Photocatalytic water splitting on hydrated layered perovskite tantalate A2SrTa2O7·nH2O (A = H, K, and Rb)

A series of layered perovskite tantalates, A2SrTa2O7 (A = H, Li, K, and Rb), were prepared as novel photocatalysts for photocatalytic water splitting into H2 and O2 under UV irradiation.

The layered perovskite tantalates with hydrated interlayer space, A2SrTa2O7·nH2O (A = H, K, and Rb), showed higher H2 formation rate than anhydrous layered tantalate, Li2SrTa2O7, and anhydrous perovskite tantalate, KTaO3.

H2SrTa2O7·nH2O and K2SrTa2O7·nH2O showed high activity for overall splitting of water without loading co-catalysts.

The reaction over H2SrTa2O7·nH2O proceeded steadily more than 70 h, demonstrating a high durability of the catalyst.

Effects of hydrated interlayer space on the catalytic activity were discussed on the basis of the results of photoluminescence spectra and the hydrogen evolution from aqueous solution of n-butylamine as a test reaction.

The results indicate that the availability of interlayer space of layered tantalate as reaction sites is an important factor to improve the photocatalytic activity of Ta-based semiconductor materials.


Among various methods of solar energy conversion, the photocatalytic decomposition of water into H2 and O2 over semiconductor materials is one of the most challenging methods.

Much attention has been paid to this reaction, as it can potentially provide clean and renewable source for hydrogen fuel in the future.

Since TiO2 electrode was first studied for water decomposition under UV-light irradiation, a variety of photocatalysts, mainly Ti,1–4 Nb5,6 and Ta7–13 based oxides, have been reported to be effective for photocatalytic decomposition of water.

A group of Domen et al5,6. first demonstrated the use of ion-exchangeable layered oxides, K4Nb6O17 and K2La2Ti3O10, as effective photocatalysts for water splitting.

Upon loading transition metal co-catalyst, NiOx, these materials show much higher activity than the “bulk type” catalysts such as Pt/TiO2.

It was explained that these layered materials use their interlayer space as reaction sites, where electron-hole recombination process could be retarded by physical separation of electron and hole pairs generated by photo-absorption.

On the other hand, Ta-based catalysts have recently been developed as a new class of photocatalyst.

Kudo et al. first studied the photocatalytic properties of many Ta-based catalysts, such as Sr2Ta2O79 and NaTaO3.10

Among these catalysts, La doped NiO/NaTaO3 showed extremely high efficiency with quantum yields as high as 50% under UV irradiation.11

Recently, Zou et al. demonstrated the direct water splitting under visible light irradiation with In–Ni–Ta oxides.12

However, for most of the photocatalysts, it is necessary to load transition metal co-catalysts to achieve high activity for water splitting.

The above Ta-based catalysts can be classified to anhydrous bulk type photocatalysts.

We expected that Ta-based materials with hydrous layered structure can be an effective photocatalysts for water splitting.

Machida et al. first studied the photocatalytic properties of layered perovskite tantalates.13

They prepared layered tantalates, ANdTa2O7 (A = H or Na) in hydrous form, whose interlayer space is potentially available for the reaction.

However, the activity of these layered tantalates was lower than that of NaNdTa2O7 in dehydrated form, indicating that hydration of this type of layered tantalate is not effective in improving the photocatalytic activity.

A2SrTa2O7 (A = H, Li, K and Rb) type tantalate are new members of Ruddlesden–Popper type layered perovskite.14–17

The structure of K2SrTa2O7 and its hydrate has been recently clarified.

These materials are composed of layers of Sr–Ta perovskite sheets and potassium ions located in the interlayer.

The layered structure allows spontaneous water intercalation when anhydrous K2SrTa2O7 is kept in air at ambient temperature.

We hypothesized that from these features, A2SrTa2O7 materials act as new Ta-based photocatalysts with two-dimensional structures in water.

In this paper, we report the first successful example of layered perovskite-type tantalates with hydrated interlayer space, A2SrTa2O7·nH2O (A = H, K and Rb), as very active photocatalysts.

The activity of these novel catalysts is compared to that of Li2SrTa2O7 (anhydrous layered perovskite-type tantalate) as well as that of the bulk type perovskite tantalate.

Photocatalysts are characterized by XRD, diffuse reflectance UV spectroscopy, photoluminescence spectroscopy, and the hydrogen evolution from aqueous solution of n-butylamine as a test reaction, and the relationship between the catalytic activity and the structural characteristics is discussed in terms of the effect of the interlayer.


Catalyst preparation and characterization

A2SrTa2O7 (A = Li, K, and Rb) powders were prepared by a conventional solid state reaction.

Stoichiometric amounts of SrCO3 (Kanto Chemical, purity; 99.9%) and Ta2O5 (Rare Metallic, purity; 99.9%) with a 150% molar excess of K2CO3 (Kanto Chemical, purity; 99.5%) or Rb2CO3 (Kanto Chemical, purity; 99.5%) were mixed together and heated in air at 1173 K for 24 h.

Li2SrTa2O7 was prepared from SrCO3 and Ta2O5 with a 150% molar excess of Li2CO3 (Kanto Chemical, purity; 99.5%) by heating at 1523 K for 12 h.17

After the calcinations, the excess alkali was washed out with deionized water, and the sample was dried in vacuo at 298 K for 20 h.

According to the previous report,16 H2SrTa2O7 was prepared by the K+/H+ exchange: K2SrTa2O7 (3 g) was added to 150 mL of 0.3 M HNO3 at room temperature for 24 h with constant stirring, followed by washing with deionized water, and by drying in vacuo at 298 K for 20 h.

After the preparation, A2SrTa2O7 (A = H, K, and Rb) materials exhibit hydrated structure and are named as A2SrTa2O7·nH2O (A = H, K, and Rb).

The crystal structure of the obtained materials was confirmed by X-ray diffraction (MAC Science; MX Labo) with Cu Kα radiation (40 kV, 25 mA) at room temperature.

Thermogravimetric (TG) analysis coupled with differential thermal analysis of the hydrated oxides was made using MTC1000 (MAC Science) operating with a heating rate of 10 K min−1.

Diffuse reflectance spectra were obtained with a UV-Vis spectrometer (Jasco; V-550) and were converted from reflection to absorbance by the Kubelka–Munk method.

The optical band-gap energy was calculated from onset of absorption edges.

Photoluminescence was measured at 77 K using a closed quartz cell and a fluorometer (HITACHI F-4500).

For K2SrTa2O7 in anhydrous form, XRD, UV-Vis and photoluminescence measurements were carried out immediately after calcination in order to avoid the hydration process.

Photocatalytic reaction

The photocatalytic decomposition of water was performed with a gas-closed circulating system.

The catalyst powder (0.5 g) was dispersed in 200 mL of pure water by a magnetic stirrer in an inner irradiation cell made of quartz.

The light source was a 400 W high-pressure mercury lamp.

Before the reaction, the mixture was degassed completely and then Ar (ca. 15 kPa) was introduced.

The amounts of evolved H2 and O2 were determined by gas chromatography (Hitachi, TCD, molecular sieve 5A column and Ar carrier) sampler, 3 mL of which was directly connected to the closed gas circulation system to avoid any contamination from air.

Gas evolution was observed only under photoirradiation.

Results and discussions

Characterization of A2SrTa2O7

The powder X-ray diffraction patterns of the samples are shown in Fig. 1.

In the XRD pattern of K2SrTa2O7 recorded immediately after calcination (Fig. 1a), all the observed lines could be indexed on a tetragonal cell (I4/mmm, a = 3.9072 Å, c = 21.6006 Å),15,16 indicating that this compound consists of a single phase of layered perovskite, K2SrTa2O7.

Fig. 2a represents the schematic crystal structure of K2SrTa2O7 before hydration.

The structure of anhydrous K2SrTa2O7 can be described as formed from two TaO6 octahedra thick slabs of a perovskite lattice cut along the c direction; these alternate layers are shifted by (a + b)/2 (body centered), with the large Sr cations fully occupying the 12-coordination sites.

The potassium ions are located in a rock salt type coordination (C.N. = 9).16

Recently, Crosnier–Lopez et al16. reported that anhydrous form of K2SrTa2O7 completely disappeared after exposing it to humid air for 7 h, and the corresponding hydrated phase appeared simultaneously.

In our case, it was also confirmed that K2SrTa2O7 allowed intercalation of water when exposed to humid air; the anhydrous form of K2SrTa2O7 completely disappeared after exposing it to humid air for 24 h or by washing it with deionized water, and the corresponding hydrated phase, K2SrTa2O7·nH2O,15 appeared simultaneously.

In the XRD pattern of K2SrTa2O7·nH2O (Fig. 1d), prepared by washing K2SrTa2O7 with deionized water and by drying at 298 K, all the observed lines could be indexed on a primitive cell (P4/mmm, a = 3.9974 Å, c = 12.133 Å).

Fig. 2b represents the crystal structure of K2SrTa2O7·nH2O. This kind of structural change has already been observed in layered perovskites, and has been attributed to a shift of alternate layers.18

The adjacent layers are stacked immediately above each other in the same arrangement leading to halving the length of the c-axis upon hydration.

The dehydration behaviour of K2SrTa2O7·nH2O was studied by TG analysis.

The result showed a distinct loss of water at a temperature range of 373 K < T < 743 K, which could be due to the transformation of hydrate to anhydrous layered material K2SrTa2O7.

The number of hydration in the formula of K2SrTa2O7·nH2O estimated by the analysis was n = 0.9.

In the XRD patterns of H2SrTa2O7·nH2O (Fig. 1c), the observed lines could be indexed on a primitive cell (P4/mmm, a = 3.9038 Å, c = 9.7742 Å).

The structure of Rb2SrTa2O7·nH2O has not been reported.

The Rietveld method with the RIETAN2000 profile refinement program21 was used to refine the structure in the space group P4/mmm (no.

139) with a starting model for the perovskite slabs similar to that observed in K2SrTa2O7·nH2O. It was found that the Rb2SrTa2O7·nH2O material consists of a single-phase layered perovskite in a hydrated form, whose X-ray diffraction pattern were indexed based on a primitive cell (P4/mmm, a = 3.9727 Å, c = 12.7632 Å).

From the thermogravimetric analysis, the hydration numbers of H2SrTa2O7·nH2O and Rb2SrTa2O7·nH2O were estimated to be n = 0.6 and n = 3.6, respectively.

In summary, the above three materials, A2SrTa2O7·nH2O (A = H, K, and Rb), consist of the layered perovskite in a hydrated form, as illustrated in Fig. 2b, though the hydration number depends on the type of alkali cation.

Among these materials, theb c-axis length increased in the order of H < K < Rb, which should be due to the increase in the ionic radius of the alkali metal cations and/or increase in the hydration number at the interlayer space.

In the XRD pattern of Li2SrTa2O7 (Fig. 1b), recorded after exposing it to humid air for 24 h, all the observed lines could be indexed on a tetragonal cell (I4/mmm, a = 3.8470 Å, c = 18.1094 Å), indicating that this compound consists of a single phase of layered perovskite, Li2SrTa2O7.17

The XRD pattern of Li2SrTa2O7 was essentially the same after the dispersion of the sample to the distilled water for 6 h, followed by drying at ambient temperature.

This indicates that water intercalation does not occur on this material.

UV diffuse reflectance spectra of the catalysts are shown in Fig. 3.

All the samples had an absorption band in the ultraviolet region and showed a clear absorption edge at around 320 ∼ 330 nm.

From the absorption edge, the band gap of H2SrTa2O7·nH2O, Li2SrTa2O7, K2SrTa2O7·nH2O, and Rb2SrTa2O7·nH2O was about 3.9 eV, irrespective of the alkali metal cations at the interlayer space.

The band gap of anhydrous K2SrTa2O7 was about 3.8 eV.

As listed in Table 1, the BET surface areas of the samples were very small (0.2–4.0 m2 g−1).

Fig. 4 shows emission and excitation spectra of the layered tantalates.

In the emission spectra with an excitation wavelength of 300 nm, the layered tantalates in anhydrous form, Li2SrTa2O7 and K2SrTa2O7, showed intense luminescence peak with a maximum around 490 and 470 nm, respectively.

Although luminescence properties of these compounds were not reported, the positions of excitation and emission peaks are very close to those of other anhydrous perovskite tantalates, such as KTaO3 and Sr2Ta2O7.22–24

The excitation and emission transitions are known to be due to charge transfer transitions within the octahedral tantalate groups.22

The layered tantalates in hydrous form, H2SrTa2O7·nH2O and K2SrTa2O7·nH2O, showed a very weak and broad band with a maximum at around 500 nm.

These results can be interpreted as the photoluminescence intensity of the layered tantalates being drastically decreased when the interlayer water is present.

The excitation spectra of Li2SrTa2O7 and K2SrTa2O7 have onsets at 320 and 330 nm, respectively, which correspond to the diffuse reflectance spectrum.

This indicates that band gap irradiation is necessary for the appearance of the luminescence spectra, and hence photogenerated electrons and holes play a role in the emissions.

It is reasonable to assume that in the hydrated layered tantalates, H2SrTa2O7·nH2O and K2SrTa2O7·nH2O, photogenerated electrons and holes can be readily transferred to the interlayer water.

Hence, we consider that significantly low luminescence efficiency of the hydrated layered tantalates is mainly due to the electron and/or hole trapping by the interlayer water rather than the non-radiative transition.

Photocatalytic activity of A2SrTa2O7

Fig. 5 shows a time course of gas evolution for the photocatalytic decomposition of water over H2SrTa2O7·nH2O without any catalyst pretreatment.

The catalyst produces H2 and O2 in a stoichiometric ratio (H2/O2 = 2 ∶ 1).

After the first run, the system was evacuated, and the reaction was repeated using the same reaction mixture.

In the second and third runs, overall splitting of water occurred with almost the same rate as the first run, which demonstrates a high durability of the catalyst.

The total amount of H2 and O2 evolved during these runs over H2SrTa2O7·nH2O reached 21.2 and 11.2 mol (mol cat.)−1, respectively.

These results clearly demonstrate that overall water splitting on this material proceeds catalytically.

As shown in Fig. 6, the K2SrTa2O7·nH2O material also produced stoichiometric mixtures of H2 and O2, and the reaction proceeded at a steady rate.

Durability test of K2SrTa2O7·nH2O was also performed for 65 h, and the total amount of H2 and O2 evolved during these runs reached 23.2 and 12.1 mol (mol cat.)−1, respectively (result not shown).

The Rb2SrTa2O7·nH2O material produced H2 and O2 simultaneously, though the rate of O2 evolution was less than the stoichiometric amount.

Li2SrTa2O7 showed very low activity for H2 formation as compared to the other layered tantalates, and O2 evolution was not observed with this material.

Table 1 summarizes the rate of gas evolution over various Ta-based catalysts.

Using H2SrTa2O7·nH2O as a photocatalyst, the rates of H2 and O2 formation were 385 and 179 μmol h−1, respectively.

The rates of H2 and O2 formation with K2SrTa2O7·nH2O were very close to those of H2SrTa2O7·nH2O. It should be noted that the present system is the first example of active Ta-based photocatalysts with hydrous layered structure.

The rate of H2 evolution over A2SrTa2O7nH2O) was strongly dependent on alkali cation and decreased in the sequence of H ≈ K > Rb ≫ Li.

We also studied the reaction over the non-hydrous perovskite tantalate, KTaO3, which was reported to catalyse this reaction.10

This material showed moderate activity for H2 evolution, though the rate of H2 evolution was lower than those of A2SrTa2O7·nH2O (A = H, K, and Rb).

Together with the result that the anhydrous layered tantalate, Li2SrTa2O7, was less effective than hydrous layered tantalate, A2SrTa2O7·nH2O (A = H, K, and Rb), one can conclude that the presence of hydrated interlayer space is important for improving the photocatalytic activity of Ta-based catalysts.

In the photoluminescence study shown in Fig. 4, it was shown that the hydrated layered tantalates, A2SrTa2O7·nH2O (A = H and K), exhibited a considerably lower luminescence efficiency than anhydrous layered tantalates, A2SrTa2O7 (A = Li and K), probably due to the electron and/or hole trapping by the interlayer water.

From this result, the positive effect of hydrated interlayer on the activity could be explained as follows.

The photogenerated electrons and holes can be effectively transferred to the interlayer water, and this results in the effective photocatalytic water splitting in the hydrated layered tantalates.

Effect of sacrificial agents

In overall water splitting, oxidation of water by holes is a slower process than reduction by electrons.

In order to facilitate the oxidation, hole-scavengers are often introduced.

In this study, n-butylamine was used as a sacrificial agent.

Fig. 7 shows the time course of H2 evolution from aqueous solution of n-butylamine (0.25 M) using 0.1 g of the catalyst.

The initial rates of H2 evolution are listed in Table 2.

For H2SrTa2O7·nH2O the rate of H2 formation (1490 μmol h−1) was much higher than that in the absence of n-butylamine (385 μmol h−1), though the amount of the catalyst used was lower in the former reaction.

The results in Table 2 show that the activity sequence for the n-butylamine/H2O system (H2SrTa2O7·nH2O ≫ K2SrTa2O7·nH2O > KTaO3) is markedly different from that observed in the H2 evolution from pure water (H2SrTa2O7·nH2O ≈ K2SrTa2O7·nH2O ≫ KTaO3).

It is well known that H+-exchanged layered oxides, such as HCa2Nb3O10, can intercalate organic bases such as amines into the interlayer space,19,20 while K+-exchanged one can not intercalate these molecules.

Table 2 includes the amount of n-butylamine adsorbed on the samples when the sample was stirred in the dark for 24 h under the same condition as the catalytic reaction shown in Fig. 7.

As expected, 0.94 mmol g−1 of n-butylamine was adsorbed in H2SrTa2O7·nH2O probably due to the interlayer acidity of this material.

In contrast, only negligible amount of n-butylamine adsorption was observed on K2SrTa2O7·nH2O and KTaO3.

This result can explain the activity sequence for the n-butylamine/H2O system.

Although the band structures of H2SrTa2O7·nH2O and K2SrTa2O7·nH2O are very close to each other (Fig. 3), one of the reactant molecules, n-butylamine, is hardly intercalated into the interlayer space of K2SrTa2O7·nH2O during the reaction, whereas n-butylamine is intercalated into the interlayer space of H2SrTa2O7·nH2O to act as the hole-scavenger over the reaction site at the TaO6 perovskite sheet.

As for the decomposition of pure water, these two catalysts show similar reactivity, because the water intercalation occurs during the reaction in both catalysts.

These results support the above proposal that the photocatalytic reaction proceeds at interlayers of the hydrated layered tantalates.

In the case of hydrous tantalates, electron–hole pairs may be easily migrated to the interlayer surface consisting of TaO6 before their recombination, and thus intercalated H2O molecules can be effectively decomposed.


We have developed novel photocatalysts A2SrTa2O7·nH2O (A = H, K and Rb) with hydrated layered perovskite structure.

These catalysts showed higher activity than Li2SrTa2O7 (anhydrous tantalate with similar layered structure) and the bulk type perovskite tantalate, KTaO3.

Overall water splitting was accomplished over A2SrTa2O7·nH2O (A = H and K) without any modification.

This is the first successful example of highly active tantalates photocatalysts with hydrated interlayer space.

The results from photoluminescence spectroscopy and H2 evolution from aqueous n-butylamine solution support the hypothesis that the high activity of A2SrTa2O7·nH2O results from their hydrated layered structure where the photogenerated electrons and holes can be effectively transferred to the interlayer water.