UV-induced formation of activated Bi₂O₃ nanoflake: an enhanced visible light driven photocatalyst by platinum loading

Abstract

Pt/Bi₂O₃ nanoflakes were designed as efficient visible light-driven photocatalyst which could be attributed to the synergy of its UV-induced nanostructure, surface phase and Pt-induced multielectron O₂reduction.

---

Tailoring the architecture of nano/microcrystals has been of great research interest due to their unique shape and size-dependent properties. In this respect, bismuth oxide (Bi₂O₃) is an attractive material due to its good optical and electrical properties and semiconductor characteristics. It has been widely used in various applications such as microelectronics, gas sensor, and optical coatings.^1,2 Furthermore, Bi₂O₃ has also proved to be an efficient photocatalyst for water splitting and pollutant decomposing under visible light irradiation. Recently, Bi₂O₃ nanocrystals with various morphologies have been achieved. For example, Liet al.³ reported the fabrication of Bi₂O₃ nanotubes from fast oxidizing the bismuth nanowires in air by a template-based heat-treatment method. Ultrathin Bi₂O₃ nanowires were prepared by thermal evaporation technique by Yang et al.⁴Bi₂O₃ microrods have been produced using BiI₃ and O₂ as a starting material.⁵ Chen et al. reported mesh-like Bi₂O₃ single crystalline nanoflakes via a bismuth oxalate precursor way.⁶ Our group also reported a free-standing 3D microjagged Bi₂O₃ prepared through a nontraditional template method.⁷ Accordingly, a direct correlation between the size, nanostructures and surface phase of the photocatalyst and its photocatalytic performance is of great significance, the availability of efficient Bi₂O₃ photocatalysts with combining the above factors is the main motive of this paper.

In this communication, we report the first example of porous bismuth carbonate microflowers prepared by urea-assisted alcoholysis of bismuth nitrate as shown in ESI.† Thermal treatment is generally considered to be the most classic way to obtain oxides. However, thermal treatment adversely affected their physical performances. Undesirable crystal aggregation was involved in the transformation from bismuth carbonate to bismuth oxide (Bi₂O₃). The porous structure of bismuth carbonate microflowers was totally destroyed forming Bi₂O₃ micronuts with compact surface and accompanying rapid decrease of specific surface area. In general, thermal-induced tubular morphology collapse of nanotubes and crystal aggregation of nanoparticles^8,9 are regarded as unavoidable thermodynamic behaviors of nanomaterials. However, our recent research demonstrated that these thermodynamic behaviors can access to be solved by special methods.¹⁰ In this paper, UV light irradiation was available to ‘make Bi₂O₃ micronuts rebloom’. The surface mesopores in Bi₂O₃ micronuts were successfully reopened forming Bi₂O₃ nanoflakes which were simultaneously introduced active species on the surface. Further platinum deposition on the surface of Bi₂O₃ nanoflakes structure that results from combining the ideas of nanostructural control with a Pt/Bi₂O₃ multicomponent composite as a heterocatalyst should therefore be more promising for visible-light-driven photocatalysis. Pt-loaded Bi₂O₃ nanoflake was proven to exhibit remarkably high photocatalytic activity for aqueous acetaldehyde (AcH) decomposition because of the synergy of its UV-induced nanostructure, surface phase and Pt-induced multielectron O₂reduction.

The X-ray diffraction (XRD) profiles shown in Fig. 1 are for products obtained from the hydrothermal reaction (Fig. 1a), after calcination and UV irradiation (Fig. 1b and c). The sample from the hydrothermal reaction was identified as crystalline bismuth carbonate (JCPDS No.41-1488). After calcination, the bismuth carbonate decomposed into Bi₂O₃, as confirmed by the presence of characteristic peaks of monoclinic Bi₂O₃ (α-Bi₂O₃, JCPDS No.41-1449) in the XRD pattern. No other phases were detected in the spectrum, indicating that the impurities in the sample after calcination were out of the detection range of XRD. The XRD profile of the irradiated Bi₂O₃ shows the main phase, α-Bi₂O₃, with two new minor peaks at 2θ = 27.9° and 32.3° (Fig. 1c). The two characteristic reflections can be clearly indexed with the formation of the oxidized Bi₂O_4-x species⁷ on the surface of α-Bi₂O₃.

Fig. 1 XRD patterns of a) bismuth carbonate microflowers before calcination and b) calcined Bi₂O₃ micronuts and c) UV-irradiated Bi₂O₃ nanoflakes.

The morphology of the collected bismuth-based compounds is observed using the field-emission scanning electron microscope (FESEM). Fig. 2a shows SEM image of bismuth carbonate, which overall reveals many uniform flowerlike textures. These flowerlike bismuth carbonates are composed of many nanosheets as the petals with an average thickness of 20–30 nm, and these nanosheets interweave together forming an open porous structure. Fig. 2b and c depict a group of fully calcined Bi₂O₃ (500 °C, 1 h). Clear observation shows that the bismuth carbonate microflowers present obvious signs of aggregation after calcination, converting into Bi₂O₃ with nutlike shape and each micronut is constructed with two or three microflowers. The top-view SEM image (Fig. 2c) shows the micronut is actually reassembled by ordered nanosheets, rather than unorderly aggregation. This inspires us the possibility to separate these nanosheets again. Very interestingly, the ordered layers peel off leading to the formation of mesporous structure again when we employ the UV radiation, as shown in Fig. 2d. Energy dispersive X-ray analysis (Fig. S1a and b in ESI) indicates that bismuth carbonate converts into Bi₂O₃ in absence of carbon peak after calcination.

Fig. 2 SEM images of bismuth carbonate microflowers (a) and Bi₂O₃ micronuts after annealing (b, c) and UV-irradiated Bi₂O₃ nanoflakes (d).

It has been reported that large surface areas are desirable for numerous applications of nanostructures to fulfil the demand of high efficiency and activity.¹¹ The porosity and specific surface area changes in the morphology evolution of the products were investigated by nitrogen cryosorption studies. Bismuth carbonate microflowers were found to hold remarkable BET surface area of 120 m² g⁻¹. However, the specific surface area of the product rapidly decreased to 25 m² g⁻¹ due to the thermal aggregation when bismuth carbonate transformed into compact Bi₂O₃ micronuts. Excitingly, a significant increase in the surface area (95 m² g⁻¹) was observed after UV irradiation, in good agreement with SEM observations. The nitrogen absorption of the products exhibits the type IV characteristics (Fig. S2a in ESI). Fig. S2b in ESI shows the pore-size distribution. The pore-size distribution of bismuth carbonate microflowers predominantly locates in the range from 10 to 50 nm (mesopores), which can be attributed to the porous structure of microflowers. The pores from Bi₂O₃ micronuts reveal the absence of mesopores with the fusion of nanosheets, whereas the pores (range in 40–100 nm) are found again as the ordered nanosheets separated after UV irradiation. These mesopores are quite suitable to accommodate nanoscale Pt particles and allow for more efficient absorption of incident photons as well as decreased bulk recombination.

In this regard, two mainly striking features were produced by the employment of UV light irradiation including the formation of Bi₂O₃ nanoflakes with laminar porous nanostructure and the introduction of surface oxidized species on the surface of Bi₂O₃ nanomaterials. The band gap value of 2.4 eV is calculated for the irradiated Bi₂O₃ (Fig. S3 in ESI). This value is much lower than that of the calcined Bi₂O₃ (2.7 eV). The high band gap value of the irradiated material is likely attributed to originating from bulk O(2p)–Bi³⁺(6p°) excitations and surface O(2p)–Bi⁵⁺(6s°) transitions made available by UV irradiation treatment. Because highly oxidized species are formed by reacting with oxygen in solution in the case of calcined Bi₂O₃ under UV irradiation. Among these reactive species, the photogenerated superoxide radicals with the strongest oxidation ability and enough long lifetime¹² can in return oxidize Bi³⁺ to Bi⁵⁺ with the formation of reactive Bi₂O_4-x at the surface. We have also proved that the surface oxidation won't occur without O₂ in the solution. On the other hand, It is worthy to note that the UV light induced valence change from Bi³⁺ to Bi⁵⁺ would lead to the accumulation of electrons between the Bi₂O₃ nanosheets. With prolongation of UV light irradiation time, accommodating enough accumulated electrons between the nanosheets tended to repel each other and finally formed the laminar porous nanostructures. To confirm the above hypothesis, the intermediate product by irradiation for shorter time (10 min) was checked. From the SEM images (Fig. S4 in ESI), the compact surface of Bi₂O₃ micronuts became partially wavy and appeared a small part of clear furrows, verifying our proposed formation process very well. It thus reveals that the adopted UV irradiation is an ideal structural modification approach for Bi₂O₃, which should be minor impact for TiO₂–based material. So far, only deliberate designed thermal treatment helps to obtain interesting anatanse/rutile nanocompositions with enhanced photocatalytic activity in TiO₂–based material.¹³

Actually, Bi₂O₃ as a promising visible light driven photocatalyst has been more and more reported.¹⁴ Though its surface state can be easily modified by some high oxidative reagent such as Na₂S₂O₈¹⁵ and H₂O₂⁷ or under some high-energy radiation like UV light,¹⁶ it is stable under visible light driven photocatalytic reaction. Thus we can say that the surface treatment of Bi₂O₃ under UV light radiation is a photooxidation process while the decomposition of organic pollution under visible light radiation is a photocatalytic reaction.

The photocatalytic activity is evaluated by the amount of CO₂ that evolved from the oxidative decomposition of acetaldehyde solution (AcH aq) upon visible light irradiation, as described in the Experimental Section in the ESI. The photocatalytic activity of UV-irradiated Bi₂O₃ nanoflakes is better than that of calcined Bi₂O₃ micronut and nitrogen-doped TiO₂ (N-TiO₂) at each Pt-loading concentration, as shown in Fig. S5 in ESI. 1 wt% Pt loading produces the highest activity for all the photocatalysts and an excessive Pt loading decreases the photocatalytic activity due to the competitive photon absorption of the photocatalytically inactive Pt nanoparticles with the photocatalysts. Fig. 3a shows that the rate of CO₂ generation over both UV-irradiated Bi₂O₃ nanoflakes and calcined Bi₂O₃ micronuts from aqueous AcH increased rapidly after Pt loading, achieving the maximum rate (131 μmol h⁻¹) at 1 wt% Pt-loaded Bi₂O₃ nanoflake. This rate of CO₂ generation is ca. 6.3 and 30 times higher than that over pure Bi₂O₃ nanoflake (21 μmol h⁻¹) and Bi₂O₃ micronut (4.2 μmol h⁻¹), respectively. It is necessary to note that the enhanced activity in UV-irradiated Bi₂O₃ nanoflake compared to calcined Bi₂O₃ micronut could be associated with the formation of surface phase junctions between the bulk Bi₂O₃ and surface Bi₂O_4-x when calcined Bi₂O₃ micronut was irradiated under UV light. We have proved that such surface phase junctions could promote the spatial charge separation in the surface region leading to high photocatalytic efficiency, and the high efficient photocatalytic activity can only be available under a certain ratio between bulk Bi₂O₃ and surface Bi₂O_4-x. Excess surface Bi₂O_4-x would instead decrease the photocatalytic activity.⁶ Furthermore, the UV-induce opening of laminar pores on the surface would be the other reason. These laminar pores could not only benefit photons' effective absorbance comparing to a flat surface due to light scattering,^17,18 but also allow the rapid diffusion of the reactants and products during the reaction. Herein, N–TiO₂ as a commercial visible light-responsive photocatalyst is also utilized in the control experiment. Although detectable amounts of CO₂ gas were liberated over N–TiO₂, the rate (ca. 25 μmol h⁻¹ with optimal Pt loading) had just little increase compared to bare N–TiO₂ and was much lower than that over irradiated Bi₂O₃ after Pt loading. The final molar amount of CO₂ generation over Pt–Bi₂O₃ nanoflake is almost twice that of AcH used, indicating that the complete decomposition of AcH solution underwent in this reaction. We also investigate the stability of the UV-irradiated Pt/Bi₂O₃ catalyst during the decomposition of AcH. After ten reaction cycles, Pt/Bi₂O₃ nanoflakes showed a negligible decrease of the activity (Fig. S6 in ESI), and the XRD and SEM images of the photocatalysts after photocatalytic reaction showed no difference compared to the original samples (data no shown), suggesting that the photocatalyst is durable under aerobic, visible-light illumination conditions.

Fig. 3 Time course of CO₂ generation from the decomposition of (a) acetaldehyde (AcH) and (b) isopropyl alcohol (IPA) over 1 wt% Pt-Bi₂O₃ nanoflake (1), 1 wt% Pt-Bi₂O₃ micronut (2), 1 wt% Pt-N-TiO₂ (3), UV-irradiated Bi₂O₃ nanoflake (4), N-TiO₂ (5) and calcined Bi₂O₃ micronut (6) under visible light.

In the end of this paper, we try to explore a plausible mechanism for the enhanced photocatalytic ability of Pt-loaded Bi₂O₃ nanoflakes. From the morphological view, UV irradiation induced reopening the bulk Bi₂O₃ micronuts forming Bi₂O₃ nanoflakes with large specific area. The shape and structure change is decisive to photocatalytic activity of materials. The calcined and UV-irradiated Bi₂O₃ after Pt loading were observed by TEM (Fig. S7 in ESI). For UV-irradiated Bi₂O₃ nanoflake, the laminar mesopores are though a little hard to observe clearly in TEM, as shown in Fig. S7b in ESI, some laminar mesopores are still visible and they do exist from SEM and pore size measurement. These laminar mesopores naturally act as Pt-buffering sites. Most Pt particles with small size (3–5 nm) preferentially entered these laminar pores leading to uniform loading on the surface of Bi₂O₃ nanoflakes or in the mesopores, while Pt nanoparticles can only be photodeposited on the smooth and compact surface of calcined Bi₂O₃ micronuts in Fig. S6a in ESI. The UV-induced formation of Bi₂O₃ nanoflakes results in a larger effective surface area and guarantees more active sites in close to the Pt nanoparticles, thus enabling diffusive transport of photogenerated electrons to these particles. Further from energetic view, the CB levels of Bi₂O₃ (+0.33 V versusNHE)¹⁹ with visible light absorption are more positive than the reduction potentials of O₂ leading to the inefficient oxidation decomposition of isopropyl alcohol (IPA) in air (Fig. 3b). IPA is known as the counterpart of holes consumption. Photogenerated electron reduction of O₂ to produce superoxide or active radicals predominately proceeds in the decomposition of IPA. Notably, the oxidation decomposition of IPA is accelerated by Pt loading. This result strongly indicates that the accumulated photogenerated electrons on the surface of Pt particles working as an electron magnet provide sufficiently negative potential to catalyze O₂reduction. This is called multielectron-induced reduction, which has been reported in the case of Pt–WO₃ photocatalysts.²⁰ This is the first time to find Pt-induced multielectron reduction occurred in Bi₂O₃ samples. While single-electron reduction has been reported to mainly proceed over TiO₂ and N–TiO₂ photocatalysts,^21–23 and this is reasonably supported by the Pt-induced improvement of the photoactivity of N-TiO₂ was remarkably less than that for Bi₂O₃ (2.7-fold versus 100-fold). The cooperation of multi-electron assembled by Pt particles for O₂reduction is superior to single-electron reduction and that's the reason why Pt loading contributes to high photocatalytic activity under visible light.

Conclusions

We have developed a simple and low cost approach to obtain Bi₂O₃ nanoflake from the surface compact Bi₂O₃ micronut through employing UV-visible radiation. The UV-visible radiation induced the formation of the nanostructure of Bi₂O₃ nanoflake with large specific area and the formation of surface active species (Bi₂O_4-x) on surface Bi₂O₃ simultaneously. After Pt loading, the Pt/Bi₂O₃ nanoflake showed higher photocatalytic efficiency for the degradation of aqueous acetaldehyde and isopropyl alcohol than commercial N–TiO₂ and optimal Pt-loaded N–TiO₂ under visible light. Such a large activity enhancement probably arises from the synergy of its UV-induced nanostructure, surface phase and Pt-induced multielectron O₂reduction. This work provides a new route for the design of multicomponent photocatalysts with enhanced photoactivity as well as a new material which can be used in solar cells, nanodevices and other applications.

References

1. T. Hyodo, E. Kanazawa, Y. Takao, Y. Shimizu and M. Egashira, Electrochemistry, 2000, 68, 24.
2. G. Bandoli, D. Barreca, E. Brescacin, G. A. Rizzi and E. Tondello, Chem. Vap. Deposition, 1996, 2, 238.
3. L. Li, Y. W. Yang, G. H. Li and L. D. Zhang, Small, 2006, 2, 548.
4. Y. F. Qiu, D. F. Liu, J. H. Yang and S. H. Yang, Adv. Mater., 2006, 18, 2604.
5. R. Chen, Z. R. Shen, H. Wang, H. J. Zhou, Y. P. Liu, D. T. Ding and T. H. Chen, J. Alloys Compd., 2011, 509, 2588.
6. T. Takeyama, N. Takahashi, T. Nakamura and S. Itoh, Solid State Commun., 2005, 133, 771.
7. Z. L. Xu, I. Tabata, K. Hirogaki, K. Hisada, T. Wang, S. Wang and T. Hori, Catal. Sci. Technol., 2011, 1, 397.
8. X. M. Sun and Y. D. Li, Chem.–Eur. J., 2003, 9, 2229.
9. J. N. Nian and H. S. Teng, J. Phys. Chem. B, 2006, 110, 4193.
10. T. Wang, S. Wang, W. X. Chen, W. Wang, Z. L. Xu and T. Hori, J. Mater. Chem., 2009, 19, 4692.
11. J. Tang, Z. Zou and J. Ye, Chem. Mater., 2004, 16, 1644.
12. T. Hirakawa, K. Yawata and Y. Nosaka, Appl. Catal., A, 2007, 325, 105.
13. J. Zhang, Q. Xu, Z. C. Feng, M. J. Li and C. Li, Angew. Chem., Int. Ed., 2008, 47, 1766.
14. (a) S. Y. Chai, Y. J. Kim, M. H. Jung, A. K. Chakraborty, D. Jung and W. I. Lee, J. Catal., 2009, 262, 144; (b) L. Zhou, W. Z. Wang, H. L. Xu, S. M. Sun and M. Shang, Chem.–Eur. J., 2009, 15, 1776; (c) X. H. Wu, W. Qin and W. D. He, J. Mol. Catal. A: Chem., 2007, 261, 167; (d) M. Ge, Y. F. Li, L. Liu, Z. Zhou and W. Chen, J. Phys. Chem. C, 2011, 115, 5220.
15. A. S. Prakash, C. Shivakumara, M. S. Hegde, L. Dupont and J. M. Tarascon, Mater. Res. Bull., 2007, 42, 707.
16. A. Hameed, T. Montini, V. Gombac and P. Fornasiero, J. Am. Chem. Soc., 2008, 130, 9658.
17. G. K. Mor, K. Shankar, M. Paulose, O. K. Varghese and C. A. Grimes, Nano Lett., 2005, 5, 191.
18. H. X. Li, Z. F. Bian, J. Zhu, D. Q. Zhang, G. S. Li, Y. N. Huo, H. Li and Y. F. Lu, J. Am. Chem. Soc., 2007, 129, 8406.
19. Y. Xu and M. A. A. Schoonen, Am. Mineral., 2000, 85, 543.
20. R. Abe, H. Takami, N. Murakami and B. Ohtan, J. Am. Chem. Soc., 2008, 130, 7780.
21. K. Ishibashi, A. Fujishima, T. Watanabe and K. Hashimoto, J. Phys. Chem. B, 2000, 104, 4933.
22. T. Hirakawa and Y. Nosaka, Langmuir, 2002, 18, 3247.
23. A. J. Colussi and M. R. Hoffmann, J. Phys. Chem. B, 2004, 108, 17269.

---

Footnote

† Electronic Supplementary Information (ESI) available: Experimental section, EDXA data, BET and pore-size distribution data, UV-vis spectra data, SEM image and TEM image of as-prepared samples. See DOI: 10.1039/c1ra00638j/

---