Sustainable all-weather CO2 utilization by mimicking natural photosynthesis in a single material

ABSTRACT Solar-driven CO2 conversion into hydrocarbon fuels is a sustainable approach to synchronously alleviating the energy crisis and achieving net CO2 emissions. However, the dependence of the conversion process on solar illumination hinders its practical application due to the intermittent availability of sunlight at night and on cloudy or rainy days. Here, we report a model material of Pt-loaded hexagonal tungsten trioxide (Pt/h-WO3) for decoupling light and dark reaction processes, demonstrating the sustainable CO2 conversion under dark conditions for the first time. In such a material system, hydrogen atoms can be produced by photocatalytic water splitting under solar illumination, stored together with electrons in the h-WO3 through the transition of W6+ to W5+ and spontaneously released to trigger catalytic CO2 reduction under dark conditions. Furthermore, we demonstrate using natural light that CH4 production can persist at night and on rainy days, proving the accomplishment of all-weather CO2 conversion via a sustainable way.


Section 1. Preparation of h-WO 3
To prepare h-WO 3 nanorods, 0.99 g Na 2 WO 4 • 2H 2 O and 1.19 g NaHSO 4 •H 2 O were dissolved in 40 mL deionized water with constant stirring.After stirring for 1 h, the mixture solution was then transferred into a 100 mL autoclave and heated in oven at 180°C for 24 h.After reaction, the precipitates were collected by centrifugation, and washed with deionized water and ethanol for several times.The powder was obtained and dried in vacuum at 70°C overnight (noted as h-WO 3 ).

Section 2. Preparation of Pt/h-WO 3
Pt-modified h-WO 3 was obtained by in situ reduction of Pt by low-valence W 5+ .Briefly, 0.2 g h-WO 3 and 100 mL water were added into a 200 mL quartz closed reactor, and then ultrasonically treated for 30 min to form a uniform light-white mixture solution.After the solution mixture was purged by Ar for 1 h, a 500 W xenon lamp was used as the light source to irradiate the solution for 1 h.The color of the mixture solution changed into light blue after 1 hour of illumination, indicating that the reduced W 5+ was formed.Then 10 mL of H 2 PtCl 6 • xH 2 O (2 mg/mL) was dropped into the obtained solution with constant stirring.After stirring for 1 min, the sample was collected and dried in vacuum at 70°C overnight (noted as Pt/h-WO 3 ).

Section 3. Synthesis of h-WO 3 loaded with Pt nanoparticles
The h-WO 3 loaded with Pt nanoparticles (Pt NPs) was prepared by photo-deposition.
Briefly, other conditions were the same as the preparation method of Pt/h-WO 3 , except that the order of addition of H 2 PtCl 6 • xH 2 O and light irradiation was switched.

Section 4. Synthesis of Au-loaded h-WO 3
The preparation process of Au-loaded h-WO 3 was similar to that of Pt-loaded h-WO 3 .
Briefly, 0. 2 g h-WO 3 and 100 mL water were added into a 200 mL quartz closed reactor, and then ultrasonically treated for 30 min to form a uniform light-white mixture solution.After the solution mixture was purged by Ar for 1 h, a 500 W xenon lamp was used as the light source to irradiate the solution for 1 h.The color of the mixture solution changed into light blue after 1 hour of illumination, indicating that the reduced W 5+ was formed.Then 10 mL of H 2 AuCl 4 • 3H 2 O (2 mg/mL) was dropped into the obtained solution with constant stirring.After stirring for 1 min, the sample was collected and dried in vacuum at 70°C overnight (noted as Au/h-WO 3 ).The Au loading content of Au/h-WO 3 is 0.15 wt% as measured by inductively coupled plasma-mass spectrometry (ICP-MS).

Section 5. Synthesis of Cu-loaded h-WO 3
The preparation method of Cu-loaded h-WO 3 (Cu/h-WO 3 ) was the same as that of Pt/h-WO 3 , except that H 2 PtCl 6 • xH 2 O was replaced by Cu(NO 3 ) 2 • 3H 2 O.The Cu loading content of Cu/h-WO 3 is 0. 17 wt% as measured by ICP-MS.

Section 6. Synthesis of Ni-loaded h-WO 3
Since Ni is difficult to be reduced by W 5+ , we extended the reaction time under dark conditions.Firstly, 0. 2 g h-WO 3 and 100 mL water were added into a 200 mL quartz closed reactor, and then ultrasonically treated for 30 min to form a uniform lightwhite mixture solution.After the solution mixture was purged by Ar for 1 h, a 500 W xenon lamp was used as the light source to irradiate the solution for 1 h.The color of the mixture solution changed into light blue after 1 hour of illumination, indicating that the reduced W 5+ was formed.Then 10 mL of (Ni(NO 3 ) 2 • 3H 2 O) (2 mg/mL) was dropped into the obtained solution with constant stirring.After stirring for 24 h, the sample was collected and dried in vacuum at 70°C overnight (noted as Ni/h-WO 3 ).
The Ni loading content of Ni/h-WO 3 is 0.21wt% as measured by ICP-MS.
Steady-state photoluminescence (PL) spectroscopy (FLS980, Edinburgh, England) was employed to investigate the lifetime of the photo-induced charges and the optical properties of the resulting samples.Nitrogen adsorption-desorption isotherms and specific surface area were measured by gas adsorption analysis system (Quantachrome Autosorb-IQ, USA).

Section 8. Characterizations of EPR spectroscopy
EPR spectra were recorded on a ELEXSYS E500 spectrometer (Bruker, German) at room temperature.The reaction process was similar to that of CO 2 reduction measurement.After a certain time, a strong vacuum pump was used to exclude the water vapor above the solution, so that the water in the reaction chamber was quickly evaporated.After all the water was removed, the sample was transferred to a quartz tube (inner diameter 3 mm) under N 2 protection and the mouth of the tube was closed with a plug to avoid contact with air.Finally, the quartz tube was put into the instrument for EPR characterization.

Section 9. Characterizations of FTIR spectroscopy
FTIR characterization was performed on a VERTEX 70 FTIR spectrometer (Bruker, Germany) equipped with a Harrick in situ diffuse reflectance cell and liquid nitrogen cooled MCT (Mercury Cadmium Telluride) detector.For the CO adsorption, all the samples were preheated at 200 C for 120 min to obtain a clean surface.The baseline for the spectra was recorded after cooling to room temperature.Subsequently, 1% CO/Ar was poured into the cell at 10 mL/min for 30 min.The cell was then purged with N 2 for 30 min to completely remove gaseous CO, and the FTIR spectrum was recorded at the end of the purging.For the in-situ CO 2 conversion measurement, after the light reaction, all the water was extracted by a mechanical pump.Subsequently, the catalyst was transferred to the in situ diffuse reflectance cell under nitrogen protection.Finally, 5% CO 2 /N 2 with the flow rate of 10 mL/min was introduced through a container filled with water, and FTIR data was collected every 1 min in dark.

Section 10. Characterizations of light-assisted Kelvin probe force microscopy (KPFM)
Atomic force microscope (AFM) topography images and light-assisted Kelvin probe force microscopy (KPFM) were collected by AFM (Bruker MultiMode-8 surface potential mode) at ambient conditions.A SCM-PIT-V2 model tip and AS-130VLR ("J" vertical) scanner model was used, and the radius and elastic coefficient of tip were 35 nm and 3 N/m, respectively.The tip lift height was 60 nm for potential mapping at tapping mode.A 300 W Xe arc lamp irradiated at the sample to provide light irradiation.In the process of testing, the samples were coated on highly oriented pyrolytic graphite (HOPG) substrate.The contact potential difference (CPD) was defined as the difference between the work function of the tip and the sample [1].The images were processed by first order flattening to eliminate errors caused by sample tilt.The surface photovoltage (SPV) was calculated as SPV = ΔCPD = CPD dark − CPD light , where CPD dark and CPD light are the CPD measured in dark and under light illumination, respectively.

Section 11. Measurements of the amount of stored H
The amount of stored H was measured by means of ion exchange.10 mg of photocatalyst powder was dispersed in 60 mL of deionized water and then ultrasonically dispersed for 30 min.Subsequently, the air in the reactor was removed by vacuum pump for 10 min while the pH of the solution was close to 7. Then the simulated sunlight was produced by a 300 W Xe arc lamp irradiated on the sample through quartz window for 10 min.After illumination, excessive KNO 3 solution (0.5 M, 4 mL) was dropped into the obtained solution in the dark with constant stirring.
After stirring for 3 h until the pH no longer changed, the sample was separated by centrifugation, and the pH of the separated solution was measured.This value was used to evaluate the amount of released H through the following equation: where M(H) is the amount of hydrogen atoms (μmol).

Section 12. Characterizations of photoelectrochemical properties
The photoelectrochemical properties of the as-prepared samples were collected using

a
Parstat 4000 electrochemical workstation (USA) in a three-electrode cell.Pt foil and saturated Ag/AgCl electrode were used as the counter electrode and reference electrode, respectively.The potential plots were measured at the open-circuit voltage in 0.2 M Na 2 SO 4 with phosphate buffer (pH = 7) as the electrolyte.The Mott-Schottky (M-S) plots were measured in the dark at fixed frequencies of 250, 500 and 750 Hz, respectively.Electrochemical impedance spectroscopy (EIS) was performed in 1 mM K 3 Fe(CN) 6 and K 4 Fe(CN) 6 solution.The photocurrent response spectra were measured at 0.2 V vs. Ag/AgCl in 0.5 M Na 2 SO 3 at room temperature.Cyclic voltammetry (CV) of Pt/h-WO 3 in 0.1 M H 2 SO 4 with a polished glassy carbon electrode (5 mm of diameter) as the working electrode.All solutions were purged with Ar or CO 2 before CV test.

Table S1
Comparison of the state-of-the-art works on photocatalytic CO 2 reduction into CH 4 .

Table S2
The calculated amount of H based on ion exchange using 10 mg catalyst.

Table S3
Weather conditions during the outdoor test.