In situ Raman spectroscopic insights of hydrogen spillover in electrocatalytic hydrogenationYan Liu1 , Ze-Yu Zhang1 , Jie Wei1,2* , Yan Liu3 , Hua Zhang1,2* , and Jian-Feng Li1,2,4*1 College of Energy, College of Materials, State Key Laboratory of Physical Chemistry of SolidSurfaces, College of Chemistry and Chemical Engineering, School of Life Sciences, College ofPhysical Science and Technology, Discipline of Intelligent Instrument and Equipment, iChEM,Fujian Key Laboratory of Advanced Materials, Xiamen University, Xiamen 361005, China2 Innovation Laboratory for Sciences and Technologies of Energy Materials of Fujian Province(IKKEM), Xiamen 361102, China3 School of Chemistry and Materials Science, Anhui Normal University, Wuhu 241000, China4 College of Chemistry, Chemical Engineering and Environment, Minnan Normal Univers ity,Zhangzhou 363000, China* Corresponding authors, E-mail: weij@xmu.edu.cn; zhanghua@xmu.edu.cn; Li@xmu.edu.cn. Experimental proceduresChemicals. Cupric chloride (CuCl2 ·2H2 O), sodium hydroxide (NaOH), ascorbic acid (C 6 H8 O6),polyvinylpyrrolidone (PVP), ethanol, (3-aminopropyl)trimethoxysilane (APTMS), sodium silicatesolution (Na2 SiO 3 , 35-40%), hydrochloric acid (HCl), tetrabutylammonium nitrate (TBAN),sodium sulfate (Na2 SO4 ), potassium nitrate (KNO 3 ), ammonium sulfate ((NH4 )2 SO4 ), salicylicacid (C7 H6 O 3 ), sodium hypochlorite solution (NaClO, available chlorine 5.2% of aqueoussolution), trisodium citrate dihydrate (C 6 H5 Na3 O7 ·2H2 O), and sodium nitroferricyanide dihydrate(C5 FeN 6 Na2 O·2H2 O), para-nitrothiophenol (C 6 H4 O2 NSH) were purchased from SinopharmChemical Reagent Co., Ltd. (Shanghai, China). Nafion solution (5 wt%) was purchased fromSigma-Aldrich. All the chemicals were used as received without further purification. All aqueoussolutions were prepared using deionized water with a resistivity of 18.2 MΩ cm-1 .Material synthesis. The Cu2 O precursors were prepared by wet chemical reduction method.Typically, PVP (0 g for cube Cu2 O and 4 g for octahedral Cu2 O) was dissolved in 90 mL ofdeionized water. 10 mL of CuCl2 ·2H2 O (0.1 mol L-1 ) was added dropwise to the above solution,followed by heating in a water bath at 55 o C. After the addition of 10 mL NaOH solution (2.0 molL-1 ), the mixture was stirred for 30 min. Then 10 mL of ascorbic acid solution (0.6 mol L-1 ) wasadded dropwise. The reaction solution was further stirred at 55 o C for 3 h. The resulting precipitatewas separated by centrifugation and washed with water and ethanol for three times. Finally, theobtained Cu2 O samples were dried under vacuum at 60 o C for 12 h. The obtained Cu2 O samples were soaked in 10 mL of RuCl3 solution for 10 min, followed bythe centrifugation to prepare Ru-doped Cu2 O. Then the samples were rinsed with water and driedat 60 o C. 4 mg of catalysts were dispersed in 2 mL of ethanol by sonication for 1 h, followed bythe addition of Nafion solution (40 µL) and sonication for 30 min. Finally, the uniform mixturewas loaded onto carbon paper with an area of 2×2 cm2 . Before electrochemical measurements, thecatalysts were reduced in 0.5 M Na2 SO 4 under the potential of -0.6 V vs RHE for 30 min to obtainRu1 /Cu(111) and Ru1 /Cu(100). The different Ru doping content were controlled by adjusting theconcentration of RuCl3 solution to 0.5, 1, 2, 10, and 50 μM, respectively.Synthesis of SHINs. 400 mL of 0.01% HAuCl4 ·4H2 O solution was added to a 500 mL round-bottom flask under stirring and heating. Once the HAuCl4 solution come to a boil, 3 mL of 1% sodium citrate solution was promptly added. After heating for 30 min, the mixture was cooled toroom temperature with continuous stirring. Then, 30 mL of 55 nm Au nanoparticle colloida lsolution was taken out. Subsequently, 0.4 mL of a 1 mM (3-aminopropyl)trimethoxysila ne(APTMS) solution was added to the Au sol at room temperature under stirring for 15 min. Then,3 mL of 0.54 wt% sodium silicate solution was introduced, and the resulting sol was transferred toa 95 °C water bath and stirred for 30 min, yielding Au nanoparticles coated with a silica shell ofapproximately 2 nm in thickness.The preparation of SHINs film on the catalysts. SHINs arrays were constructed at liquid- o ilinterfaces according to the previous literature. 1 In a typical synthesis, 10 mL of SHINs dispersionwas purified by centrifugation and re-dispersed in 5 mL of water, followed by the addition ofcyclohexane (5 mL) and of 0.1 mM TBAN (200 μL). The mixture was vigorously shaken for 3min to complete the self-assembly process. After aging for 10 min, the upper cyclohexane wasremoved. The mixture was poured immediately into a petri dish. The Au nanoparticle arrays wouldstay on the surface of the water. The glass carbon loaded with catalysts was dipped into themonolayer film of SHINs floating on the water at a small angle (5-10°) and pulled out slowly,leading to the formation of SHINs-catalyst nanocomposites. By replacing the SHINs with anaqueous dispersion of Cu2 O particles and following the same procedure, a monolayer film of Cu2Oparticles can be fabricated. The SHINs film was prepared on a gold-sputter Si wafer to obtain thecross-sectional SEM images.Material characterization. SEM images and EDS elemental mapping images were taken on acold field emission scanning electron microscope (SU8220) operating at 5 kV. HAADF-STEMand the corresponding EDS elemental mapping were carried out on a Talos F200X field-emiss io ntransmission electron microscope operated at an accelerating voltage of 200 kV using Mo-basedTEM grids. XRD patterns were collected using a Rikagu MiniFlex X-ray diffractometer with Cu-Kα radiation (λ = 1.54059 Å). ICP-MS (Thermo Fisher iCAP RQ) analysis was employed tomeasure the concentration of metal species. The Raman spectra were conducted via LabRAM HREvolution (Horiba) Raman system with a 532 nm excitation laser. The absorbance data wasmeasured on a UV-vis spectrophotometer (Agilent Technologies, Cary 60). XANES and EXAFSspectra at the Ru K-edge and Cu K-edge were recorded at 1W1B station in Beijing Synchrotron Radiation Facility (https://cstr.cn/31109.02.BSRF.1W1B) and BL11B station in ShanghaiSynchrotron Radiation Facility (https://cstr.cn/31124.02.SSRF. BL11B), respectively. Theproduction of H2 was measured using an on-line gas chromatograph (GC2060) equipped with athermal conductivity detector.Electrochemical OH- adsorption. Electrochemical OH- adsorption were carried out in Ar-saturated 1 M KOH electrolyte. The CV measurements was performed between 0 and +0.6 V vsRHE with a sweep rate of 100 mV s-1 .Pb Underpotential Deposition (UPD). Pb UPD experiments were carried out in Ar-saturated 0.1M HClO 4 aqueous solution containing 10 mM Pb(ClO 4 )2 . First, the sample was hold for 150 sunder -0.15 V vs RHE. Then, the CV measurements was performed from -0.10 to +0.25 V vs RHEwith a sweep rate of 10 mV s-1 .In situ SHINERS study of hydrogenation reaction. The Raman spectra were obtained using aconfocal microscope Raman system (Horiba LabRAM HR Evolution). All Raman measureme ntsutilized a 633 nm excitation wavelength and a 50× microscope objective with a numerical apertureof 0.55. Prior to measurements, the Raman shift range was calibrated against the 520.6 ± 0.5 cm-1peak of silicon. A homemade Teflon Raman cell was used for the in situ electrochemical Ramanmeasurements with a 3 mm glassy carbon electrode, a Pt wire, and an Ag/AgCl electrode as theworking, counter, and reference electrodes, respectively. Before the measurements, the electrodeloading with SHINs-catalyst nanocomposites was placed in 5 mL of ethanol containing 1.0 mMpNTP and incubated at room temperature for 30 min. Before the in situ Raman experiments, theelectrodes adsorbed pNTP were washed several times by ethanol to remove the free molecules onthe surface. The potential was controlled by an electrochemical workstation (CHI 660E, ShanghaiCH Instruments).Estimation of hydrogen spillover distance and site density. The calculated hydrogen spillo verdistance (d) was derived from the conversion of pNTP (w) and site density (ns) of Ru atoms. Inthis model, we assume that the conversion is positively related with d and ns. The calculatio nequation can be expressed as w = 4 × d2 × ns / 100. The ns represents the number of Ru sites per 100 nm2 , which is estimated from the ICP of Ru in the catalysts. The calculation equation of ns canbe expressed as ns = VCu × ρCu × LRu × NA × 100 / (MRu × SCu)where VCu and ρCu represent the volume and mass density of one Cu nanoparticle, LRu representsthe loading content of Ru, NA is the avogadro’s constant, MRu is the relative atomic mass of Ru,and SCu represents the surface area of one Cu nanoparticle.Electrochemical measurements. The electrochemical measurements were carried out in an H-cell system which was separated by Nafion 115 membrane with a CHI660E electrochemica lworkstation (Chenhua, Shanghai). The area of working electrodes used in the electrochemica lmeasurements was 0.25 cm2 with mass loading was calculated to be 1 mg cm-2 . Ag/AgCl electrodeand graphite rod were used as the reference electrode and counter electrode, respectively. For NO3-electroreduction, 0.5 M Na2 SO4 containing 0.1 M KNO 3 solution (60 mL) was evenly distributedto the cathode and anode compartments. All potentials were measured against the Ag/AgClreference electrode and converted to the RHE reference scale by E (vs RHE) = E (vs Ag/AgCl) +0.21 V + 0.0591 × pH. Before the electroreduction test, CV curves were performed until thepolarization curves achieved steady-state ones with a scan rate of 10 mV s-1 . Before the electrolys is,Ar gas was delivered into the cathodic compartment at a rate of 10 mL min-1 to remove dissolvedO2 . The LSVs of the catalysts were recorded at a scan rate of 5 mV s -1 in 0.5 M Na2 SO4 containing0.1 M KNO 3 . The controlled potential electrolysis was performed at applied potentials for 10 min.The calculation method for FE. The FE for NH3 was calculated at a given potential as follows: FE = C × V × N × F / (Q × M)C: the measured concentration of product (mg mL-1 ),V: the volume of the electrolyte (mL),N: the number of electrons transferred for the product, which is 8 for NH 3 ,F: Faraday constant, 96485 C mol-1 ,Q: total electric charge (C),M: the relative molecular mass, which is 17 g mol-1 for NH3 . The calculation method for the yield rate of NH3 product. The yield rate of NH3 product wascalculated at a given potential as follows: ν NH3 = (CNH3 × V ) / (S × t) × 60ν NH3 : the yield rate (mgNH3 h-1 cm-2 ),CNH3 : the measured concentration of NH3 (mg mL-1 ),V: the volume of the electrolyte (mL),S: the area of the catalyst (cm2 ),t: the reduction reaction time (min).Determination of NH3 concentration with indophenol blue method. After the electroreductio nprocess, a certain amount of electrolyte was taken out from the electrolytic cell and diluted to thedetection range. Then, 2 mL of 1 M NaOH solution containing salicylic acid (5 wt%) and sodiumcitrate (5 wt%) were added into the aforementioned solution, followed by the addition of 1 mL of0.05 M NaClO and 0.2 mL of C5 FeN 6 Na2 O (1 wt%). After standing in darkness for 2 h, theabsorption spectra were measured using a UV-vis spectrophotometer. The concentration ofindophenol blue was determined using absorbance at the wavelength of 650 nm. Theconcentration-absorbance curve was calibrated using standard (NH4 )2 SO 4 solution with a series ofconcentrations.EPR measurements. EPR measurements were carried out using a Bruker EMX plus-9.5/12.Before the electrolysis, Ar gas was delivered into the cathodic compartment (containing 15 mL ofelectrolyte) at a rate of 10 mL min-1 to remove dissolved O 2 . The electrocatalytic test on catalystswas conducted in 0.5 M Na2 SO4 solution with or without 0.1 M KNO 3 at -0.3 V vs RHE for 3 min.After that, 50 L of DMPO (Dojindo) was added into the electrolyte. Then 1 mL of electrolyte wastaken and immediately transferred to a capillary for detection.DFT calculations. The first-principles calculations based on spin-polarized density functio na ltheory (DFT) were performed using the Vienna ab initio simulation package (VASP) version 5.4.4.2 Exchange-correlation effects were treated within the generalized gradient approximatio n(GGA) using the Perdew-Burke-Ernzerhof (PBE) exchange-correlation functional, and theelectron-ion interactions were described using the projector augmented wave (PAW)pseudopotentials.3, 4 A plane-wave basis set with an energy cutoff of 450 eV was adopted, and Γ-centered k-point sampling with a 3 × 3 × 1 mesh was implemented for Brillouin zone integratio n.The convergence criteria of structure optimization were chosen as the maximum force on eachatom less than 0.02 eV/Å with an energy change less than 1 × 10 −5 eV. Van der Waals interactio nswere corrected using Grimme’s DFT-D3 method.5 The activation barrier for each reaction wasobtained from climbing- image nudged elastic band (CI-NEB) calculations.6 Figure S1. SEM images of (a) octahedral Cu2 O and (b) cubic Cu2 O. (c) XRD patterns Cu2Oprecursors. Figure S2. XRD patterns of Ru1 /Cu(111) and Ru1 /Cu(100). Figure S3. EXAFS fitting results at the Ru K-edge of (a) Ru1 /Cu(111) and (b) Ru1 /Cu(100). Table S1. EXAFS fitting results at the Ru K-edge of Ru1 /Cu(111) and Ru1 /Cu(100). Sample Shell CN R (Å) σ2 E0 (eV) Ru-O 3.2 0.6 2.03 0.03 0.003 5.0 Ru1 /Cu(111) Ru-Cu 1.0 0.2 2.64 0.03 0.005 -3.2 Ru-O 3.6 0.7 2.03 0.03 0.005 3.1 Ru1 /Cu(100) Ru-Cu 2.0 0.4 2.62 0.03 0.007 -4.7CN represents coordination number, R represents bond distance, σ2 represents Debye-Waller factor,and E0 represents edge-energy shift. The apparently low Ru-Cu coordination numbers was causeby the inevitable surface oxidation of catalysts after exposure to the air during the XAFSmeasurements. It has been reported that Cu surfaces, including those decorated with atomicallydispersed noble metals, are highly susceptible to rapid oxidation under ambient conditions 7 . Suchsurface oxidation can partially block Ru-Cu interfacial bonding and consequently reduce theapparent metal-metal coordination number, which is consistent with the previous studies 8 . Apartfrom the Ru-Cu coordination, the Ru-O coordination was observed for Ru1 /Cu catalysts due to thestrong oxophilicity of Ru atoms. The Ru-Cu and Ru-O coordination contributed together to thestructural characteristic of atomically dispersed Ru species on air-exposed Cu surfaces. Figure S4. Wavelet transformed EXAFS spectra of (a) Ru1 /Cu(111), (b) Ru1 /Cu(100), (c) Ru foil,and (d) RuO 2 . Figure S5. In situ Cu K-edge XANES spectra for Ru-doped octahedral Cu2 O. Figure S6. In situ Cu K-edge (a) XANES and (b) EXAFS spectra for Ru-doped cubic Cu2 O. Figure S7. The SEM images of monolayer SHINs on (a) Au and (b) Pt electrodes. Figure S8. (a) Cross-section SEM and (b) typical SEM images of SHINs-assembled Ru1 /Cu(100)nanocomposites. Figure S9. HAADF-STEM and corresponding EDS mapping of SHINs-assembled Ru1 /Cu(100). Figure S10. The Raman signal of pNTP under different conditions. Figure S11. In situ Raman spectra of hydrogenation of pNTP on SHINs-assembled (a) Ru1 /Cu(100)and (b) Cu(100) nanocomposites. Figure S12. SEM images of (a) Cu(111) and (b) Cu(100). Figure S13. (a) The Raman spectra of pNTP under different concentrations. (b) The in situ Ramanspectra of pNTP hydrogenation under 0.1 mM pNTP. (c) The conversion of pNTP under differe ntconcentrations. To explore the influence of adsorbed pNTP on hydrogen spillover, we investigated theconversion of pNTP with a lower coverage by decreasing the concentration of pNTP solution from1 mM to 0.1 mM. As shown in Figure S13a, the signal intensity of 0.1 mM pNTP was significa ntlyweaker than that of 1 mM pNTP, indicating the lower coverage of pNTP molecules on the catalystsurface. Regardless of the concentration of pNTP, the emergence of pATP were both observed atthe potential of -0.1 V vs RHE for Ru1 /Cu(111) (Figs. 2a and S13b). Besides, the conversions ofpNTP at all the applied potentials were nearly equal under the two different concentrations (Fig.S13c), demonstrating that the coverage of pNTP had no significant impact on hydrogen spillo verdistance. Figure S14. The Raman spectra of pNTP on Ru1 /Cu(100) without the applied potential. Table S2. The loading content of Ru in Ru1 /Cu(111) and Ru1 /Cu(100). Sample Loading content (wt%) ns 0.001 3.6 0.0013 4.7 Ru1 /Cu(111) 0.0016 5.8 0.009 32.7 0.046 167.2 0.001 4.5 0.0014 6.2 Ru1 /Cu(100) 0.0016 7.1 0.011 48.9 0.048 213.6 Figure S15. The Raman spectra of pNTP hydrogenation on (a) Ru1 /Cu(111) and (b) Ru1 /Cu(100)with a sufficient surface Ru site density (ns = 32.7 and 48.9 for Ru1 /Cu(111) and Ru1 /Cu(100),respectively). (c) The conversion of pNTP at different reaction time. Figure S16. Free energy diagram of the hydrogen spillover over Ru1 /Cu(100) following the Cu-top migration pathway. The yellow, purple, red, and white spheres represent Cu, Ru, O, and Hatoms, respectively. The asterisk represents an adsorption site. Figure S17. The (a) UV-vis curves and (b) concentration-absorbance curve of NH4 + solution witha series of standard concentrations. The standard curve showed linear relation of absorbance withNH4 + concentration (y = 0.1169x + 0.0076, R2 = 0.9997). Figure S18. The FE for H2 of Ru1 /Cu(100) with different ns values. Figure S19. The structure models of intermediates for Cu(111). The yellow, blue, red, and whitespheres represent Cu, N, O, and H atoms, respectively. The asterisk represents an adsorption site. Figure S20. The structure models of *NOH and *N with H2 O. The yellow, blue, red, and whitespheres represent Cu, N, O, and H atoms, respectively. When the adjacent active *H slightlydeviated from its equilibrium position on Cu(111) surface, the O atom in *NOH intermediateswould spontaneously bind with *H after the structure optimization, generating H2 O and *N with aquite low energy barrier of -1.10 eV. Figure S21. The structure models of *N and *NH. The yellow, blue, and white spheres representCu, N, and H atoms, respectively. In the presence of *N species and *H, the adjacent *H wouldreadily bind with *N species on the surfaces and spontaneously transformed into *NH after thestructure optimization. References(1) Wei, J.; Qin, S.-N.; Liu, J.-L.; Ruan, X.-Y.; Guan, Z.; Yan, H.; Wei, D.-Y.; Zhang, H.; Cheng, J.; Xu, H.; Tian, Z.-Q.; Li, J.-F. In situ raman monitoring and manipulating of interfac ia l hydrogen spillover by precise fabrication of Au/TiO 2 /Pt sandwich structures. Angew. Chem. Int. Ed. 2020, 59 (26), 10343-10347.(2) Kresse, G.; Furthmüller, J. Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set. Comp. Mater. Sci. 1996, 6 (1), 15-50.(3) Blöchl, P. E. Projector augmented-wave method. Phys. Rev. B 1994, 50 (24), 17953-17979.(4) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized gradient approximation made simple. Phys. Rev. Lett. 1996, 77 (18), 3865.(5) Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. A consistent and accurate ab initio parametrization of density functional dispersion correction (DFT-D) for the 94 elements H- Pu. J. Chem. Phys. 2010, 132 (15), 154104.(6) Henkelman, G.; Uberuaga, B. P.; Jónsson, H. A climbing image nudged elastic band method for finding saddle points and minimum energy paths. J. Chem. Phys. 2000, 113 (22), 9901- 9904.(7) Ji, K.; Xu, M.; Xu, S. M.; Wang, Y.; Ge, R.; Hu, X.; Sun, X.; Duan, H., Electrocatalytic hydrogenation of 5-hydroxymethylfurfural promoted by a Ru1 Cu single-atom alloy catalyst. Angew. Chem. Int. Ed. 2022, 61(37), e202209849.(8) Zheng, T., Liu, C., Guo, C.; Zhang, M.; Li, X.; Jiang, Q.; Xue, W.; Li, H.; Li, A.; Pao, C.-W.; Xiao, J.; Xia. C.; Zeng, J., Copper-catalysed exclusive CO 2 to pure formic acid conversion via single-atom alloying. Nat. Nanotechnol. 2021, 16(12), 1386-1393.