Tantalates have attracted considerable interest owing to their attractive roles in optic, optoelectronic, and electronic applications ^[1]. Recently, tantalates have been extensively studied in view of their high photocatalytic activities for water splitting into H₂ and the degradation of toxic substances ^{[2, 3, 4, 5, 6, 7, 8, 9]}. Particularly, the quantum yield of NiO/NaTaO₃:La was estimated to be 56% at 270 nm, which is the highest quantum yield ever reported for catalysts in pure water splitting ^[9]. However, the tantalates are usually synthesized by solid state methods at high temperatures of approximately 1000 °C and require long synthesis times ^{[9, 10, 11]}. Such requirements may result in large grain growth (thus reducing the surface areas), which leads to a decrease in the photocatalytic activity ^[12]. Another possible approach for the synthesis of tantalates is the use of simple solution-based methods, in which soluble tantalum (Ta) precursors are used as starting materials. An alkoxide of Ta has been adopted to prepare LiTaO₃; however, this precursor is expensive and its solution is extremely sensitive to moisture ^{[13, 14]}. A polymerized complex method from TaCl₅ has been employed to prepare the A₂Nb_xTa_2-xO₇ solid solutions (A = Ba, Sr) ^[15]. However, this approach is complicated and usually involves poisonous organic solvents. Others have used similar solution-based processes to synthesize tantalates by converting Ta₂O₅ to a Ta-based solution, although Ta₂O₅ is hardly dissolved by regular chemical routes. For example, HF acid (40%), which is strongly corrosive and toxic, has been used to dissolve Ta₂O₅, and soluble tantalic acids (Ta₂O₅·nH₂O) have been prepared starting either from Ta₂O₅ or from metallic Ta by the KHSO₄-oxalic acid method ^[16]. Among these solution-based preparations, a high temperature is also usually necessary to obtain the final product. Therefore, it is highly desirable to develop a low-temperature route to synthesize tantalates from cheap, nontoxic starting materials for their further application. Nevertheless, so far, only a few types of tantalates, such as ATaO₃ (A = Li, Na, K) ^[2], Sr₂Ta₂O₇ ^[3], and Ba₅Ta₄O₁₅ ^[6], have been synthesized from Ta₂O₅ by a hydrothermal method under high concentrations of OH^-, and the obtained products often exhibit an irregular-shaped morphology with a broad size distribution. Despite some success in the preparation of tantalates, it remains a major challenge to synthesize high-quality powder tantalates ^[17].

Morphology control of tantalates is another potentially important issue to meet different scientific and technological needs because of the strong correlation between the morphology and the physical or chemical properties. However, to the best our knowledge, it is still difficult to achieve morphology-controlled synthesis of tantalates. It is even more of a challenge to achieve morphology-controlled synthesis of tantalates from Ta₂O₅ at a relatively low temperature. Herein, using Ta₂O₅ as a starting material, we report on the facile synthesis of K_1.9Na_0.1Ta₂O₆·2H₂O (KNTO) crystals with controllable morphologies by a two-step synthesis technique, namely, the molten salt in combination with the hydrothermal method. Importantly, the morphology control of KNTO is achieved by simply adjusting the concentrations of KOH addition in the hydrothermal reaction. Taking an example, KNTO nanospheres were found to exhibit a distinguished photocatalytic activity for degrading rhodamine B (RhB) under ultraviolet (UV)-light irradiation.

K_1.9Na_0.1Ta₂O₆·2H₂O (KNTO) was synthesized by a two-step technique as published previously ^{[4, 5]}, namely, the molten salt and hydrothermal methods. Ta₂O₅ was converted to a soluble Ta precursor at 240 °C in the KOH/NaOH melt. The aqueous solution of the obtained Ta precursor is then transferred as a starting source of Ta to synthesize KNTO by the hydrothermal process at 180 °C. Table 1 lists the experimental details for the KNTO crystals with different morphologies, which are defined as samples 1-4. In a typical synthesis (sample 1 in Table 1), a mixture of Ta₂O₅ (0.5 g), NaOH (3.855 g), and KOH (5.651 g) was fused at 240 °C in a Teflon cup for 20 h and air-cooled to room temperature. Deionized water (50 mL) was then added into the cup to dissolve the alkalis. The solid collected after centrifugation was dissolved in deionized water (100 mL) with magnetic stirring. Then, 1.2 mL of 5 mol/L KOH aqueous solution (V_KOH) was added dropwise into a Teflon cup (50 mL) containing 33.8 mL (V_Ta) of the above clear solution under constant stirring at room temperature. With constant magnetic stirring for 8 h, the Teflon stainless autoclave was then maintained at 180 °C for 3 h and then air-cooled to room temperature. The samples obtained were collected by centrifugation, washed several times with deionized water, and air dried at 60 °C. Other products were prepared under the same conditions, except for the values of V_Ta and V_KOH (V_Ta + V_KOH = 35 mL).

Table 1 Composition and surface areas of the KNTO samples. Sample VTa (mL) VKOH (mL) K:Na:Ta atomic ratio Surface area (m2/g) 1 33.8 1.2 1.83:0.09:2 7.86 2 33.0 2.0 1.85:0.09:2 2.58 3 31.8 3.2 1.88:0.08:2 0.89 4 27.0 8.0 1.91:0.08:2 0.65

Table 1 Composition and surface areas of the KNTO samples.

The samples were characterized by powder X-ray diffraction (XRD) on a Bruker D8 X-ray diffractometer with Cu K_α radiation, 0.02° step size, and 0.2 s step time. The compositions of the as-synthesized samples were determined by inductively coupled plasma emission spectrometry (ICP; Perkin-Elmer, Optima 5300 DV) after the sample was dissolved in a mixture of HNO₃ and HF solutions. Thermogravimetric (TG) characterization was performed on a NETZSCH STA 449F3 instrument. The sample was heated in an alumina crucible under air flow with a heating rate of 10 °C/min. Scanning electron microscope (SEM) images were collected on a field emission SEM (FESEM, ZEISS SUPRA55VP). Ultraviolet-visible (UV-vis) absorption spectra of the samples were recorded on a spectrophotometer (Shimadzu SolidSpec-3700DUV). Fourier transform infrared (FT-IR) spectra were obtained from a Perkin-Elmer 1600 FT-IR spectrometer with a KBr disk. The Brunauer-Emmett-Teller (BET) surface area was measured with a gas sorption instrument (Autosorb-iQ, Quantachrome Instruments).

The photodegradation of RhB was carried out with 20 mg of powdered photocatalyst suspended in 100 mL of RhB solution (5 mg/L) at room temperature. An 8 W bactericidal lamp, with a maximum emission at 254 nm, was used as the light source. Prior to irradiation, the solution was first ultrasonicated for 5 min and then magnetically stirred in the dark for 30 min to ensure the establishment of an adsorption-desorption equilibrium between the photocatalyst and RhB dye. During the photocatalytic progress, after each appropriate time, a 3-mL solution was sampled and centrifuged to remove the particles. The filtrate was analyzed by measuring the change of maximum absorption (553 nm) in the UV-vis spectrum (Shimadzu UV-1800).

XRD patterns of the hydrothermally derived samples in the presence of different KOH concentrations (Table 1) are shown in Fig. 1(a). The XRD patterns were assigned to the pyrochlore structure with space group Fd3m ^[18], indicating that these compounds consist of a single phase. As shown in Table 1, the ICP analysis indicates that the atomic ratios of K:Na:Ta in the samples are close to 1.9:0.1:2. TG analysis of all prepared samples showed a total weight loss of 6.1%-6.9% between room temperature and 600 °C for the products (Fig. 1(b)). The TG curves are in agreement with the previous reports and the weight loss can be attributed to loss of water ^{[18, 19]}. Based on the TG and ICP analysis, the chemical formula of the samples prepared in different KOH concentrations was determined as K_1.9Na_0.1Ta₂O₆·2H₂O (KNTO) for concision. The FT-IR spectra of the obtained samples recorded in the range 400-4000 cm⁻¹ are shown in Fig. 1(c). The broad band around 3300 cm⁻¹ is normally ascribed to the O-H stretching vibrations of water molecules in the crystal structure. FT-IR peaks appearing around 1720 cm⁻¹ can be assigned to the bending vibration of water. The FT-IR results are consistent with a previous report ^[19].

Fig. 1. XRD patterns (a), TG curves (b), and FT-IR spectra (c) of KNTO synthesized under different volumes of KOH added in the hydrothermal reaction.

In our synthetic process, KNTO particles with various sizes and shapes are achieved under different KOH concentrations in the hydrothermal process. Fig. 2 shows the FESEM images of the KNTO synthesized by varying the volume of 5 mol/L KOH used from 1.2 to 2.0, 3.2, and 8.0 mL. It can be seen that the KNTO products are composed of a large quantity of monodispersed particles. As shown in Fig. 2(a), nanospheres with good uniformity of shape were obtained when 1.2 mL of 5 mol/L KOH was added under hydrothermal conditions. The diameter of the nanospheres was approximately 150 nm. As shown in the inset of Fig. 2(a), the nanosphere was assembled by many tiny particles of diverse sizes, which resulted in a relatively rough surface of the spheres. When the volume of KOH solution was increased to 2.0 mL, spheres with a diameter from approximately 200 nm to 2.5 μm were produced, as shown in Fig. 2(b). The magnified view clearly indicates that the surfaces of the spheres were very smooth. Upon increasing the volume of the aqueous KOH solution from 2.0 to 3.2 mL, the morphology of the KNTO sample changed from spherical to cuboctahedral (or truncated octahedral) with an edge length from approximately 3 to 4 μm (Fig. 2(c)). The FESEM image in the inset of Fig. 2(c) also reveals that the vertices of the cuboctahedron are not perfectly flat, but are gently upward-curving. These features distinguish them from the typical cuboctahedral structure with flat faces. More interestingly, in the presence of 8.0 mL aqueous KOH, a durian-like morphology with numerous pyramid vertices was obtained, as shown in Fig. 2(d), which proves that this simple approach can enable the acquisition of unusual complex microstructural features. The broad size distribution shown in Fig. 2 might be caused by the quick nucleation of the KNTO under hydrothermal conditions ^[20].

Fig. 2. SEM images of KNTO synthesized with various morphologies. (a) Nanospheres; (b) Microspheres; (c) Cuboctahedra; (d) Durian-like morphology. The insets are magnified individual microstructures.

First, the soluble (K,Na)₈Ta₆O₁₉·mH₂O was obtained by pretreatment of the Ta₂O₅ in hydroxide melts ^{[4, 5]}. When the concentration of K⁺ reached or exceeded supersaturation, KNTO nuclei were formed under hydrothermal conditions, which served as seeds for subsequent growth into larger particles ^[6]. As reported, the chemical potential of the reaction system, as a driving force for the crystal growth, is mainly determined by the concentration of KOH ^[20]. Theoretically, the KOH quantities can cause the transformation of the KNTO morphology, which is consistent with the observations from FESEM images. When V_KOH = 1.2 and 2.0 mL, the formation of the spheres followed the nucleation, growth, and aggregation processes in the hydrothermal condition, which has been discussed in detail previously ^[5]. The nonspherical structures shown in Fig. 2 result from the anisotropic growth rates of the crystal along the crystallographic directions. It is easy to determine that the cuboctahedra is bounded by eight (111) planes of the cubic lattice of KNTO. In our experiments, the formation of the cuboctahedral KNTO arises from the slow growth rate of the (111) faces, which is induced by increasing the amount of KOH from 2.0 to 3.2 mL. From the observation of the KNTO hierarchical structure in Fig. 2(d), it is presumed that this morphology can be considered to be generated by the interpenetrated growth of the octahedrons. According to a previous report ^[21], the geometrical shape of a crystal with cubic lattice is determined by the ratio (R) of the growth rate along the <100> versus <111> directions. A perfect octahedron will result when R is 1.73, and the cuboctahedra is formed when R is 0.87-1.73 ^[21].

The UV-vis diffuse reflectance spectra of KNTO synthesized under different conditions (V_KOH = 1.2-8.0 mL) are shown in Fig. 3. It reveals that there is a small red shift of the absorption edge as the V_KOH used in the hydrothermal reactions is increased. The bandgaps of nanospheres, microspheres, cuboctahedral particles, and the durian-like sample are 4.88, 4.82, 4.77, and 4.73 eV, respectively. Generally, the electronic band structures are affected by the morphology, size, and defect of the photocatalysts ^[22]. The larger band gap energy of KNTO synthesized under the condition of V_KOH = 1.2 mL might result from the small particle sizes. A similar phenomenon has been observed for the tantalite Sr₂Ta₂O₇ ^[3]. It can be seen from Fig. 2 that the sizes of the other three samples (V_KOH = 2.0-8.0 mL) are all in the micrometer range. Therefore, for these three samples, the effect of particle size on the electronic structure can be neglected. The difference in the band gap energy is probably caused by the different surface atomic arrangement because each morphology possesses its own unique surface atomic arrangement ^[23].

Fig. 3. UV-vis diffuse reflectance spectra of the KNTO synthesized under different volumes of KOH added in the hydrothermal reaction.

The photocatalytic activities of the as-prepared catalysts with different morphologies are evaluated using RhB as the probe molecule under UV light irradiation. Fig. 4(a) shows the time profiles of C_t/C₀ under UV irradiation of KNTO samples, where C₀ is the initial concentration of RhB at the absorption equilibrium of the photocatalyst before irradiation, and C_t is the corresponding concentration after irradiation for time t. As shown in Fig. 4(a), the nanospheres (synthesized at V_KOH = 1.2 mL) with the largest surface area (7.86 m²/g) exhibited the highest photoactivity for RhB photodegradation. The photoactivity of the samples decreased with the increasing V_KOH added in the hydrothermal process. The temporal evolution of the spectrum of RhB during the photodegradation over KNTO nanospheres is shown in Fig. 4(b). It can be seen that the characteristic peak of RhB at 553 nm is gradually decreased with increasing photocatalytic reaction time. After 60 min of irradiation with UV light, RhB is almost completely degraded by the KNTO nanospheres.

Fig. 4. (a) Variation of RhB concentration during photocatalytic reaction with KNTO samples; (b) Temporal absorption spectral changes of RhB in the presence of KNTO nanospheres under UV light irradiation; (c) Repeated photocatalytic experiments of the KNTO nanospheres.

It is known that the activity of a photocatalyst is affected by many factors, including crystallinity, surface area, and active sites ^[5]. From Fig. 1(a), it can be seen that the sharp Bragg peaks of the four samples exhibit similar intensity, indicating that the crystallinity of the four samples is similar. Therefore, we can conclude that the different performance of the four samples is not as a result of crystallinity. As shown in Table 1, the surface area of the KNTO samples decreases with the amount of KOH used in the hydrothermal process. This phenomenon is consistent with the consequence of the photocatalytic activity of the samples. The larger surface area can provide more active sites for the photocatalytic reaction, thus, the surface area may be the main factor that can contribute to the photocatalytic performance of the four samples. Moreover, as shown in Fig. 2(a), the KNTO nanospheres with a rough surface are composed of many tiny particles of diverse sizes. As reported, the boundaries among the primary nanoparticles might be beneficial for the separation of photogenerated electron-hole pairs, which might be helpful for the enhancement of photocatalytic activity of the KNTO nanospheres. Therefore, the KNTO nanospheres show a significant advantage over the other three samples in RhB degradation. The stability of the KNTO nanospheres in RhB degradation was also tested under the same reaction conditions. After complete degradation, the sample was separated by centrifugation and washed several times with deionized water and methanol. The recovered catalyst was used for the next experiment. Fig. 4(c) shows the results of RhB degradation for four runs. It suggests that no obvious decline in RhB degradation occurs after four cycling tests for the KNTO nanospheres, indicating their excellent photostability. The XRD pattern and FESEM image of the sample after the reaction are shown in Fig. 5. No significant changes were found in the crystal structure and morphology of the used photocatalyst, confirming the stability of the photocatalyst.

Fig. 5. (a) XRD patterns of KNTO nanospheres before (1) and after (2) photocatalytic reaction; (b) FESEM image of KNTO nanospheres after photocatalytic reaction.

A two-step synthetic method has been used for the controlled synthesis of KNTO at a low temperature. By fine tuning a single parameter, the amount of KOH added in the hydrothermal process, KNTO with spherical, cuboctahedral, and durian-like morphologies can be selectively prepared. Interestingly, the as-obtained crystals showed different photocatalytic performance, mainly depending on the morphology. The surface areas of the samples have been confirmed to be responsible for their final activity. This technique is expected to be applicable for the precise control of the shape of other tantalate materials that exhibit promising physicochemical properties.