For the synthesis of tubular PbSnO₃nanostructures, two same surfactant–water solutions were first prepared by dissolving 0.2 g poly(vinyl pyrrolidone) (PVP) surfactant in 25 mL distilled water, respectively. Then, equivalent amounts of Pb(AC)₂and Na₂SnO₃(2 mmol) were dissolved in the above surfactant–water solution at room temperature, separately. After stirred for 30 min, the solutions were mixed together and kept stirring for another 30 min, which were then transferred into a Teflon-lined stainless steel autoclave and subsequently heated at 180 °C for 16 h in an oven. After cooling to room temperature, the yellow precipitate was filtered and washed for several times with distilled water and ethanol, respectively, then dried in air at 70 °C. PbSnO₃nanoparticles were also synthesized in this work using a similar process without the use of surfactant PVP. Brief flowcharts illustrating the formation of PbSnO₃nanostructures are shown in Scheme 1.

To compare the photocatalytic properties, bulk PbSnO₃was also synthesized by selecting optimal experimental parameters including calcinations temperature and time. For the synthesis of PbSnO₃bulk material, we first dissolved equivalent amounts of Pb(AC)₂and Na₂SnO₃into distilled water under stirring, and then mixed them to obtain the white precursor. Heating the white precursor at 500 °C for 5 h in a quartz tube under Ar flow resulted in yellow powders. In this process, temperature is very important for the formation of yellow powders due to the instability of PbSnO₃at high temperature.

The crystal structure of the as-prepared sample was confirmed by the X-ray diffraction pattern (JEOL JDX-3500 Tokyo, Japan). The morphology and size of the sample were characterized by transmission electron microscope (HRTEM, JEM-3000F) equipped with an X-ray dispersive spectrometer (EDS). UV–Vis diffuse reflectance spectra were recorded on a UV/Vis spectrometer (UV-2500, Shimadzu) and were converted from reflection to absorbance by the standard Kubelka–Munk method. The surface area of the sample was measured by the BET method (Shimadsu Gemini Micromeritics).

The photoactivities of the obtained PbSnO₃nanostructures were evaluated by decomposition of gaseous IPA under visible light irradiation. Typically, 0.1 g PbSnO₃catalyst was spread uniformly in a quartz-made vessel with an irradiation area of 7.8 cm². Prior to light irradiation, the vessel was kept in dark for 2 h until an adsorption–desorption equilibrium was finally established. The visible light with light intensity of about 1.8 mW/cm²was obtained by using a 300 W Xe lamp with a set of combined filters (L42 + B390 + HA30) and a water filter. The products in the gas phase were analyzed with a gas chromatograph system (GC-14B, Shimadzu, Japan), using a flame ionization detector (FID) for organic compounds determination.

The crystal structure of both as-synthesized PbSnO₃nanostructures from the hydrothermal process and bulk material from the solid-state route were characterized by XRD and the results are shown in Fig. 1. In these patterns, all peaks can be indexed as cubic phase PbSnO₃with pyrochlore-type structure (space group:Fd3m). The calculated lattice constanta = 10.67 Å is in agreement with previously reported value (JCPDS 17-060). From the XRD patterns, it can be clearly seen that the PbSnO₃nanostructures are of better crystallinity than the bulk material, which might be one of the reasons why nanostructured PbSnO₃show higher photocatalytic activities (detailed contents in the part of discussion). Inset in Fig. 1is a typical SEM image of the product from the SSR. Scheme 2 shows the crystal structure of pyrochlore-type PbSnO₃, an anion-deficient three-dimensional framework consisting of corner-sharing SnO₆octahedra.

Figure 2a shows a TEM image of as-prepared PbSnO₃nanoparticles from the hydrothermal process. Obviously, the products are consisted of many small nanoparticles with dimensions in the range of 10–15 nm. The corresponding selected-area electron diffraction (SAED) pattern (Fig. 2b) can be readily indexed as cubic phase PbSnO₃, which is in agreement with the XRD result. An EDS spectrum in Fig. 2c depicts the presence of Pb, Sn, and O elements, indicating the formation of PbSnO₃. In this spectrum, the signals corresponding to Cu arise from the TEM grid. The microstructures of the produced PbSnO₃nanoparticles were investigated using high-resolution TEM. As indicated in Fig. 2d, the nanoparticles are well-crystallized and of good crystallinity. The marked lattice fringes of 0.32 and 0.25 nm correspond well to the (311) and (331) crystalline planes of cubic PbSnO₃.

In the presence of surfactant PVP, polycrystalline PnSnO₃nanotubes were obtained instead of nanoparticles. Panels (a) and (b) of Fig. 3are typical TEM images of as-obtained PnSnO₃nanotubes, which reveal that the nanotubes are polycrystalline with typical diameters of 300–340 nm and wall thickness of 40–80 nm. Figure 3c is the corresponding SAED pattern taken from a single PbSnO₃nanotube, confirming the formation of polycrystalline nanotube. The three polycrystalline rings are in accordance with those of (311), (400), and (533) of cubic phase PbSnO₃. Typical HRTEM images of the nanotubes are shown in Fig. 3d and e. It can be seen that the polycrystalline PbSnO₃nanotubes are composed of numerous nanoparticles with diameters of several to ten nanometers. The interplanar spacing was calculated to be about 0.32 nm, corresponding to the (311) plane of cubic PbSnO₃, in accordance with the SAED result.

UV–Vis spectra of all three PbSnO₃samples were checked and the spectra are displayed in Fig. 4. It is evident that PbSnO₃nanostructures could absorb much more visible light than bulk sample at the present condition. Corresponding band gaps of PbSnO₃are determined to be 2.8 eV for bulk material, 2.8 eV for nanotubes, and 2.7 eV for nanoparticles from the absorption edges, respectively (as shown in Table 1).

^aBulk PbSnO₃are prepared from the solid-state route

One-dimensional micro- or nanosized tubular materials with hollow interior structure have attracted extraordinary attention owing to their unique properties and potential applications 14 15 16. Many kinds of growth mechanisms have been proposed for the formation of nanotubes. For example, the rolling mechanism and template-assisted mechanism have been reported to explain the formation of tubular structure with layered or pseudo-layered structures such as BN 17 , NiCl₂ 18 , Nb₂O₅ 19 , Se 20 , etc. During the growth of PbSnO₃ nanotubes, surfactant PVP was used and was found to be the key issue for nanotube growth. Thus, the surfactant-assisted growth process can be used to explain the formation of these nanotubes. The possible formation process of PbSnO₃ nanotubes may involve three following distinctive stages: (i) the generation of PbSnO₃ particles, (ii) the adsorption of PVP molecules on the surface of particles and subsequently self-assembly into tubular microstructure, and (iii) the formation of uniform PbSnO₃ nanotubes. In the initial stage, cubic PbSnO₃ tiny nuclei could easily crystallize and serve as the seeds for the growth of nanotubes. Meanwhile, PVP molecules in the solution would strongly and rapidly adsorb on the surfaces of these nascent nuclei, which confined the crystal growth and efficiently controlled the dimension and morphology of the final products. Then, these particles with high free energy aggregated and self-assembled into tubular structures with the help of PVP template molecules. As a result, the growth of PbSnO₃ nanotubes would form eventually by a typical oriented attachment process under the hydrothermal conditions. Meanwhile, the existence of PVP in this solution can alter the surface energies of various crystallographic surfaces to promote selective anisotropic growth of nanocrystals 21 .

The photocatalytic activities of the PbSnO₃nanostructures were evaluated by IPA mineralization under visible light irradiation. Under visible light irradiation, gaseous IPA was gradually oxidized through an acetone intermediate to CO₂, and the concentration changes of IPA, acetone, and CO₂versus time over PbSnO₃nanoparticles are shown in Fig. 5. It was clear that the concentration of IPA in the reaction system almost decreased from the initial concentration to zero; the concentration of acetone also decreased continually while the concentration of CO₂increased with the long-term irradiation. Inset in Fig. 5shows that almost no additional CO₂gas was detected under dark test, suggesting that degradation of IPA over the catalyst was driven by light irradiation. Figure 6further displays the concentration changes of evolved acetone over different PbSnO₃nanostructures and bulk material with the increasing of irradiation time. Clearly, acetone was detected over all these catalysts when light was turned on. Among them, particulate PnSnO₃performs the best activity for degradation of IPA under the present conditions.

In this case, the photocatalytic activities for IPA degradation over these catalysts were in the order of nanoparticle > nanotube > bulk material, which was in consistent with that of BET surface areas. As mentioned earlier, BET surface area of catalyst is closely related to its photoactivity. Usually, larger surface area means much more active sites, at which the photocatalytic reaction occurs. Thus, as shown in Table 1, PbSnO₃nanostructures with larger surface areas as 68 m²/g for nanoparticles and 50 m²/g for nanotubes, respectively, resulted in enhanced photocatalytic activities than bulk material with 10 m²/g of surface area. Meanwhile, the improved crystallinity of PbSnO₃nanostructures (shown in XRD patterns) resulted in the increase of photocatalytic activity since it could reduce electron-hole recombination rate.

The wavelength dependence of the rate of acetone evolution from IPA degradation over PbSnO₃ nanoparticles was investigated by using different cutoff filters, as shown in Fig. 7. The intensity variation of the incident light with different cutoff filters is given as an inset figure for reference. It is notable that the rate of acetone evolution decreased with increasing cutoff wavelength, which is in good agreement with the UV–Vis diffuse reflectance spectra of PbSnO₃ nanoparticles, indicating the present reaction is driven by a visible light absorption. The used catalysts were again checked by XRD and UV–Vis reflectance spectroscopy to explore the stabilities of samples. There was no detectable change between the spectra of PbSnO₃ before and after the photodegradation of IPA gas, suggesting that the catalyst was fairly stable for the degradation of organic compounds. For many p-block metal oxides photocatalysts with d¹⁰ configuration, the VB and CB are the 2p orbital of the oxygen atom and the lowest unoccupied molecular orbital (LUMO) of p-block metal center, respectively 22 23 24. Meanwhile, for the lead-containing compounds, it was found that an additional hybridization of the occupied Pb 6s and O 2p orbitals seems to push up the position of the valence band and result in a narrower band gap 25 . Based on the above depiction, we assumed that the VB of PbSnO₃ is composed of hybridized Pb 6s and O 2p orbitals, whereas the CB is composed of Sn 5s orbitals, and these bands meet the potential requirements of organic oxidation.