Two-dimensional (2D) nanosheets consisting of a few atomic layers have attracted significant attention for potential application in various fields due to their unique properties. Certain metal oxide nanosheets can be readily prepared by exfoliation of the corresponding layered metal oxides with organic bases such as tetra(n-butyl)ammonium hydroxide, TBA+OH−. Exfoliated nanosheets are amenable to restacking by reactions with acid or base. Compared to the bulky host layered solid, the resulting material possesses a much larger surface area, which could potentially be useful for heterogeneous photocatalysis.[3-10]
Water oxidation involving a four-electron process is a particularly important step in artificial photosynthesis for solar fuel production, not only for water splitting, but also for CO2 fixation. It is known that certain metal oxide semiconductors exhibit photocatalytic activity for the reaction under band-gap irradiation. In bulk type metal oxide photocatalysts such as TiO2 and SrTiO3, high crystallinity is a prerequisite for efficient water oxidation. While higher crystallinity leads to lower probability of undesirable electron–hole recombination,[12a] there is a smaller surface area that limits available sites for redox reactions. On the other hand, applying transition metal oxide nanosheets as components of the water oxidation reaction is of interest, since the exfoliated materials have both high crystallinity and large surface area. So far, photocatalytic water reduction to form H2 from water using oxide nanosheets has been reported by several researchers.[3-9] However, it should be stressed that non-sacrificial water oxidation using nanosheets in the presence of a reversible electron acceptor has been little explored, although there are a few reports that describe photocatalytic water oxidation based on nanosheets in the presence of an irreversible electron acceptor (e.g., Ag+ or Ce4+).[8b],[9, 10] To achieve an artificial photosynthetic reaction where the Gibbs energy change is positive, a water oxidation photosystem that works in the presence of a reversible electron acceptor is indispensable.[11b]
In this work, we investigated non-sacrificial water oxidation using KCa2Nb3O10-based solid photocatalysts. KCa2Nb3O10 consists of negatively charged calcium niobate sheets that stack along the c axis to form a 2D layered structure, in which K+ cations are located between the triple perovskite layers to compensate for the charge balance. Photocatalytic activities for water oxidation were tested under band-gap irradiation of KCa2Nb3O10 in the presence of IO3− ions that work as reversible electron acceptors. The half reactions of the present water oxidation reaction are:
The total reaction is energetically uphill:
Details of the preparation of layered KCa2Nb3O10 and the nanosheets are included in the Supporting Information. Figure 1A shows the XRD patterns of the restacked KCa2Nb3O10 nanosheet samples, along with a reference pattern of the parent lamellar solid. The XRD pattern of the restacked sample gives a very weak (002) diffraction peak and a complete absence of (00l) (l ≥ 6) peaks, indicating that the periodic layered structure of the parent solid almost disappears upon exfoliation with TBA+OH− and the subsequent restacking process. This is apparent in the SEM observations (Figure 1B), which indicates that the original plate-like layered structure is completely destroyed, producing aggregated solids with a disordered structure. However, (100) and (110) diffraction peaks corresponding to in-plane diffraction are preserved in the XRD patterns. This indicates that the two-dimensional structure of perovskite sheets is preserved after the exfoliation–restacking procedure. The positions of the (002) and (004) diffraction peaks in the restacked nanosheets appear at lower 2θ angles than those in the corresponding layered material, suggesting hydration of the nanosheet material. The specific surface area of the restacked material was determined to be ca. 26 m2 g−1, much larger than that of layered KCa2Nb3O10 (∼1 m2 g−1).
The UV-visible diffuse reflectance spectrum of the restacked KCa2Nb3O10 nanosheets differs from that of the parent layered solid (Figure 1C). Layered KCa2Nb3O10 has a steep absorption edge at around 350 nm, which is due to electron transitions from the valence band formed by oxygen 2p orbitals to the conduction band that consists of empty orbitals of niobium 4d. The band gap of layered KCa2Nb3O10 is estimated at ca. 3.5 eV, based on the onset wavelength of the diffuse reflectance spectrum. On the other hand, there are at least two absorption edges (ca. 320 and 340 nm) in the restacked material, both of which are blue-shifted compared to that in the layered material. This is most likely due to a quantum-confinement effect. The generation of two absorption edges after the exfoliation–restacking process may originate from an increase in the distortion of the 2D nanosheet structure that consists of triple perovskite slabs of ∼1 nm thickness, changing the local structure of NbO6 octahedrons in the perovskite block, although the long-range ordering in the perovskite block (shown in XRD) appears to be maintained.
Using the as-prepared KCa2Nb3O10 materials, photocatalytic water oxidation was conducted in the presence of IO3− ions as reversible electron acceptors. As shown in Figure 2, layered KCa2Nb3O10 without modification exhibited no activity. With exfoliation of layered KCa2Nb3O10 and subsequent restacking, however, O2 evolution was clearly observable. The amount of evolved O2 increased over time, indicating that water oxidation and reduction of IO3− both took place on the surface of the calcium niobate nanosheets. The amount of I− produced as the result of the reduction of IO3− was quantified by means of ion chromatography. This showed that the amount of I− found in the liquid phase corresponds to two thirds of the O2 production, consistent with the amount expected from the stoichiometry (see Equation (3)). No noticeable change in the reaction pH occurred before and after reaction (pH = 8–9). It was also confirmed that no gas evolution took place in the dark or under > 380 nm irradiation. Without NaIO3, no reaction occurred as well.
Another important finding is that loading RuO2 as an additional cocatalyst enhanced the water oxidation rate on restacked KCa2Nb3O10 nanosheets. As shown in Figure S1, the reaction rate increased abruptly with Ru content to a maximum at around 0.1 wt% and then decreased gradually upon further loading. The rate of O2 evolution recorded using the optimal catalyst was an order of magnitude higher than that achieved with the unmodified sample. Scanning electron microscopy (SEM) observation showed that the size of the RuO2 deposits was typically 10–40 nm, with some aggregation (Figure S2). With an increase in the loading amount of RuO2, more coverage of the surface of restacked KCa2Nb3O10 nanosheets was observed. Unfortunately, it was difficult to find RuO2 deposits in the samples with lower RuO2 content, because RuO2 deposits and the surface roughening of the restacked nanosheets were not distinguishable. It is also noted here that no intercalation of RuO2 could be identified (see Figure S3, XRD patterns before and after RuO2 loading), indicating that RuO2 cocatalysts were deposited almost exclusively on the external surface of restacked KCa2Nb3O10. In the 1.0 wt% sample, a small diffraction peak assigned to RuO2 was observed. This means that the loaded Ru species are indeed RuO2, and excess loading results in the generation of bulky RuO2 deposits. We have previously elucidated that RuO2 loaded on TaON acts as a bifunctional cocatalyst to oxidize water into O2 and to reduce IO3− into I−. Therefore, the functionality of RuO2 anchored on the surface of KCa2Nb3O10 nanosheets would be identical.
On the basis of these results, nanoparticulate RuO2 with an optimal distribution is concluded to be essential to efficiently promote the water oxidation reaction. Indeed, RuO2-loading was also effective for enhancing water oxidation by layered KCa2Nb3O10 (Table 1, entry 4). However, the promotional effect was much more pronounced for restacked KCa2Nb3O10 than for the layered one (Table 1, entries 1–4), indicating that the exfoliation–restacking procedure, which peels off the lamellar structure of KCa2Nb3O10, is essential to inducing the water oxidation activity.
|Entry||KCa2Nb3O10 form||Loading amount of RuO2/wt%||O2 evolution rateb)/μmol h−1|
As indicated in Figure 1B, the exfoliation–restacking process resulted in aggregation of individual sheets, which causes bending of the perovskite sheets to different extents, although the in-plane crystallinity (in other words, long-range atomic ordering) is preserved (Figure 1A). This reflects the change in the optical absorption profile of KCa2Nb3O10 before and after exfoliation–restacking (Figure 1C). Because water oxidation to form O2 molecules on the surface of a metal oxide single-crystal is sensitive to the crystal face, such a change in surface local structure after the exfoliation–restacking process (including the modulation of the atomic arrangement and/or the formation of suitable reaction sites) might contribute to inducing the water oxidation activity of KCa2Nb3O10. The enlarged band gap of restacked KCa2Nb3O10 nanosheets that can lead to an increase in the driving forces for surface redox reactions is another possible reason for the activation of KCa2Nb3O10 for water oxidation.
If the water oxidation rate is governed simply by the difference in specific surface area, just increasing the amount of layered KCa2Nb3O10 used for the reaction would enhance the reaction rate. However, this was not the case. As listed in Table 1, the water oxidation rate recorded using 500 mg of layered KCa2Nb3O10 (entry 5) was much lower than that achieved by 20 mg of the restacked one (entry 6), even though the total surface areas of both catalysts were similar (∼0.5 m2). This result further confirms that the exfoliation–restacking process is essential to inducing non-sacrificial water photo-oxidation activity of KCa2Nb3O10.
The water oxidation behavior of RuO2-modified restacked KCa2Nb3O10 nanosheets was found to be stable. After recovering the used photocatalyst powder by centrifugation and re-dispersing a fresh NaIO3 solution, almost the same rate of O2 evolution was observed (Figure S4), although the activities recorded in the second and third run were slightly lower than that in the first run. In addition, no noticeable change could be identified in the XRD pattern of RuO2/restacked KCa2Nb3O10 before and after the water oxidation reaction (Figure S5). Based on these results, it is concluded that the photocatalyst itself is stable for water oxidation. The apparent quantum yield for the water oxidation by 0.1 wt% RuO2/restacked KCa2Nb3O10 was measured to be ca. 3.6 ± 0.1% at 300 nm.
In summary, we applied restacked 2D-nanosheets of KCa2Nb3O10 as a photocatalyst to oxidize water into molecular O2 in the presence of IO3− ions as reversible electron acceptors under band-gap irradiation. Our experimental results led us to conclude that the activation of lamellar KCa2Nb3O10 by exfoliation–restacking for non-sacrificial water oxidation results from the structural modulation of the 2D perovskite sheets. To the best of our knowledge, such nanostructure-sensitive behavior in water oxidation has not been observed for oxide nanosheet photocatalysts, let alone bulk-type metal oxides. In the case of nanostructured photocatalysts such as oxide nanosheets for water oxidation, the flexible 2D-sheet structure that possesses single-crystalline texture appears to be important and differentiates the nanosheet-based systems from conventional bulk-type metal oxide systems, reflecting the unique photocatalytic property of nanosheet photocatalysts. We believe this new finding will facilitate further research on nanostructured photocatalysts to achieve optimal functions for desired reactions, as there are a wide variety of oxide nanosheets with controllable compositions and structures.