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

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 Al2 O3 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 13 CO2 . The present work shows that the use of a proper support material is another method of catalyst activation toward the selective reduction of CO2 .

In recent years, CO2 emissions arising from the consumption of fossil fuels has become a serious problem. Among various proposed methods and schemes to address this problem, photocatalytic CO2 reduction is expected to be a key technology in the future. The production of transportable fuels such as formic acid (HCOOH) has attracted attention because they can act as energy carriers of hydrogen, which releases a high density of energy after combustion, without generating byproducts other than water.[ , ]

A photocatalytic CO2 reduction system typically consists of a light‐absorbing substrate (e.g., a molecular redox photosensitizer or semiconductor) and a catalyst (Scheme 1).[ , , ] Because the reduction of CO2 into value‐added chemicals involves the transfer of at least two electrons, a catalyst that enables multi‐electron transfer is necessary. Thus far, various metal complexes based on abundant metals such as MnI, FeII, CoI, and NiII have been reported as efficient catalysts for photocatalytic CO2 reduction.[ , ]

Scheme 1

A photochemical CO2 reduction system consisting of a catalyst and a photosensitizer. D indicates an electron donor. The catalyst may be in the form of a molecule or nanoparticle.

The potential recyclability of solid catalysts distinguishes them from homogeneous molecular catalysts, and the development of a catalyst that is sufficiently active, nontoxic, and inexpensive is an important subject in this research field. Recently, solid materials such as ZnCo2O4, NiCo2O4, Co3O4, CoSn(OH)6, LaCoO3, and metal–organic frameworks (MOFs) containing Co[ , , ] or Ni[ , ] have been reported to catalyze CO2 reduction in the presence of a RuII photosensitizer and an electron donor under visible light.

For future practical applications, Fe‐based solid catalysts have advantages over catalysts based on other early transition metals such as Co and Ni. Some Fe‐containing solids have recently been reported to function as CO2 reduction catalysts with [Ru(bpy)3]2+. For example, a composite derived from a Ni‐ and Fe‐containing MOF, further combined with SnO2, promoted the reduction of CO2 to CO. ZnFe2O4 nanoparticles grown in situ on the surface of iron porphyrin covalent triazine‐based frameworks and Co3O4/CoFe2O4 nanoparticles have been reported to exhibit a similar functionality, although the carbon source of the reaction product has not been fully clarified in these works because of the lack of 13CO2 experiments, which is one of the most important aspects in photochemical CO2 reduction studies (especially in heterogeneous systems). MOFs containing Fe2+ and another metal cation (e.g., Mn, Co, Ni, or Zn) have been reported to function as catalysts for CO2 reduction with [Ru(bpy)3]2+, giving CO as the main product with 75–85 % selectivity, as confirmed by a tracer experiment with 13CO2. As such, Fe‐based catalysts reported thus far produce CO as the main product; in addition, HCOOH‐generating Fe‐based solid catalysts for photochemical CO2 reduction have rarely been reported.

Here, we report that α‐FeOOH‐loaded Al2O3 can function as a recyclable catalyst for visible‐light‐driven CO2 reduction in the presence of a RuII photosensitizer ([Ru(bpy)3]2+, abbreviated as Ru]) and 1‐benzyl‐1,4‐dihydronicotinamide (BNAH) as an electron donor. Under these conditions, photoexcited Ru undergoes reductive quenching by reaction with BNAH, which was also confirmed in the present study (see Figure S1 and additional discussion). Therefore, CO2 reduction proceeds if a proper catalyst is present in the reaction system. In fact, the α‐FeOOH/Al2O3 catalyst produces HCOOH as the main product via CO2 reduction, with 80–90 % selectivity.

A catalyst sample, Fe‐loaded Al2O3, was prepared by a simple impregnation and H2‐reduction method. As evident in Figure 1a, the Al2O3 support was composed of well‐crystallized ≈1 μm particles, which form larger secondary particles with some interparticle space. Upon loading of Fe, the “interconnection” of particles was more pronounced and the interparticle space was reduced, accompanied by some surface roughening (see additional images in Figure S2). Energy‐dispersive X‐ray spectroscopy (EDS) measurements revealed that Fe species were well distributed on the Al2O3 surface (Figure 1b). The X‐ray diffraction (XRD) pattern of the Fe‐loaded Al2O3 was recorded to investigate the crystal structure of the Fe species in the sample. However, no diffraction peak other than those of Al2O3 was observed in the pattern of the Fe‐loaded Al2O3 (Figure S3), suggesting that the Fe species loaded onto Al2O3 was amorphous and/or in the form of a thin layer below the diffraction limit.

Figure 1

Characterization of Fe‐loaded Al2O3: a) SEM images and b) EDS mapping images. Fe K‐edge c) XANES spectra and d) EXAFS oscillation.

To identify the local structure of Fe species loaded onto Al2O3, X‐ray absorption fine structure (XAFS) measurements were conducted. Figure 1c shows the Fe K‐edge X‐ray absorption near edge structure (XANES) spectrum of the Fe‐loaded Al2O3, along with the spectra of α‐Fe2O3 and α‐FeOOH for reference. From the characteristic feature of the pre‐edge region (7113 eV), the Fe species loaded onto Al2O3 were found to be very similar to α‐FeOOH but different from α‐Fe2O3. FeOOH has several stable polymorphs, and α‐FeOOH is one of the most stable phases. In the present work, the existence of α‐FeOOH on the Al2O3 surface was also supported by the extended X‐ray absorption fine structure (EXAFS) oscillation. As shown in Figure 1d, the EXAFS oscillation of the Fe‐loaded Al2O3 was similar to that of the α‐FeOOH reference, although the oscillation was relatively weak at larger k regions, most likely because of the thin‐layer form of the loaded Fe species on the Al2O3. The existence of α‐FeOOH on Al2O3 was also supported by the result of X‐ray photoelectron spectroscopy (Figure S4). On the basis of these results, the prepared sample is hereafter represented as α‐FeOOH/Al2O3.

Photocatalytic CO2 reduction was performed at room temperature using α‐FeOOH/Al2O3 in an N,N‐dimethylacetamide (DMA)/BNAH mixed solution in the presence of Ru under visible light (λ>400 nm). As listed in Table 1, the α‐FeOOH/Al2O3 gave HCOOH as the main product; CO and H2 were also produced as secondary products. The CO2 reduction selectivity to HCOOH was 82 %. The apparent quantum yield (AQY) for HCOOH formation was 4.3 % at 460 nm. No reaction occurred in the dark or in the absence of α‐FeOOH (only Al2O3) (entries 2 and 3). No CO2 reduction product was obtained in the absence of either Ru, CO2, or BNAH (entries 4–6). As shown in Figure S5, the α‐FeOOH/Al2O3 showed semiconductor‐like absorption in the visible‐light region; however, this absorption was not responsible for the visible‐light CO2 reduction. Notably, Ru can undergo structural transformation during photochemical reaction to become catalytically active species for HCOOH production.[ , ] However, we confirmed that HCOOH production in the “blank” system was negligible under the present reaction conditions (entry 7).

Table 1

Results of visible‐light CO2 reduction experiments (λ>400 nm).[a]

[a] Reaction conditions: catalyst, 4 mg (Fe loaded 10.0 wt % to catalyst); solution, 4 mL DMA containing 1.0 mM Ru and 0.1 M BNAH; reaction time, 3 h. [b] In the dark. [c] Under an Ar atmosphere. n.d.=Not detected.

Figure 2a shows a typical time course of CO2 reduction using the α‐FeOOH/Al2O3 catalyst. The amount of HCOOH produced increased with reaction time, but changed little after 3 h. One of the most probable reasons for the deactivation is the decomposition and/or structural change of [Ru(bpy)3]2+ during the reaction (see Figure S6). Nevertheless, the performance of the α‐FeOOH/Al2O3 catalyst was found to be stable in consecutive runs without loss of the high selectivity to HCOOH (85–90 %), as shown in Figure 2b. The Fe‐based turnover number (TON) exceeded 6 after the five consecutive runs, confirming the catalytic cycle of the reaction. It was also confirmed by isotope tracer experiments with 13CO2 that both HCOOH and CO originated from CO2 introduced into the reaction system (see Figure S7).

Figure 2

a) A typical time course of CO2 reduction using α‐FeOOH/Al2O3 under visible light (λ>400 nm). Reaction conditions: catalyst, 4 mg (Fe loaded 10.0 wt % to catalyst); solution, 4 mL DMA containing 1.0 mM Ru and 0.1 M BNAH. b) Amounts of reaction products and the HCOOH selectivity in photocatalytic CO2 reduction using α‐FeOOH/Al2O3 under visible light (λ>400 nm). Each run was conducted for 3 h, and the α‐FeOOH/Al2O3 catalyst was subsequently recovered by centrifugation. The next reaction was then started using a new reaction solution and the recovered catalyst.

The CO2 reduction performance was also dependent on the loading amount of α‐FeOOH on Al2O3. As shown in Figure S8, the performance improved as the α‐FeOOH loading amount was increased to 10.0 wt %; at greater loading amounts, the performance was adversely affected. A commercially available bulk α‐FeOOH achieved CO2 reduction to HCOOH at a rate comparable to that of the α‐FeOOH/Al2O3 catalyst; however, the selectivity to HCOOH was lower (entry 8). These results imply that highly dispersed α‐FeOOH is important for selective CO2 reduction. A well‐known iron(III) oxide, α‐Fe2O3, also functioned as a catalyst but exhibited lower CO2 reduction performance (entry 9).

As mentioned earlier, it is most likely that the present CO2 photoreduction system with the [Ru(bpy)3]2+ photosensitizer works according to a reductive quenching mechanism, in which one‐electron‐reduced species of the Ru complex donates an electron to solid catalysts and the concomitant CO2 reduction occurs. It is also noted that the α‐FeOOH/Al2O3 catalyst produced H2 in the absence of CO2 (Table 1, entry 5). A similar result has been reported when a molecular catalyst was employed instead of the solid catalyst under the condition identical to the present. Therefore, reduction of CO2 and proton can compete with each other, depending on the environment of the catalyst surface as well as CO2 adsorption capability of the catalyst.

CO2 adsorption capabilities of the catalyst samples were examined. A shown in Figure 3, the bulk α‐FeOOH had the ability to adsorb CO2, consistent with early reports,[ , ] which was superior to α‐Fe2O3. More importantly, the α‐FeOOH/Al2O3 catalysts captured CO2 more efficiently than the bulk α‐FeOOH. It is also noted that catalysts having higher CO2 adsorption capability tended to give higher CO2 reduction activity and selectivity. However, this was not the case for the 30.0 wt % sample, which had a little higher CO2 adsorption capability, but lower CO2 reduction activity than the optimal 10.0 wt % sample (Figure S8). In the 30.0 wt % sample, aggregation of the loaded Fe species was observed (Figure S9), and the lower activity is in line with a general trend in heterogeneous catalysis; i.e., aggregated species give lower catalytic activity. Nevertheless, the 30.0 wt % sample still maintained its high CO2 reduction selectivity. Considering these results, the primary role of the Al2O3 support is concluded to provide α‐FeOOH with suitable dispersion, thereby improving the adsorption of CO2, which is essential to increasing the CO2 reduction activity and selectivity.

Figure 3

CO2 sorption isotherms of catalyst samples at 298 K.

As a homogeneous photocatalytic system for selective visible‐light CO2 reduction with the use of [Ru(bpy)3]2+ and BNAH as a photosensitizer and an electron donor, a molecular catalyst of [Ru(bpy)2(CO)2](PF6)2 has been reported by Ishida et al. According to that report, the homogeneous system produced nearly 1 : 1 HCOOH and CO from a DMA/BNAH mixed solution, with the total quantum yield for CO2 reduction being estimated to be ≈8.9 % at 460 nm. Therefore, the AQY of our system (4.3 %) was roughly half that of the representative homogeneous system. Considering the nature of heterogeneous catalysis where the reaction rate is generally slower compared to a homogeneous system, nevertheless, the present result would be a good starting point of further research on Fe‐based solid catalyst for efficient CO2 reduction.

In conclusion, we demonstrated that α‐FeOOH/Al2O3 functions as a recyclable catalyst for CO2 reduction to HCOOH with ≈90 % selectivity in the presence of [Ru(bpy)3]2+ and BNAH as a photosensitizer and an electron donor, respectively, under visible light. This catalyst is the first example of an Fe‐based solid catalyst for HCOOH generation that can function in a photochemical CO2 reduction scheme with the aid of a proper redox photosensitizer. More importantly, the results of this work suggest that, with the use of a proper support material (here, Al2O3), well‐known, earth‐abundant compounds can be used as selective catalysts for CO2 reduction without a complicated catalyst preparation method. Such a support effect has not been explored sufficiently in the search for solid catalysts applicable to photochemical CO2 reduction schemes. Therefore, much room exists for developing precious‐metal‐free, abundant compounds—which are, of course, not limited to α‐FeOOH—for reducing CO2 to energy‐rich chemicals, although mechanistic studies will be important.

Conflict of interest

The authors declare no competing financial interest.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supporting Information

Click here for additional data file.