. The obtained sample was washed with DI water to remove the metal salts before drying in a vacuum oven at 60 °C for 6 h. The final product was denoted as Cu9S5/Cu-CCN. Particularly, the loading amount of Cu9S5 could be controlled by changing the dosage of CuSCN (55 and 220 mg). Preparation of crystalline graphitic carbon nitride (CCN) CCN was prepared with the similar procedure for Cu9S5/Cu-CCN, except the absence of CuSCN. Preparation of Cu single atom incorporated CCN (Cu-CCN) Typically, as-prepared Cu9S5/Cu-CCN (50 mg) was dispersed in water (10 mL) in a beaker and treated under sonication for 10 min. Subsequently, HCl (15 mL) and H2O2 (15 mL) were added into the same beaker. Then the mixed suspension was stirred for another 12 h. Afterward, the product was centrifuged and washed by water until pH = 7. Finally, the sample was dried at 60 °C for 6 h in a vacuum oven. Photocatalytic reaction A 300 mL stainless-steel high-pressure batch-reactor with a 200 mL Teflon lining (Xi’an Taikang Biotechnology Co., LTD, China) was employed in photocatalytic CH4 oxidation. The reactor is equipped with a quartz glass top window to introduce light irradiation from a 300 W Xe lamp (incident light density = 400 mW cm–2) (Supplementary Fig. 13). Typically, 10 mg photocatalyst was well-dispersed in 100 mL DI H2O under ultrasonic treatment. After the reactor was sealed, ultrapure oxygen (99.99 vol.%) was continuously injected for 20 min to remove air. Subsequently, ultrapure oxygen (99.99 vol.%, 0.1 MPa) and CH4 (99.99 vol.%, 1.9 MPa) were charged into the reactor in sequence. The corresponding pressures before reaction and during reaction were monitored by the mechanical pressure gauge and pressure sensor equipped in the reactor (Supplementary Fig. 13). No pressure drop was observed after the reaction. Then, the reactor was top-irradiated by full spectrum (AM 1.5) and visible light (λ ≥ 420 nm), respectively. The reaction temperature during the reaction was monitored by a thermocouple, which was inserted into the photocatalytic suspension. The system temperature was determined to be 50 °C under light irradiation with 400 mW cm−2. The gaseous products were quantified by gas chromatograph (Agilent 8890, Ar as carrier gas, 5 A molecular sieve column, HayeSep Q packed column and HayeSep A packed column), equipped with flame ionization detector (FID, Agilent) and thermal conductivity detector (TCD, Agilent). A methane converter is built in the instrument to hydrogenate trace amounts of CO2 and CO into CH4 assisted by Ni catalyst and enters FID for detection. Specifically, the gas outlet of high-pressure reactor was connected the injection port of GC by a PTFE tube, thus a certain volume of gas would automatically flow into GC. H2, O2 and N2 were detected by TCD, while carbon-containing gas (CO2, CO, CH4) was detect by FID. The liquid products (CH3CH2OH and CH3OH) were quantified by 1H nuclear magnetic resonance (NMR) (Bruker AVANCE AV III 500) spectroscopy through a water suppression mode, with dimethyl sulfoxide (DMSO) in D2O as the internal standard. Specifically, the lid of high-pressure reactor would be removed, then the suspension was taken out by injection syringe (1 mL), following by filtering with syringe filter to obtain the liquid supernatant. 0.5 mL post-reaction liquid supernatant was mixed with 0.1 mL D2O containing dimethyl sulfoxide (DMSO, 100 ppm), which was then transferred to NMR tube. For time-coursed reaction, an extended stainless-steel tube which can insert into the liquid was connected to the gas outlet, and the liquid product was pressed out with the high-pressure gas in the upper part of the reactor. HCHO was also identified by the colorimetric method based on the reaction between acetylacetone and HCHO in the presence of acetic acid and ammonium acetate. No HCHO was detected. The apparent quantum efficiency (AQE) for CH3CH2OH evolution was also performed at the same reaction setup, using the incident light with monochromatic wavelengths (350, 365, 380, 420, 450, 500, 520, 580 and 650 nm). The AQE was calculated according the following Eq. (1):1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${AQE}( \% )={N}_{e}/{N}_{p}\times 100 \%={4N}_{{C}_{2}{H}_{5}{OH}}/{N}_{p}\times 100 \% $$\end{document}AQE(%)=Ne/Np×100%=4NC2H5OH/Np×100%where Ne stands for the number of converted holes to CH3CH2OH, while \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${N}_{{C}_{2}{H}_{5}{OH}}$$\end{document}NC2H5OH is the mole number of CH3CH2OH, and Np represents for the number of incident photons, respectively. Photocatalytic reaction with external heating Photocatalytic reaction with external heating was the same with those of photocatalytic reaction, except a furnace is used to control the temperature from room temperature to 200 °C. For determined the photothermal effect, 10 mg photocatalyst was homogeneously dispersed into water (3 mL) under ultrasonic treatment for 30 min. Afterwards, the suspension was spread onto the glass fiber filter membrane (diameter 25 mm, pore size 0.22 µm) uniformly and dried at ambient temperature for 3 h. Subsequently, the glass fiber membrane was exposed to the 300 W Xeon light source and surface temperature was measured using an infrared camera. The distance between glass fiber membrane and light outlet is fixed at 14 cm (the distance between light outlet of Xeon light and liquid level in high pressure reactor), and the incident light intensity was set as 400 mW cm−2. Photoelectrochemical measurement Photoelectrochemical tests were performed in a three-electrode system using a CHI660E electrochemical workstation (Chenhua Instruments Co., Shanghai) under irradiation of a 300 W Xe lamp, with sample, saturated calomel electrode (SCE) and carbon rod as working electrode, reference electrode and counter electrode. Na2SO4 aqueous solution (0.5 M) was employed as the electrolyte. For preparation of working electrodes, 5 mg of samples were dispersed in a mixture of isopropanol (990 μL) and Nafion solution (5 wt.%, 10 μL) under ultrasonication for 1 h to form homogeneous catalyst ink. Afterwards, 10 μL of the ink was dropped onto a L-type glassy carbon electrode (GCE, diameter = 5 mm). After natural drying, the working electrode would be obtained. Instrumentation High-resolution transmission electron microscopy (HRTEM) images were captured on FEI Tecnai F20 and Thermo Fisher Scientific Spectra 300 scanning transmission electron microscope equipped with a Super-X EDS detector system. X-ray powder diffraction (XRD) patterns were acquired on Rigaku Smart Lab diffractometer with Cu Kα source (wavelength = 1.5418 Å, 30 kV/40 mA). Raman spectra was obtained over a Jobin-Yvon LabRam HR 800 micro-Raman system equipped with a liquid N2 cooled detector. Atomic Force Microscopy (AFM) images were acquired from a Bruker Dimension Icon using the Peakforce Tapping model. Electron paramagnetic resonance (EPR) measurements were conducted on a Bruker EMX X-band spectrometer with microwave frequency=9.40 GHz at room temperature. X-ray photoelectron spectroscopy (XPS) was measured by Kratos spectrometer (AXIS Ultra DLD) with monochromatic Al Kα micro-fused monochromatized source (hν = 1486.6 eV). In-situ near ambient pressure XPS (NAP-XPS, SPECS Surface Nano Analysis GmbH, Germany) was carried out on a SPECS system equipped with a differentially pumped Phoibos hemispherical electron energy analyzer using monochromatic Al Kα radiation (1486.6 eV). The binding energy was calibrated relative to the signal of C 1 s peak at 284.8 eV. The UV-vis absorption spectra were collected from an Agilent Cary 5000 UV-vis-near-IR spectrophotometer (BaSO4 as reference). Steady-state photoluminescence (PL) and time-resolved photoluminescence (TRPL) spectra were both characterized by time-resolved PTIQM-4 fluorescence spectrophotometer at an excitation wavelength of 370 nm. X-ray absorption spectroscopy (XAS) data were acquired from NSRRC TPS44A beamline of the National Synchrotron Radiation Research Center (NSRRC, Hsinchu, Taiwan). Data were processed according to standard procedures using the Demeter program package (Version 0.9.24). Finite element simulation A commercial finite element method (FEM) package, COMSOL, was employed to calculate optical response of Cu9S5/Cu-CCN. In typical simulation, Cu9S5/Cu-CCN is placed in a tetragonal dielectric environment with a perfect matched layer (PML) shell. The FEM domain is filled with water (refractive index ne = 1.32). The catalyst is modeled to mimic their nanostructures. Specifically, Cu9S5 NPs (radius = 50 nm) is set on the top of Cu-CCN nanosheets (thickness = 10 nm). The radiation source is located at the top of the domain, transmitting downward vertically. The dielectric functions of Cu9S5 and CCN are from the literatures57,58. Plane waves are launched to excite the optical modes of Cu9S5/Cu-CCN and the FEM packages are used to solve for the stationary solutions in the frequency domain. The model computes the scattering, absorption, and extinction cross sections of the Cu9S5 particle on CCN substrate. The scattering cross-section is defined as Eq. (2):2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\sigma }_{{sc}}=\frac{1}{{I}_{0}}\iint \left({{\bf{n}}}\cdot {{{\bf{S}}}}_{{{\bf{sc}}}}\right){{\rm{d}}}S$$\end{document}σsc=1I0∬n⋅SscdSwhere n is the normal vector pointing outward from the nanodot, Ssc is the scattered intensity (Poynting) vector, and I0 is the incident intensity. The integral is taken over the closed surface of the scatterer. The absorption cross section can be calculated as Eq. (3):3\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\sigma }_{{abs}}=\frac{1}{{I}_{0}}\int\!\!\!\iint Q\,{dV}$$\end{document}σabs=1I0∫∬QdVwhere Q is the power loss density in the particle and the integral is taken over its volume. The extinction cross section is simply the sum of the two others, as Eq. (4):4\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\sigma }_{{ext}}={\sigma }_{{abs}}+{\sigma }_{{sc}}$$\end{document}σext=σabs+σsc Density function theory (DFT) calculation detail The Vienna Ab initio Simulation Package (VASP) based on DFT was employed in the entire calculations59–62. Perdew-Burke-Ernzerhof exchange-correlation functional within generalized gradient approximation was applied, while a plane-wave basis set was implemented in the framework of the projector augmented wave (PAW) method to represent the core and valence electrons with a kinetic cutoff energy of 400 eV63,64. The atomic structures were fully relaxed until energy and forces were converged to 10–5 eV·Å–1 and 0.02 eV·Å–1, respectively. A 1 × 1 × 1 Monkhorst-Pack k-point mesh was utilized to sample the Brillouin zone both for Cu-CCN and Cu9S5/Cu-CCN. The DFT-D3 method was adopted to describe the van der Waals (vdW) interaction65. In the case of Cu9S5/Cu-CCN model, the three bottom layers of Cu9S5 were fixed during the calculation. To avoid the periodic interaction along out-plane direction, a vacuum layer of 15 Å was used for the slab model. Post-computation analysis was carried out using the Vaspkit code66. The species adsorption free energy (ΔGs) on disparate active sites can be defined as Eq. (5):5\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\varDelta G}_{s}={\varDelta E}_{s}+{\varDelta E}_{{ZPE}}-{T\varDelta S}_{s}$$\end{document}ΔGs=ΔEs+ΔEZPE−TΔSswhere ΔEs is the species absorption energy, ΔEZPE is the difference in zero point energy between the surface species and dissociative species, T is the temperature, and ΔSs is the entropy change between the adsorbed state of the system and the gas phase standard state. Supplementary information Supplementary Information Transparent Peer Review file Supplementary Information Transparent Peer Review file Source data Source Data 1 Source Data 2 Source Data 3 Source Data 4 Source Data 5 Source Data 1 Source Data 2 Source Data 3 Source Data 4 Source Data 5