Mechanism for catalytic ozonation of p-nitrophenol in water with titanate nanotube supported manganese oxide

Abstract

Manganese oxide supported on titanate nanotubes (TNT) was prepared by an impregnation method and used as a catalyst for ozonation of p-nitrophenol (PNP) in an aqueous solution. Characterization results indicated that the manganese oxide was highly dispersed on the surface of TNT. The synthesized catalyst exhibited high activity for the mineralization of PNP with ozone, and about 95% of the total organic carbon was removed at 45 min. The degradation of PNP was mainly due to the oxidative process in solution, and the hydroxyl radical reaction played an important role for the degradation of its ozonation products (formic acid and oxalic acid). The negatively charged surface and surface acid sites of the support favored the adsorption of ozone, while the highly dispersed MnO_x accelerated the decomposition of adsorbed ozone into hydroxyl radicals. A possible mechanism for the catalytic ozonation of PNP was proposed.

---

Introduction

Water pollution with organic pollutants has become a serious environmental problem in the past few decades. Heterogeneous catalytic ozonation has recently attracted considerable attention as a promising technology for the degradation and mineralization of refractory organic pollutants in water.^1,2 Significant efforts have been made on the preparation of highly active heterogeneous ozonation catalysts. These catalysts can promote the generation of hydroxyl radicals from ozone decomposition and lead to higher efficiency of ozone consumption. However, their application is limited because the mechanisms of these processes are still unclear.³

The efficiency of catalytic ozonation depends on the structure and surface properties of the catalysts. Transition metal oxides with variable electronic structures have exhibited excellent catalytic activity for ozonation of organic pollutants.^4–8 In order to increase the active surface, mesoporous materials with large surface areas were used as the supports of metal oxides.^9–13 The supported oxides with multivalence oxidation states and high dispersion can enhance the interfacial electron transfer, causing the high catalytic activity. In aqueous systems, metal oxide particles can be hydrated and form surface hydroxy groups, which are hypothesized to be active sites in catalytic ozonation. Nevertheless, there are full of contradictory reports on the catalytic activity of metal oxides. These surface hydroxy groups tend to dissociate or protonate, which is dependent on the solution pH. At pHs lower than the pH of the point of zero charge (PZC), the particle surface is positively charged, while it has negative charges above it.¹⁴ Zhang et al. suggested that not all surface hydroxy groups of FeOOH possessed the same high catalytic activity, but the weak surface MeO–H bonds were favorable sites for promoting ˙OH generation from aqueous ozone.¹⁵ Li et al. reported that the positively charged surface of PdO/CeO₂ played an important role for the catalytic ozonation of pyruvic acid.¹⁶ On the contrary, some researchers found that negatively charged surface groups had a major effect on ozone decomposition.^17,18 On the other hand, the surface hydroxy groups of metal oxide can behave as Brönsted acid sites. Some metal oxides have Lewis acid sites and/or Lewis bases. Since ozone can be regarded as a Lewis base, it can react with surface acid groups of metal oxide.¹⁹ For instance, Bing et al. reported that the combination of both Lewis acid sites of iron and aluminium onto Fe₂O₃/Al₂O₃@SBA-15 enhanced the formation of ˙OH and O₂˙⁻ radicals.²⁰ However, some researchers found that alumina did not decompose ozone, and the adsorption of reactants (ozone and/or organic molecules) on the surface might be necessary for its catalytic effect.^21–24 Therefore, the fate of the catalytic activity of metal oxide should be further investigated.

In the present work, manganese oxide was highly dispersed on the surface of titanate nanotube (TNT) by a simple impregnation method. TNT was selected as the support because it contains acid sites and has unusual morphology. The structure, morphology and surface properties of the catalyst were analyzed by different characterization methods. p-Nitrophenol (PNP), a common compound in waste waters, was selected to evaluate the catalytic ozonation activity of the catalyst. The results indicated that the catalyst significantly enhanced the mineralization of PNP. The correlation between the surface properties and catalytic activity of the catalyst was discussed. A possible reaction mechanism was also proposed.

Experimental

Catalyst preparation

All chemicals were of analytical grade and used without further purification. Deionized water was used in this study. TNT was prepared by a hydrothermal method.²⁵ 1 g of anatase phase TiO₂ powders were added into 50 mL of 10 mol L⁻¹ NaOH aqueous solution. The suspension was transferred into an autoclave and heated at 150 °C for 24 h. The product was washed with dilute HCl and distilled water. Supported manganese oxide was prepared by an impregnation method with Mn(NO₃)₂ as the metal precursor. 0.67 g of Mn(NO₃)₂ was dissolved in 100 mL of distilled water. Then, 1 g of TNT was added to this solution. After impregnation, the sample was washed with deionized water to remove the excess manganese that was not connected tightly to the surface of titanate, and dried at 110 °C for several hours and calcined in air at 400 °C for 2 h. As a reference, TNT was also calcined in air at 400 °C for 2 h.

Catalyst characterization

X-ray diffraction (XRD) analysis was performed on a Bruker D8-Advance using Cu Kα radiation. The morphologies were examined on a Hitachi H-7500 transmission electron microscopy (TEM) at 80 kV and a JEM-2100 high-resolution TEM at 200 kV. The content of Mn was determined by induced coupled plasma spectrometer (ICP)-MS (Thermo Fisher). X-ray photoelectron spectroscopy (XPS) data were obtained by an AXIS-Ultra instrument using monochromatic Al Kα radiation (225 W, 15 mA, 15 kV). Nitrogen adsorption–desorption measurements were carried out with a Quantachrome NOVA 4000e surface area analyzer at 77 K. Infrared spectra were recorded on a Nicolet iS 50 Fourier transformation infrared (FTIR) spectrometer with KBr disks in the range of 4000–400 cm⁻¹ at room temperature. The zeta potential of the catalyst suspensions was measured with a Malvern 3000 Zetasizer with three consistent readings.

Catalytic activity measurements

In a typical experiment, 0.5 L of 40 mg L⁻¹ PNP solution and 0.1 g catalyst were added into a 1 L cylindrical reactor. Ozone gas (4 mg min⁻¹ output) generated from a 3S-A3 laboratory ozonizer (Tonglin Technology, China) was continuously fed to the suspension under magnetical stirring. Samples were withdrawn and filtered through a millipore filter (pore size 0.22 μm) for analysis. The residual ozone was removed by adding 0.1 mol L⁻¹ Na₂S₂O₃ solution. The concentration of aqueous ozone was measured with the indigo method (Ozone Standards Committee method). The total organic carbon (TOC) of the solution was determined with a Phoenix 8000 TOC analyzer. PNP was measured by a high performance liquid chromatography (HPLC, Agilent 1200) using a ZORBAX Eclipse XDB-C18 column (4.6–250 mm, 5 μm) at a flow rate of 1 mL min⁻¹ and wavelength of 210 nm. The mobile phase consisted of 60% acetonitrile and 40% water. Organic intermediates were determined using a Poroshell 120 SB-Aq column (4.6–100 mm, 5 μm) at a flow rate of 0.5 mL min⁻¹. The mobile phase consisted of the solution of 0.1% H₃PO₄/H₂O. The concentration of 7-hydroxycoumarin was measured with a Hitachi F-4600 fluorescence spectrometer at 455 nm (excited at 332 nm). In order to avoid the degradation of 7-hydroxycoumarin by excessive ˙OH, the ozone output was decreased to 0.2 mg min⁻¹. Each experiment was run in triplicate. Data were the arithmetic mean of three measured values.

Results and discussion

Characterization of catalysts

The XRD patterns of the products are shown in Fig. 1. The diffraction peaks of TNT located at 11°, 25° and 49° can be assigned to the (001), (011) and (020) faces of Na_xH_2−xTi₃O₇·nH₂O (JCPDS 31-1329), respectively. According to previous report, layered hydrogen titanates nanotubes would be transformed to anatase with the loss of the tubular morphology after calcination at 400–600 °C.²⁶ Therefore, the TNT sample calcined at 400 °C should be Na_xH_2−xTi₃O₇·nH₂O. The XRD pattern of MnO_x/TNT is similar to that of pure TNT. No XRD diffraction peaks of manganese oxide were observed due to its low content and good dispersion. EDX analysis reveals that the Mn content loaded on TNT is about 4 wt%.

Fig. 1 XRD patterns of TNT, MnO_x/TNT and Na_xH_2−xTi₃O₇·nH₂O.

TEM images in Fig. 2 show that both TNT and MnO_x/TNT are nanotubes with diameters of about 10 nm and lengths of several hundred nanometers. Comparing to TNT, the nanotubular structure of MnO_x/TNT became uniform and the tube length became long. The reason might be that some nanotubes collapsed during calcination and the loaded manganese increased the thermal stability of TNT. Manganese oxide was not detected by HRTEM imaging, which could be due to its low content and amorphous nature. Manganese oxide might be dispersed into the pore channels of TNT, which is similar with the literatures.^27,28

Fig. 2 TEM images of (a) TNT, (b) MnO_x/TNT, and HRTEM image of (c) MnO_x/TNT.

MnO_x/TNT was characterized by XPS to confirm the chemical state of Mn and Ti (Fig. 3). The peaks located at 458.1 and 463.9 eV were attribute to Ti⁴⁺ 2p_3/2 and Ti⁴⁺ 2p_1/2, respectively.^29,30 The broad and asymmetric Mn 2p peaks imply multiple oxidation states of Mn in MnO_x/TNT. The peaks around 653.2 and 641.3 eV were assigned to the Mn 2p_1/2 and Mn 2p_3/2 transitions of Mn^2+/3+.³¹ The surface concentration of Na on MnO_x/TNT was much lower than that on TNT, suggesting that Mn ion was loaded on TNT via an ion exchange with Na ion. Then, highly dispersed MnO_x was formed by calcination.

Fig. 3 (a) Ti 2p and (b) Mn 2p spectra of MnO_x/TNT, and (c) Na 1s spectra of TNT and MnO_x/TNT.

The surface area of TNT was calculated to be about 143.6 m² g⁻¹, smaller than that of MnO_x/TNT (171.2 m² g⁻¹). The reason is that some nanotubes of TNT collapsed during calcination. As shown in Fig. 4, both of the isotherms are type IV. The hysteresis loop is type H3, indicating the existence of narrow slit-shaped pores.³²

Fig. 4 Nitrogen adsorption–desorption isotherms of TNT and MnO_x/TNT.

Surface properties of a catalyst, such as surface charge and acidity, are very important for the heterogeneous catalytic ozonation of organic pollutants. Fig. 5 shows the zeta potential as a function of pH for different catalyst suspensions. The PZC of TNT and MnO_x/TNT were measured to be about 1.9 and 1.7, respectively. The introduction of MnO_x did not change the surface charge property of TNT. The surface of the catalysts was negatively charged at pH above 2, and the surface charge density increased significantly with increasing pH.

Fig. 5 Plot of the zeta potential as a function of pH for different catalyst suspensions in the presence of KNO₃ (10⁻³ mol L⁻¹).

The adsorption of base molecules (e.g. pyridine or ammonia) combined with vibrational spectroscopic techniques, is well known for characterizing the surface acid sites.³³ In order to study the interaction between base molecules and catalyst surface in aqueous system, the catalyst was added to an aqueous pyridine or ammonia solution, and the suspension was vigorously stirred for 2 h to achieve adsorption/desorption equilibrium. Then the catalyst was separated from solution by centrifugation and dried in an oven for analysis. The FTIR spectra of the catalysts are shown in Fig. 6. The band at 910 cm⁻¹ corresponded to the stretching vibration of Ti–O bonds.³⁴ The weak band around 1630 cm⁻¹ was assigned to the H–O–H bending vibration of water. For TNT after adsorption of pyridine, the intensity of this band around 1630 cm⁻¹ became stronger, which might be due to the protonation of pyridine molecule by the Brönsted acid sites.³³ According to the literature, the coordination interaction of water molecules with oxide surface is accompanied with protonization, and the strong proton centers on the TNT surface can interact with pyridine molecules with a creation of pyridinium ion (PyH⁺).³⁵ The broad band at 400–600 cm⁻¹ further confirms the adsorption of pyridine on TNT.³⁵ The TNT sample after adsorption of ammonia exhibited similar FTIR spectra of the sample after adsorption of pyridine except for a new band at 1402 cm⁻¹, assigned to the asymmetric deformation mode of NH₄⁺ cations adsorbed on Brönsted acid sites.^36,37 The strong band around 1630 cm⁻¹ could be due to ammonia adsorbed on Lewis acid sites.³⁸ For MnO_x/TNT after adsorption of pyridine and ammonia, the bands at 400–600 and 1630 cm⁻¹ were much weaker and the band at 1402 cm⁻¹ was not observed. The result indicates that TNT exhibited higher adsorption capacities for base molecules and might have more acid sites than MnO_x/TNT.

Fig. 6 FTIR spectra of different catalysts: (a) TNT, (b) TNT after adsorption of ammonia, (c) TNT after adsorption of pyridine, (d) MnO_x/TNT, (e) MnO_x/TNT after adsorption of ammonia, (f) MnO_x/TNT after adsorption of pyridine.

Catalytic ozonation of PNP

Fig. 7a displays the degradation of PNP in single ozonation and catalytic ozonation using various catalysts at pH 6.5. Only 81% of PNP was removed at 45 min in the absence of catalyst. The addition of catalyst enhanced the degradation of PNP. In the presence of TNT, 91% of PNP was removed at 45 min, while with MnO_x/TNT PNP was completely removed at 30 min. The adsorption efficiencies of PNP on TNT and MnO_x/TNT were about 10% and 3%, respectively. The increased removal of PNP in TNT suspension was due to its adsorption on TNT, while the degradation of PNP in MnO_x/TNT suspension was mainly due to the oxidative process in solution. The effect of catalytic ozonation in the presence of MnO_x/TNT is much higher than the combined effect of adsorption on the catalyst surface and ozonation without catalyst, demonstrating that MnO_x/TNT is an effective ozonation catalyst. Catalytic ozonation usually exhibits high effectiveness in the mineralization of organic pollutants. The TOC removal efficiencies during the degradation of PNP are shown in Fig. 7b. About 58% of TOC was removed at 45 min in the presence of ozone alone. The addition of TNT did not enhance the mineralization of PNP, and the TOC removal efficiency was only 52%. However, the TOC removal increased dramatically with the addition of MnO_x/TNT, and about 95% of TOC was removed at a reaction time of 30 min. The maximum concentration of leached Mn in the solution was 0.85 mg L⁻¹. For comparison, the catalytic activity of the dissolved Mn was tested. However, only about 60% of TOC was removed. The contribution of homogeneous catalysis could be ignored. The results indicate that the TNT support is inactive and the supported Mn is the active component for this reaction. In our previous work, various mesoporous composite metal oxides were prepared and used as catalyst for ozonation of PNP.¹⁸ The highest TOC removal efficiency (95%) was achieved by Mn–Co–Fe at a reaction time of 60 min. MnO_x/TNT exhibited much higher activity than Mn–Co–Fe. The stability of MnO_x/TNT was also examined. As shown in Fig. 8, the TOC removal efficiencies were all above 85% in each cycle, and no apparent reduction was observed after six cycles. The result indicates that MnO_x/TNT is a highly effective and stable ozonation catalyst.

Fig. 7 (a) Degradation of PNP and (b) TOC removal during ozonation and catalytic ozonation of PNP at pH 6.5.

Fig. 8 TOC removal for six successive catalytic ozonation experiments with MnO_x/TNT.

Previous studies confirmed that the main ozonation products of nitrophenol are carboxylic acids that could not be oxidized in ozonation.^39,40 The solution pH decreased from 6.5 to 3.3 during the single ozonation process, indicating the accumulation of organic acids in solution. While the solution pH in the presence of MnO_x/TNT changed less (from 6.5 to 4.0) than that in ozonation, which might be due to the removal of the produced organic acids. The organic intermediates generated during the degradation process were analyzed by HPLC and identified by comparison with standards. Formic acid and oxalic acid were detected during the degradation of PNP, and their retention times were 2.8 and 2.9 min, respectively. The concentrations of the acids in catalytic ozonation process were much lower than that in ozonation process. It has been demonstrated that formic acid and oxalic acid are hardly oxidized by ozone molecules and their ozonation rate constants are only 5.0 and <4.0 × 10⁻² M⁻¹ s⁻¹, respectively.⁴¹ Because hydroxyl radicals are highly potent chemical species with a little selectivity, formic acid and oxalic acid can be readily oxidized by hydroxyl radicals with the reaction rate constant of 1.3 × 10⁸ and 1.4 × 10⁶ M s⁻¹, respectively.⁴¹ Fig. 9 presents the catalytic ozonation of formic acid and oxalic acid with MnO_x/TNT at different pH. Both formic acid and oxalic acid were significantly degraded in the catalytic ozonation process, suggesting that the catalyst can enhance the generation of active radicals from ozone decomposition. When the initial pH decreased from 6.5 to 3.5, the TOC removal efficiencies for formic acid and oxalic acid increased from 54% to 92% and 36% to 65%, respectively. This may be associated with the adsorption capacity of the catalyst toward the acids at different pH values. On one hand, the increasing pH promotes the dissociation of the acids into anion species that are prone to be adsorbed on the catalyst surface. On the other hand, the catalyst surface is negatively charged in the tested pH range, and the charge density increases with increasing pH, which inhibits the adsorption of anions. Surface reaction involving adsorbed pollutants is suggested to play an important role in heterogeneous catalytic ozonation process.^10,18 When the initial pH decreased from 6.5 to 3.5, the adsorption efficiencies for formic acid and oxalic acid increased from 10.7% to 23.5% and 5.4% to 9.1%, respectively. The result indicates that the contribution of surface reactions to the removal of formic acid and oxalic acid would be strengthened with decreasing pH, which is similar with the phenomena in literature.⁴²

Fig. 9 Catalytic ozonation of formic acid and oxalic acid with MnO_x/TNT at different pH.

Mechanism discussion

It is well known that coumarin can react with hydroxyl radicals to form fluorescent 7-hydroxycoumarin. Fig. 10a shows the formation of 7-hydroxycoumarin in different ozonation systems. Only a small amount of 7-hydroxycoumarin was formed in ozonation and TNT catalytic ozonation, while MnO_x/TNT resulted in significant 7-hydroxycoumarin generation. The result suggests that the supported MnO_x indeed promoted the decomposition of ozone into ˙OH. In addition, sodium sulfite (Na₂SO₃, 4 mM) was selected as a ˙OH scavenger to investigate its influence on TOC removal in ozonation and catalytic ozonation processes. Although Na₂SO₃ can also react with ozone with a significant rate, this effect can be neglected due to the relatively high concentration of ozone in the reaction medium.^43,44 As shown in Fig. 10b, the addition of Na₂SO₃ had slight effect on the single ozonation of PNP, which indicates that the organic pollutants were mainly degraded by ozone molecules. In contrast, the TOC removal in the presence of MnO_x/TNT was significantly inhibited by adding Na₂SO₃, indicating that ˙OH was the main active species in this catalytic ozonation reaction. Meanwhile, a series of ozone decomposition experiments were carried out. As shown in Fig. 11, about 90% of O₃ was decomposed within 30 min without catalyst, while the ozone decomposition rate was markedly enhanced in the presence of catalysts. The concentration of ozone decreased more rapidly in TNT suspension than MnO_x/TNT suspension before 15 min, and then reached a plateau. In contrast, the concentration of ozone in MnO_x/TNT suspension continuously decreased after 15 min, and ozone was completely decomposed within 25 min. Previous experiments have shown that TNT did not enhance the mineralization of PNP, and the corresponding TOC removal efficiency was even less than that without catalyst (Fig. 7). The reason might be that ozone molecules were only adsorbed on the surface of TNT and were not decomposed into ˙OH. Organic molecules in TNT suspension were mainly degraded by ozone molecules. Characterization results have demonstrated that TNT has more acid sites than MnO_x/TNT and its surface is negatively charged at pH above 2. Since ozone can be regarded as a Lewis base and an electrophilic agent, it can be readily adsorbed on the surface of TNT. On the contrary, the organic intermediates of PNP were hardly adsorbed on TNT. The reactions between ozone and organic molecules were inhibited to some extent, leading to lower reactivity. In comparison with TNT, MnO_x/TNT markedly enhanced both the mineralization of PNP and the decomposition of ozone, indicating that the supported MnO_x can catalytically decompose ozone into ˙OH. It can be deduced that the surface acid sites and negatively charged surface groups of the TNT support are active sites for the adsorption of ozone, while the highly dispersed MnO_x is the active component for the decomposition of ozone into ˙OH.

Fig. 10 (a) Formation of 7-hydroxycoumarin during ozonation and catalytic ozonation of coumarin (20 mg L⁻¹), and (b) effect of sodium sulfite (4 mM) on the ozonation and catalytic ozonation of PNP at pH 6.5.

Fig. 11 Decomposition of ozone in aqueous dispersions of various catalysts at pH 6.5.

Based on the results obtained under different experimental conditions, we proposed a possible mechanism for the mineralization of PNP in this catalytic ozonation reaction. Ozone molecules were firstly adsorbed onto the support, and some of them were transformed into ˙OH by MnO_x highly dispersed on TNT. The TNT support with negatively charged surface and acid sites favored the adsorption of ozone, while the highly dispersed MnO_x with multiple oxidation states enhanced the electron transfer involved in the catalytic decomposition reaction of ozone. The transformation of ozone into ˙OH was significantly accelerated due to the synergic effect between TNT and MnO_x. On the other hand, PNP was oxidized into carboxylic acids by ozone and ˙OH in solution. Subsequently, the produced carboxylic acids were further oxidized into CO₂ by active radicals. Additionally, the solution pH decreased due to the accumulation of organic acids in solution, and some of the produced carboxylic acids were adsorbed on the catalyst, which facilitate their surface reactions with radicals.

Conclusions

MnO_x was highly dispersed on the surface of TNT by a simple impregnation method. The synthesized sample was highly effective for the mineralization of PNP in aqueous solution by ozone. PNP was oxidized by ozone molecules and active radicals in aqueous solution, and its ozonation intermediates were identified to be small molecular carboxylic acids such as formic acid and oxalic acid. The produced acids were further oxidized by hydroxyl radicals. The TNT support with negatively charged surface and acid sites favored the adsorption of ozone, while the highly dispersed MnO_x accelerated the decomposition of ozone into hydroxyl radicals. This study offers a new approach for the optimization of supported catalyst with potential applications in heterogeneous catalytic ozonation.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (no. 21207032), the Natural Science Foundation of Hebei Province of China (no. B2013205028), and the Foundation of Hebei Education Department (no. ZD20131005).

Notes and references

1. J. Nawrocki and B. Kasprzyk-Hordern, Appl. Catal., B, 2010, 99, 27.
2. B. Legube and N. K. V. Leitner, Catal. Today, 1999, 53, 61–72.
3. J. Nawrocki, Appl. Catal., B, 2013, 142–143, 465.
4. L. Ciccotti, L. A. S. do Vale, T. L. R. Hewer and R. S. Freire, Catal. Sci. Technol., 2015, 5, 1143.
5. L. Zhao, Z. Sun, J. Ma and H. Liu, J. Mol. Catal. A: Chem., 2010, 322, 26.
6. R. C. Martins and R. M. Quinta-Ferreira, Appl. Catal., B, 2009, 90, 268.
7. X. Liu, Z. Zhou, G. Jing and J. Fang, Sep. Purif. Technol., 2013, 115, 129.
8. S. Tong, R. Shi, H. Zhang and C. Ma, J. Hazard. Mater., 2011, 185, 162.
9. M. Sui, J. Liu and L. Sheng, Appl. Catal., B, 2011, 106, 195.
10. Q. Sun, L. Li, H. Yan, X. Hong, K. S. Hui and Z. Pan, Chem. Eng. J., 2014, 242, 348.
11. R. Huang, H. Yan, L. Li, D. Deng, Y. Shu and Q. Zhang, Appl. Catal., B, 2011, 106, 264.
12. R. Rosal, M. S. Gonzalo, A. Rodríguez, J. A. Perdigón-Melón and E. García-Calvo, Chem. Eng. J., 2010, 165, 806.
13. S. Xing, C. Hu, J. Qu, H. He and M. Yang, Environ. Sci. Technol., 2008, 42, 3363.
14. E. Illes and E. Tombacz, J. Colloid Interface Sci., 2006, 295, 115.
15. T. Zhang, C. Li, J. Ma, H. Tian and Z. Qiang, Appl. Catal., B, 2008, 82, 131.
16. W. Li, Z. Qiang, T. Zhang, X. Bao and X. Zhao, J. Mol. Catal. A: Chem., 2011, 348, 70.
17. J. S. Park, H. Choi and J. Cho, Water Res., 2004, 38, 2285.
18. Z. Ma, L. Zhu, X. Lu, S. Xing, Y. Wu and Y. Gao, Sep. Purif. Technol., 2014, 133, 357.
19. B. Kasprzyk-Hordern, M. Ziołek and J. Nawrocki, Appl. Catal., B, 2003, 46, 639.
20. J. Bing, C. Hu, Y. Nie, M. Yang and J. Qu, Environ. Sci. Technol., 2015, 49, 1690.
21. J. Lin, A. Kawai and T. Nakajima, Appl. Catal., B, 2002, 39, 157.
22. B. Kasprzyk-Hordern and J. Nawrocki, Ozone: Sci. Eng., 2003, 25, 185.
23. B. Kasprzyk-Hordern, U. Raczyk-Stanisławiak, J. Świetlik and J. Nawrocki, Appl. Catal., B, 2006, 62, 345.
24. P. M. Álvarez, F. J. Beltrán, J. P. Pocostales and F. J. Masa, Appl. Catal., B, 2007, 72, 322.
25. T. Kasuga, M. Hiramatsu, A. Hoson, T. Sekino and K. Niihara, Langmuir, 1998, 14, 3160.
26. A. R. Armstrong, G. Armstrong, J. Canales and P. G. Bruce, Angew. Chem., Int. Ed., 2004, 43, 2286.
27. M. Sui, J. Liu and L. Sheng, Appl. Catal., B, 2011, 106, 195.
28. S. Xing, X. Lu, J. Liu, L. Zhu, Z. Ma and Y. Wu, Chemosphere, 2016, 144, 7.
29. C. Huang, X. Liu, L. Kong, W. Lan, Q. Su and Y. Wang, Appl. Phys. A, 2007, 87, 781.
30. B. Murugan and A. V. Ramaswamy, J. Phys. Chem. C, 2008, 112, 20429.
31. G. S. Qi and R. T. Yang, J. Phys. Chem. B, 2004, 108, 15738.
32. K. S. W. Sing, D. H. Everett, R. A. W. Haul, L. Moscou, R. A. Pierotti, J. Rouquerol and T. Siemieniewska, Pure Appl. Chem., 1985, 57, 603.
33. X. Wang, J. C. Yu, P. Liu, X. Wang, W. Su and X. Fu, J. Photochem. Photobiol., A, 2006, 179, 339.
34. Y. C. Chen, S. L. Lo and J. Kuo, J. Colloid Interface Sci., 2010, 361, 126.
35. T. Bezrodna, G. Puchkovska, V. Shimanovska, I. Chashechnikova, T. Khalyavka and J. Baran, Appl. Surf. Sci., 2003, 214, 222.
36. J. Zawadzki and M. Wiśniewski, Carbon, 2003, 41, 2257.
37. S. M. Park, M. Y. Kim, E. S. Kim, H. S. Han and G. Seo, Appl. Catal., A, 2011, 395, 120.
38. M. Piumetti, M. Armandi, E. Garrone and B. Bonelli, Microporous Mesoporous Mater., 2012, 164, 111.
39. A. Goi, M. Trapido and T. Tuhkanen, Adv. Environ. Res., 2004, 8, 303.
40. B. Liu, S. Li, Y. Zhao, W. Wu, X. Zhang, X. Gu, R. Li and S. Yang, J. Hazard. Mater., 2010, 176, 213.
41. L. Li, W. Zhu, L. Chen, P. Zhang and Z. Chen, J. Photochem. Photobiol., A, 2005, 175, 172.
42. Z. Liu, J. Ma, Y. Cui, L. Zhao and B. Zhang, Sep. Purif. Technol., 2011, 78, 147.
43. Q. Sun, Y. Wang, L. Li, J. Bing, Y. Wang and H. Yan, J. Hazard. Mater., 2015, 286, 276.
44. M. S. Yalfani, S. Contreras, J. Llorca and F. Medina, Appl. Catal., B, 2011, 107, 9.

---