Heterostructure Engineering of a Reverse Water Gas Shift Photocatalyst

Abstract To achieve substantial reductions in CO2 emissions, catalysts for the photoreduction of CO2 into value‐added chemicals and fuels will most likely be at the heart of key renewable‐energy technologies. Despite tremendous efforts, developing highly active and selective CO2 reduction photocatalysts remains a great challenge. Herein, a metal oxide heterostructure engineering strategy that enables the gas‐phase, photocatalytic, heterogeneous hydrogenation of CO2 to CO with high performance metrics (i.e., the conversion rate of CO2 to CO reached as high as 1400 µmol g cat−1 h−1) is reported. The catalyst is comprised of indium oxide nanocrystals, In2O3− x(OH)y, nucleated and grown on the surface of niobium pentoxide (Nb2O5) nanorods. The heterostructure between In2O3− x(OH)y nanocrystals and the Nb2O5 nanorod support increases the concentration of oxygen vacancies and prolongs excited state (electron and hole) lifetimes. Together, these effects result in a dramatically improved photocatalytic performance compared to the isolated In2O3− x(OH)y material. The defect optimized heterostructure exhibits a 44‐fold higher conversion rate than pristine In2O3− x(OH)y. It also exhibits selective conversion of CO2 to CO as well as long‐term operational stability.

The aqueous solution was ultra-sonicated for 30 min and then stirred for 15 min. The solution was subsequently placed in a Teflon-lined stainless steel autoclave with 100 mL capacity. The hydrothermal reaction was performed at T = 200 °C for 24 h. After cooling to room temperature, the white product was collected through a centrifugation process and washed three times with deionized water to remove non-reacted residues.
Finally, the sample was dried in a vacuum oven at T = 70 °C for 12 h. Nb 2 O 5 nanorods were obtained by calcining of as-prepared Nb 3 O 7 (OH) Nanorods at 400 o C at air for 6 h.

Synthesis of In(OH) 3 @Nb 2 O 5
In a typical procedure, 0.2 g of as-prepared Nb 2 O 5 nanorods were dispersed in 25 mL of ethanol with sonicating for 1 h. Subsequently, a varying amount of urea dissolved in 10 mL H 2 O was added, the mixed solution was further sonicated for 30 minutes. Finally, a varying amount of anhydrous InCl 3 (the molar ratio of urea to InCl 3 is fixed as 2:1) dissolved in 15 mL ethanol was added dropwise to above mixture solution and then heated at 80 °C in an oil bath under magnetic stirring for 12h. Following cooling to room temperature, the red products were collected though centrifugation and washed with water to remove non-reacted residues. The In(OH) 3 @Nb 2 O 5 precursors were obtained and dried for 48 h at ambient temperature.
The added urea amounts for S1, S2 and S3 are 0.96 g, 2.88 g and 3.8 g, respectively.

Synthesis of In 2 O 3-x (OH) y @Nb 2 O 5 Heterostructures
The as-synthesized In(OH) 3 @Nb 2 O 5 precursors were placed into an oven and treated at 350 °C in air for 3h to obtain the corresponding In 2 O 3−x (OH) y @Nb 2 O 5 samples.

Characterization
Powder X-ray diffraction was performed on a Bruker D2-Phaser X-ray diffractometer, using Cu Kα radiation at 30 kV. Nitrogen adsorption isotherms were obtained at 77 K using a Quantachrome Autosorb-1-C. The surface area of each sample was determined using BET theory, and pore size distributions were determined with NLDFT. The amount of In 2 O 3-x (OH) y. in heterostructures were quantified by inductively coupled plasma-atomic emission spectroscopy (ICP−AES, Thermo Electron Corp. Adv. ER/S). Before the ICP measurements, the heterostructures were immersed in concentrated nitric acid with agitation for 12 h to dissolve the In 2 O 3-x (OH) y . X-ray photoelectron spectroscopy (XPS) was performed using a PerkinElmer Phi 5500 ESCA spectrometer in an ultrahigh vacuum chamber with a base pressure of 1 × 10 -9 Torr. The spectrometer uses an Al Kα X-ray source operating at 15 kV and 27 A. The samples used in XPS analyses were prepared by drop-casting aqueous dispersions onto p-doped Si (100) wafers in the case of the In 2 O 3-x (OH) y . samples. All data analyses were carried out using the Multipak fitting program, and the binding energies were referenced to the NIST-XPS database and the Handbook of X-ray Photoelectron Spectroscopy [S2-S3] The carbon peak was calibrated as 285 eV. Thermogravimetric analysis was conducted by placing approximately 6 mg of samples on TA Instruments SDT Q600 thermogravimetric analyzer/differential scanning calorimeter in an alumina pan under 100 mL/min flow of compressed air. The temperature was steadily increased from room temperature to 700 °C at a rate of 5 °C/ min. The Fermi level position with respect to the vacuum level is determined by measuring the work function of a thin film sample. The work function is determined by the binding energy of the secondary electron cut-off (SECO) on the high binding energy side of the measured UPS spectrum. The work function is equal to the difference between photon energy and SECO binding energy. in air atmosphere at room temperature. In a typical measurement, the sample chamber was connected to an ITO glass at the top electrode and a steel substrate at bottom electrode, and a thick mica spacer was placed between the ITO glass and the sample to decrease the space charge region at the ITO-sample interface. The metal is a ground electrode that is not connected to a battery. The function of which is to ensure that the state surface photovoltage at the metal side is zero. Upon light irradiation, photoinduced carriers will diffuse to the surface of the sample. Adsorbed O 2 on the surface of the samples capture photogenerated electrons, while the photogenerated holes preferentially diffuse to the collector electrode surface, which manifested a positive TPV response. The samples were excited by a laser pulse at 355 nm or 532 nm with 10 ns width from the second harmonic of a neodymium-doped yttrium aluminum garnet (Nd:YAG) laser. The signals were amplified with a pre-amplifier, and registered with a 1 GHz digital phosphor oscilloscope.

Gas-Phase Catalytic Measurements
Gas-phase CO 2 reductive testing samples were prepared by drop casting catalysts (~3 mg) from an aqueous dispersion onto binder free borosilicate glass microfiber filters having an area of ~1.2 cm 2 (Whatman, GF/F, 0.7 µm). Two tests were then performed on catalyst samples and the duration time of each testing was 4 h: (1) in the dark at room temperature (RT), (2) under irradiation from a 300 W Xe lamp at 25 kW m -2 . The gas-phase catalytic measurements were conducted in a custom-built 1.8 mL stainless steel batch reactor with a fused silica view port sealed with a Viton O-ring.
The reactor was evacuated using an Alcatel dry pump prior to being purged with H 2 (99.9995%) at a flow rate of 20 mL min -1 . After purging, the reactor was filled with H 2 and CO 2 gas at a 1:1 pressure ratio to a total pressure of 27 psi prior to being sealed. The pressure inside the reactor was monitored during the reaction using an Omega PX309 pressure transducer. Reactors were irradiated with a 300 W Xe lamp for a duration of 4 h for each testing. The spectral output from the 300 W Xe lamp was measured using a StellarNet Inc spectrophotometer and the power of the incident irradiation was measured using a Spectra-Physics Power meter (model 407A).
Pressure and temperature inside the reactor after 4 h testing stabilized at 29-30 psi and 60°C, respectively. This is consistent assuming a closed system with no significant change of the number of moles due to the reaction and validity of the ideal gas law.
Note that the temperature of the catalyst is well above the gas temperature due to the photothermal effect. The sample temperature depends on the incident radiation, the design of the test cell, the operating conditions and the characteristics of the material.
The sample temperature has been investigated by Raman spectroscopy for other catalyst samples prior to this work, but not measured here. Yet due to the unchanged conditions a large deviation among the temperatures of the different samples discussed here is considered unlikely. Product gases were analyzed with a flame ionization detector (FID) installed in a SRI-8610 gas chromatograph (GC) with a 6'Haysep D column. The CO formation rate shown in Fig. 3 was obtained from the measured CO content in the gas after 4 h testing by dividing with the time and the catalyst mass. Note that the so defined CO formation rate is a time-averaged value.
According to kinetics and thermodynamics the CO formation rate declines with increasing conversion until it reaches zero when approaching the equilibrium. This means that the comparison of the time-averaged CO rate generally underestimates the true activity differences between the samples. The size of this effect increases with conversion. It will be particularly large if the system approaches equilibrium during testing. The experimental conversions obtained with samples S1 to S4 were much higher than calculated from the equilibrium of the reverse water gas shift reaction at the measured gas phase temperature. For example, S3 reached 3.7 % conversion, which can only be explained by a catalyst temperature of 155 °C or beyond or by an effect of the incident photons on the reaction equilibrium. Isotope tracing experiments were performed using 13 CO 2 (99.9 at%; Sigma Aldrich). The reactor was evacuated prior to being injected with 13 CO 2 followed by H 2 . Isotopically labeled product gases were measured using an Agilent 7890A gas chromatographic mass spectrometer (GC-MS) with a 60 m GS-Carbonplot column fed to the mass spectrometer.

Electronic structure calculations
Calculations on the crystal structures were carried out under periodic boundary conditions, using the projector-augmented wave (PAW) (S4), pseudopotential approach in the VASP (S5) code. The structures were initially relaxed using the PBEsol (S5) functional, with a cut-off energy of 500 eV and k-point mesh sampling density with a target length cut-off of 25 Å, as prescribed by Moreno and Soler (S6).
To obtain a quantitative electronic structure the resultant systems were then treated using the HSE06 (S7) functional, mixing 25% of screened exact exchange. Band alignment was performed by creating 2D slabs of each material and calculating the vacuum potential. The highest occupied and lowest unoccupied eigenvalues from bulk calculations were then placed relative to the vacuum potential, using the MacroDensity code (S8).    can be seen clearly that there is no PXRD peaks shifts of In(OH) 3 from S1' to S4' samples.       S10. Mass spectroscopy of S3 sample generated 13 CO from 13 CO 2 . The 28 AMU mass fragment peak at approximately 1.32 min corresponds to N 2 and the 29 AMU mass fragment peak at approximately 1.35 min corresponds to 13 CO. The fact that there is no peak near 1.35 min retention time for the 28 AMU curve shows that there is no 12 CO in the products generated from sources of adventitious 12 C.