Alumina‐Supported Alpha‐Iron(III) Oxyhydroxide as a Recyclable Solid Catalyst for CO2 Photoreduction under Visible Light

Abstract Photocatalytic conversion of CO2 into transportable fuels such as formic acid (HCOOH) under sunlight is an attractive solution to the shortage of energy and carbon resources as well as to the increase in Earth's atmospheric CO2 concentration. The use of abundant elements as the components of a photocatalytic CO2 reduction system is important, and a solid catalyst that is active, recyclable, nontoxic, and inexpensive is strongly demanded. Here, we show that a widespread soil mineral, alpha‐iron(III) oxyhydroxide (α‐FeOOH; goethite), loaded onto an Al2O3 support, functions as a recyclable catalyst for a photocatalytic CO2 reduction system under visible light (λ>400 nm) in the presence of a RuII photosensitizer and an electron donor. This system gave HCOOH as the main product with 80–90 % selectivity and an apparent quantum yield of 4.3 % at 460 nm, as confirmed by isotope tracer experiments with 13CO2. The present work shows that the use of a proper support material is another method of catalyst activation toward the selective reduction of CO2.

Preparation of FeOOH/Al2O3. Prior to use, Al2O3 was calcined at 1573 K for 1 h in air. The specific surface area of the treated Al2O3 was 13 m 2 g −1 . The Al2O3 powder was dispersed in 2 mL of water containing an appropriate amount of Fe(NO3)3·9H2O in an evaporating dish. The resultant suspension was stirred using a glass rod until the water was completely evaporated. The obtained powder was collected and heated in a H2 stream (20 mL min −1 ) at 473 K for 1 h. The loading amount of Fe was 10.0 wt% unless otherwise stated. In addition to Al2O3, other oxides were also tested as supports.
Characterization. X-ray diffraction (XRD) patterns were acquired using a Rigaku MiniFlex600 powder diffractometer equipped with a monochromatic Cu Kα radiation source. Scanning electron microscopy (SEM) images were acquired using a Hitachi SU9000 field-emission scanning electron microscope and a Jeol JSM-IT100LA microscope equipped with an energy-dispersive X-ray spectroscopy (EDS) apparatus. UV-visible absorption and diffuse-reflectance spectra were recorded using spectrophotometers (V-670 and V-565, JASCO). Steady-state emission spectra were measured at room temperature under an Ar atmosphere using and spectrofluorometer (Fluorolog-3-21, Horiba and FP-8600, Jasco). N2 and CO2 adsorption measurements were conducted using a BELSORP-MAX II (MicrotracBEL) apparatus at liquid-N2 temperatures (77 K) and 298 K, respectively. The samples were heated at 373 K for 1 h under reduced pressure before adsorption measurement. 13 C nuclear magnetic resonance ( 13 C NMR) spectra were acquired using a JEOL ECA400II (400 MHz) NMR spectrometer. XPS spectra were acquired using an ESCA-3400 X-ray photoelectron spectrometer (Shimadzu). The binding energies were calibrated by referencing the C 1s peak (285.0 eV) for each sample.
X-ray absorption fine structure (XAFS) measurements were conducted on the BL12C beamline of the Photon Factory at KEK (Inter-University Research Institute Corporation High Energy Accelerator Research Organization; Proposal No. 2020G597) using an electron energy of 2.5 GeV with an average current of 450 mA. The X-ray absorption spectra were acquired in transmission mode at room temperature using a Si(111) two-crystal monochromator. A pair of Ni-coated mirrors was used to eliminate higher harmonics. The X-ray absorption near-edge structure (XANES) spectra and extended X-ray absorption fine structure (EXAFS) oscillation were analyzed using the Athena software package. [2] Photocatalytic Reactions. Reactions were performed at room temperature using an 8 mL test tube that contained 4 mg of catalyst powder and 4 mL of DMA solution containing 1.0 mM [Ru(bpy)3]Cl2 and 0.1 M BNAH. Prior to irradiation, the suspension was purged with CO2 (Taiyo Nippon Sanso, >99.995%) for 30 min. A 400 W high-pressure Hg lamp (SEN lights Co.) was used as a light source, in combination with an aqueous NaNO2 solution to allow for visible-light irradiation (λ > 400 nm). The formate generated in the liquid phase was analyzed using a capillary electrophoresis system (Otsuka Electronics, Agilent 7000). The gaseous reaction products were analyzed using a gas chromatograph equipped with a thermal conductivity detector (TCD) (GL Science, model GC323) and an activated carbon column; Ar was used as the carrier gas.
To investigate the effect of residual HCOOH in the as-prepared α-FeOOH/Al2O3, the powder sample (4 mg) was stirred in a 0.1 M NaOH solution (4 mL) for 3 h in the dark, and the supernatant solution was examined by capillary electrophoresis. The results showed that the supernatant contained 0.06 ± 0.008 µmol of HCOOH, which is much smaller than the amount produced by the photocatalytic reaction using α-FeOOH/Al2O3.
The turnover number (TON) for HCOOH production was calculated as TON = Amount of products / Amount of Fe atom in the catalyst. (1) The apparent quantum yield (AQY) for HCOOH formation was measured using a 300 W Xe lamp (Asahi Spectra, MAX-303) fitted with a band-pass filter (460 nm) and was estimated as where R and I represent the rates of HCOOH production and incident photons, respectively. The total number of incident photons (12.1 mW) was measured using a spectroradiometer (Eko Instruments, LS-100). For the measurement, 8 mg of the α- Selectivity toward HCOOH production during the CO2 reduction reaction was calculated on the basis of the ratio between the amount of HCOOH generated and the total amount of reduction products (i.e., HCOOH, CO, and H2): Selectivity to HCOOH / % = HCOOH produced / Reduction products × 100, Isotope Tracer Experiment. 13 CO2 ( 13 C 99%, Watari CO., Ltd.) was purchased from Sigma-Aldrich. The 13 CO2 gas was introduced into a DMA solution (2 mL) containing 1.0 mM [Ru(bpy)3](PF6)2 and 0.1 M BNAH, along with 4 mg of photocatalyst powder, after the liquids were degassed through freeze-pump-thaw cycling. After the samples were irradiated, the gas phase was analyzed using a gas chromatograph-mass spectrometer (Shimadzu, QP-2010-Ultra) equipped with a Molsieve5A capillary column. Prior to the measurement, no contamination of 13 CO and H 13 COOH in the 13 CO2 gas was detected. The liquid phase was analyzed by 13 C NMR. Before NMR analysis, the reacted solution was mixed with 0.1 M NaOH-D2O solution in a volume ratio of 1:1. The solution was further diluted with 2 mL D2O. After mixing, the solution was filtered using a micropore filter and the filtrate was used for NMR measurement. CD3CN (>99.9%, Kanto Chemical) was used as an internal standard at 118.  Figure S1. Steady-state emission spectra of [Ru(bpy)3] 2+ in DMA solution containing catalysts and/or BNAH at room temperature. The excitation wavelength was 532 nm.

Results and Discussion
As shown in Figure S1, a clear emission peak was observed at around 640 nm, which originates largely from the lowest 3 MLCT excited state of [Ru(bpy)3] 2+ . In the presence of Al2O3, the emission quantum yield of [Ru(bpy)3] 2+ without BNAH was 9.4%. When α-FeOOH/Al2O3 existed, the emission quantum yield was 4.4%, which was lower than that obtained with Al2O3. This is probably because α-FeOOH absorbs the excitation light (532 nm), thereby decreasing the number of photons that are supposed to be absorbed by [Ru(bpy)3] 2+ . This also makes it difficult to precisely measure the emission quantum yield. Adding BNAH in the solution resulted in a very low emission quantum yield of ~0.1% in both systems. The apparent quenching efficiency, which is defined by the ratio of the emission quantum yield with BNAH to that of without BNAH, was 99 and 98% for Al2O3 and α-FeOOH/Al2O3, respectively. The values are very close to the quenching efficiency of [Ru(bpy)3] 2+ measured in a DMA/BNAH mixed solution (98%). [3] Summarizing above results, it is concluded that almost all [Ru(bpy)3] 2+ are quenched reductively by BNAH and the one-electron-reduced species of [Ru(bpy)3] 2+ donates an electron to the catalysts.  Wavelength / nm  As displayed in Figure S4, the catalyst sample shows two peaks at 711.4 and 725.2 eV, which are from Fe 2p3/2 and for Fe 2p1/2 photoelectrons, respectively. The peak positions are consistent with those reported previously. [4] The peak position of the catalyst sample in the Al 2s XPS (119.0 eV) is very close to that of Al2O3 (118.9 eV). [5]

ɑ-FeOOH
Fe/Al 2 O 3  As shown in Figure S6, the solution before reaction exhibits a clear 1 MLCT absorption band at 400-500 nm region that is typical of Ru, although an additional absorption extending to longer wavelengths, which arises from the absorption of BNAH, [6] appears. The 1 MLCT absorption band of Ru became weaker and shifted to longer wavelengths after the reaction. This spectral change before and after the reaction is attributable to photochemical ligand substitution of the Ru(II) photosensitizer unit giving the corresponding Ru(II) bisdiimine-type complex(es). [7] The reversible deactivation of the α-FeOOH/Al2O3 system during the CO2 reduction ( Figure 2b) could therefore be explained in terms of the decomposition and/or structural change of Ru.  1 M), and 10 mg of α-FeOOH/Al2O3, which was measured after filtration. Data for HCOOH reference is also shown. A clear peak at δ = 166.1 ppm is assignable to H 13 COOH. [8] (b) Gas chromatograms of the gas-phase products, as obtained using a mass spectrometer as a detector (m/z 28 and 29). The photocatalyst suspension was subject to visible-light irradiation from a 400 W high-pressure Hg lamp with a NaNO2 solution filter for 15 h under 13 CO2 (610 Torr) and under saturated, unlabeled CO2.