Efficient and stable catalyst of α-FeOOH for NO oxidation from coke oven flue gas by the catalytic decomposition of gaseous H₂O₂

Abstract

Goethite (α-FeOOH) possesses excellent catalytic activity, high selectivity and good stability as a catalyst for NO oxidation through the catalytic decomposition of gaseous H₂O₂. as the primary reactive oxygen species is involved in the NO oxidation process together with ·OH, and N₂O₅ is found for the first time in the products of NO oxidation.

---

Sulphur dioxide (SO₂) and nitrogen oxides (NO_x) yielded by fossil fuels and ore combustion are the major air pollutants produced by traditional chemical industries, including iron and steel, coking, and boilers, and these pollutants cause acid rain, fog, and haze. Wet flue gas desulphurisation (WFGD) and selective catalytic reduction (SCR) are efficient methods used in power plants for desulphurisation and denitration, respectively.¹ However, in traditional chemical industries, for example, coke oven flue gas yielded by coke oven gas combustion has a flue gas temperature of about 200–230 °C and the composition is complex,² which limits the utilisation of SCR.³

Recently, we proposed a promising process of gas-phase oxidation combined with wet scrubbing using steel slag slurry to treat NO_x from low-temperature flue gas.^4,5 In this process, the sparingly soluble NO could be oxidised into soluble NO₂, HNO₃ or N₂O₅ through the gas-phase oxidation process, and then the oxidation products could be absorbed together with SO₂ in a wet scrubbing device. This method can achieve the simultaneous removal of SO₂ and NO_x using the oxidisers. Some oxidisers, such as O₃,^6,7 H₂O₂,^8–12 NaClO₂,¹³ NaClO,¹⁴ persulphate salt (S₂O₈²⁻)¹⁵ and ferrate (Fe[VI]),¹⁶ can be used as the gas-phase oxidiser for NO oxidation after being gasified if needed.

H₂O₂ is a green and low-cost oxidizer, which can be used as the gas-phase oxidizer for NO oxidation after liquid H₂O₂ is gasified at low temperature. Furthermore, the oxidation potential of H₂O₂ (1.77 eV) is lower compared with that of O₃ (2.08 eV) and ·OH (2.80 eV) and thus its efficiency in oxidising NO is low.^12,17 Thermal decomposition of gaseous H₂O₂ can decompose H₂O₂ into radicals (·OH or ) with high oxidation potential for NO oxidation. However, this technology requires high H₂O₂ consumption (H₂O₂/NO = 80) and excessive residence time (34 s) and results in low NO oxidation efficiency (∼60%).¹⁸ Introducing catalysts into the H₂O₂ decomposition process can effectively decompose H₂O₂ into radicals and considerably reduce the consumption of H₂O₂.¹¹ Iron-based materials, such as hematite (α-Fe₂O₃),⁸ nanoscale zero-valent iron,⁹ Fe₃O₄,¹⁰ γ-Fe₂O₃@Fe₃O₄,¹¹ Fe₂(MoO₄)₃ ¹⁹ and Fe₂(SO₄)₃,¹² have been used as catalysts to decompose gaseous H₂O₂ for NO oxidation. These catalysts have high removal efficiencies as heterogeneous catalysts for the simultaneous removal of NO and SO₂ in an integrated catalytic oxidation/wet scrubbing process. However, the use of these catalysts results in relatively high H₂O₂ consumption^8,9,12,20 and relatively low gas hourly space velocity (GHSV)¹⁹ and catalytic stability.^9,12 Therefore, a catalyst with low H₂O₂ consumption and high GHSV and catalytic stability for NO oxidation through catalytic decomposition of gaseous H₂O₂ should be developed.

Goethite (α-FeOOH) is a ubiquitous natural mineral in soils and sediments at the Earth's surface that is widely used as a heterogeneous Fenton catalyst for wastewater treatment due to its abundance, availability, relative stability and low cost.²¹ He et al. explored the catalytic performance of α-FeOOH and found that it can be used to catalyse H₂O₂ vapour for NO oxidation under low-temperature (<160 °C) flue gas;²² however, coke oven flue gas has a higher flue gas temperature (200–230 °C), and the reaction products and catalytic mechanism may be different under different temperature regions. Therefore, the performance of α-FeOOH for NO oxidation through catalysing gaseous H₂O₂ under high flue gas temperature should be investigated, and the SO₂ oxidation efficiency of this process should also be evaluated. Herein, NO and SO₂ conversions and NO₂ yield using α-FeOOH with gaseous H₂O₂ were performed under the wide temperature range of 100–350 °C, and the catalytic stability and reaction mechanism were determined.

Fresh α-FeOOH catalyst was prepared via precipitation reaction of Fe³⁺ followed by crystal transformation according to the method of Böhm.²³ All chemicals used in the experiments were of analytical grade, and deionised water was used. NO, NO₂, and SO₂ concentrations in simulated flue gas were analysed using a UV differential optical absorption spectroscopy (DOAS) flue gas comprehensive analyser (Laoying3023, Qingdao Laoying Environmental Technology Co., Ltd.). The presence of HNO₃ and N₂O₅ in the flue gas were detected using a Fourier transform infrared (FTIR) spectrometer (Tensor 27, Bruker Optik, Inc.), which was equipped with a 2.4 m gas cell. Catalysts were characterised via X-ray diffraction (XRD) spectroscopy (Empyrean, PANalytical Instruments) and FTIR spectrometry. Hydroxyl radical (·OH) and superoxide anion radical decomposed by H₂O₂ were detected via electron paramagnetic resonance (EPR) spectroscopy (A300-10/12, Bruker Optik Inc.), and 5,5-dimethyl-1-pyrroline N-oxide (DMPO) was used for capturing ·OH and . Specifically, DMPO–H₂O solvent was used to capture the ·OH generated after H₂O₂ was catalysed, whereas DMPO–CH₃OH was used to capture The experimental equipment for evaluating the catalytic performance of α-FeOOH was established (Fig. S1, ESI†). In brief, the bypass of N₂ carried the gaseous H₂O₂ generated by the evaporation of H₂O₂ solution and mixed with the simulated flue gas, and then the mixture gas contacted with the catalyst, in which gaseous H₂O₂ was decomposed into radicals over the catalyst and oxidised NO and SO₂. At the outlet of the catalytic reactor, the simulated flue gas after being oxidised was detected using a UV DOAS flue gas comprehensive analyser and an FTIR spectrometer equipped with a gas cell. Each experiment was conducted for 20 min after the temperature was stabilised.

According to Christensen et al., α-FeOOH is transformed to hematite (α-Fe₂O₃) within the temperature range of 171–311 °C, and α-FeOOH is totally converted to α-Fe₂O₃ at 350 °C.²⁴ The fresh catalyst and the catalyst calcinated at 350 °C were characterised via XRD and FTIR spectrometry. The FTIR and XRD spectra (Fig. S2 and S3, ESI†) indicated that the fresh catalyst was α-FeOOH and the catalyst calcinated at 350 °C was α-Fe₂O₃ in an N₂ atmosphere. The colour of the fresh catalyst (yellow) and the calcinated catalyst (red) also proved the above results (Fig. S4, ESI†). An FTIR spectrometer equipped with a gas cell was used to analyse the water vapour and further investigate the transformation temperature of α-FeOOH. Results (Fig. S5, ESI†) showed that α-FeOOH began to decompose into α-Fe₂O₃ and H₂O when the temperature was above 200 °C in an N₂ atmosphere. However, when H₂O(g) and H₂O₂(g) were existed in N₂ atmosphere, they could improve the thermal stability of α-FeOOH. The reason may be that (1) multiple types of surface hydroxyls (–OHs) generated by the adsorption of water on α-FeOOH surface prevented the dehydration of –OHs;^25,26 and (2) H₂O(g) from the injected H₂O₂ solution were adsorbed on the reduced ≡Fe(II) and generated ≡Fe(II)–OH via eqn (1) and (2), and ≡Fe(II)–OH was converted to ≡Fe(III)–OH via eqn (3).^{12,19,22,27,28} The catalyst α-FeOOH was still stable when temperature was up to 225 °C, and little α-FeOOH close to the wall of the reactor was converted to α-Fe₂O₃ (the color of catalyst was changed from yellow to red) when temperature was up to 350 °C. Therefore, the catalyst α-FeOOH possesses great thermal stability under the simulated flue gas condition (e.g., the experimental condition in Fig. 1).

≡Fe(II) + V_o + H₂O → ≡Fe(II)–OH₂⁺ → ≡Fe(II)–OH + H⁺ (2)

≡Fe(II)–OH + H₂O₂ → ≡Fe(III)–OH + ·OH (3)

Fig. 1 Effects of temperature on NO and SO₂ conversions and NO₂ yield (H₂O₂/NO, 2.0; H₂O₂ solution feeding rate, 148.9 μL min⁻¹; catalyst dosage, 0.5 g; GHSV, 137 747 h⁻¹; NO concentration, 200 ppm; SO₂ concentration, 660 ppm; O₂ concentration, 6%).

As shown in Fig. 1, NO conversion and NO₂ yield were both lower than SO₂ conversion when gaseous H₂O₂ only was used for NO oxidation. This result indicated that SO₂ was more easily oxidised by gaseous H₂O₂ than NO, which consumed a large amount of H₂O₂. NO conversion and NO₂ yield were higher, but SO₂ conversion was considerably lower than using gaseous H₂O₂ only when α-FeOOH was used to catalyse gaseous H₂O₂ for NO oxidation. Therefore, α-FeOOH performed excellent catalytic activity and high selectivity for NO oxidation. Specifically, NO conversion achieved 98.8% and NO₂ yield reached 77.0% at H₂O₂/NO of 2.0, reaction temperature of 225 °C, catalyst dosage of 0.5 g and GHSV of 137 747 h⁻¹. The catalyst of α-FeOOH achieved a higher NO oxidation efficiency under low H₂O₂ consumption and high GHSV compared with those reported by previous studies (Table S1, ESI†).^{8,9,12,17,19,20}

Fig. 1 shows that NO conversion was higher than NO₂ yield within the temperature range of 100–225 °C, and NO conversion was closer to NO₂ yield when the temperature further increased from 250 °C to 350 °C. This trend indicated that the products of NO oxidation was not just NO₂ within the temperature range of 100–225 °C, and the product of NO oxidation might only be NO₂ within the temperature range of 250–350 °C. The products of NO oxidation were determined via FTIR spectroscopy. As shown in Fig. 2, the peaks for NO₂ (1599 cm⁻¹), HNO₃ (886 cm⁻¹) and N₂O₅ (1719 cm⁻¹) were observed in the FTIR spectra within the temperature range of 100–200 °C.^29,30 However, no obvious peaks for NO₂, HNO₃, and N₂O₅ were detected in the FTIR spectra when H₂O₂ was not added. This result indicated that NO₂, HNO₃ and N₂O₅ were the main products of NO oxidation through the catalytic decomposition of gaseous H₂O₂ over α-FeOOH in low-temperature (100–200 °C) region. Furthermore, the peaks for HNO₃ and N₂O₅ in the FTIR spectra disappeared when the temperature increased from 225 °C to 350 °C. The reason is that HNO₃ and N₂O₅ decomposed in the high-temperature region (225–350 °C) (eqn (4) and (5)).²⁹ In summary, NO₂, HNO₃ and N₂O₅ were the main products of NO oxidation in the low-temperature region (100–200 °C), whereas NO₂ was the main product in the high-temperature region (225–350 °C).

N₂O₅ → NO + NO₂ + O₂ (4)

2HNO₃ → NO + NO₂ + O₂ + H₂O (5)

Fig. 2 FTIR spectra of simulated flue gas oxidised by gaseous H₂O₂ over α-FeOOH under different temperature conditions. (Experimental condition was the same as that in Fig. 1.)

Fig. 1 shows that NO conversion sharply increased when the temperature increased from 100 °C to 200 °C and then decreased when the temperature further increased from 225 °C to 350 °C. The increase in NO conversion with temperature was because high temperatures enhance H₂O₂ decomposition into radicals according to the Arrhenius law,¹² which further accelerates NO conversion. The decrease in NO conversion at temperatures above 200 °C could be explained by (1) the thermal decomposition of gaseous H₂O₂ under high temperature and/or (2) the transformation of α-FeOOH under high temperature. On the one hand, gaseous H₂O₂ can be decomposed under high temperature. An FTIR spectrometer equipped with a gas cell was used to detect gaseous H₂O₂ under different temperature conditions (free of catalyst) to investigate the decomposition temperature of gaseous H₂O₂,³¹ and the variations in H₂O₂ content were measured by their homologous infrared absorption characteristic peaks (Fig. S6, ESI†). Results (Fig. S7, ESI†) indicated that H₂O₂ almost did not decompose at 100–300 °C, and the thermal decomposition of gaseous H₂O₂ occurred when the temperature increased further to 300 °C. On the other hand, NO-TPD (Fig. S8, ESI†) showed that the absorbed NO began to desorb when the temperature was above 200 °C. Therefore, when the temperature was above 200 °C, NO conversion decreased when the temperature was above 200 °C. Overall, the transformation of α-FeOOH under high temperature (>200 °C) led to the decrease in NO conversion when the temperature was above 200 °C, and the thermal decomposition of gaseous H₂O₂ also resulted in the decrease in NO conversion when the temperature was above 300 °C.

The catalytic stability of α-FeOOH was also investigated. As shown in Fig. 3, NO conversion was maintained at a high level (>97.0%) within the first 15 h and then fluctuated slightly but remained at >90.0% at 15–45 h. SO₂ conversion remained stable at about 2.0%. Results indicated that α-FeOOH possesses excellent catalytic activity, good stability and high selectivity for NO oxidation. According to the results in Fig. 1 and 3, the catalyst has an ideal temperature window of 175–250 °C; therefore, α-FeOOH can be applied in coke oven flue gas (200–230 °C) treatment.

Fig. 3 Catalytic stability of α-FeOOH. (H₂O₂/NO, 2.0; H₂O₂ solution feeding rate, 148.9 μL min⁻¹; catalyst dosage, 0.5 g; GHSV, 137 747 h⁻¹; temperature, 225 °C; NO concentration, 200 ppm; SO₂ concentration, 660 ppm; O₂ concentration, 6%.)

Fresh and used (after the 45 h test) α-FeOOH were characterised via FTIR and XRD spectroscopy. The bands at 3363, 3127 and 1628 cm⁻¹ are the O–H stretching mode in α-FeOOH, the stretching mode of surface water and the bending mode of H₂O, respectively. The characteristic strong bands at 795 and 891 cm⁻¹ were assigned to the Fe–O–H bending vibrations of α-FeOOH. The band at 638 cm⁻¹ was assigned to the Fe–O stretching vibration of pure α-FeOOH. The characteristic absorption peaks of α-FeOOH in the catalyst after the 45 h test did not change compared with that of the fresh catalyst (Fig. S9, ESI†), indicating that the catalyst maintained the structure of α-FeOOH even after the 45 h test. The band at 999 cm⁻¹ was assigned to ν₁(SO₄) frequency, and the bands at 1076, 1136 and 1229 cm⁻¹ were interpreted as ν₃(SO₄) frequencies. These vibrational frequencies are attributed to specifically adsorbed SO₄²⁻ ions on the external and internal surfaces of catalyst particles after the 45 h test.^32,33 The XRD patterns showed that the phase of the catalyst before and after the stability test was still α-FeOOH (Fig. S10, ESI†). This result also indicated that α-FeOOH possessed good stability in the NO oxidation process.

EPR test was conducted to detect the radicals generated by H₂O₂ decomposition and analyse the oxygen species of the α-FeOOH/H₂O₂ system. Fig. 4 shows the 4-fold characteristic peak of DMPO–·OH adducts with an intensity ratio of 1 : 2 : 2 : 1 and four notable signals ascribed to adducts. These peaks and signals indicate the formation of ·OH and in the α-FeOOH/H₂O₂ system.^21,34 Furthermore, the intensity of the 4-fold characteristic peak of DMPO–·OH adducts was significantly weaker than that of adducts, indicated that the quantity of produced by the α-FeOOH/H₂O₂ system was larger than that of ·OH. Therefore, α-FeOOH can decompose H₂O₂ into ·OH and and may be dominant among ·OH and for NO oxidation. Radical trapping experiments (Fig. S11, ESI†) further proved the roles of radicals (·OH and ) in the NO oxidation process. Benzoquinone (BQ, scavenger²⁰) was added into the H₂O₂ solutions, and H₂O₂ and BQ in the mixture solution was vapoured together. NO conversion substantially decreased from 92.9% to 6.1% when the molar ratio of BQ to H₂O₂ increased from 0 to 1.0. The reason was that the addition of BQ captured the generated by gaseous H₂O₂ decomposition over α-FeOOH, thereby decreasing NO conversion. NO conversion decreased from 92.9% to 8.0% when isopropanol (i-PrOH, ·OH scavenger²⁰) was introduced into the same system as the molar ratio of i-PrOH to H₂O₂ increased from 0 to 8.0. The reason was that ·OH generated in the α-FeOOH/H₂O₂ system was captured by i-PrOH instead of oxidised NO. These results indicated that when the addition concentration of the ·OH scavenger (i-PrOH) was eight times that of scavenger (BQ), both systems obtained the similar decrease in NO conversion. Therefore, as the primary reactive oxygen species was involved in the NO oxidation process together with ·OH.

Fig. 4 EPR spectra of the α-FeOOH/H₂O₂ system.

In this catalytic process, the reaction between H₂O₂ as an oxidant and iron ions as a catalyst to produce highly active species (·OH and ).³⁵ The conversion between Fe²⁺ and Fe³⁺ was proceed according to the Haber–Weiss mechanism. ≡Fe(III)–OH and ≡Fe(III) are reduced by H₂O₂ and generate ≡Fe(II)–OH and ≡Fe(II) (eqn (6) and (1)), the resulting ≡Fe(II)–OH and ≡Fe(II) also can be oxidized by H₂O₂ and produce ≡Fe(III)–OH and ≡Fe(III) (eqn (3) and (7)).^{12,19,22,27,28,36}

≡Fe(II) + H₂O₂ → ≡Fe(II) + ·OH + OH⁻ (7)

The proposed mechanism of NO oxidation by H₂O₂ decomposition over α-FeOOH is presented in Fig. 5 based on the study of the products of NO oxidation and reactive oxygen species ( and ·OH) in the NO oxidation process. (1) Gaseous H₂O₂ decomposed into ·OH and on the active sites (≡Fe(II)–OH and ≡Fe(II), ≡Fe(III)–OH and ≡Fe(III)) of the catalyst (eqn (1)–(3), (6) and (7)) according to the Haber–Weiss mechanism, a result that was proved through EPR analysis and scavenger experiments. (2) NO was oxidised by the generated ·OH and and produced NO₂ and HNO₃ via eqn (8) and (9). (3) The produced HNO₃ reacted with ·OH and produced NO₃ or NO₂ (eqn (10) and (11)). (4) NO₂ reacted with NO₃ and produced N₂O₅ via eqn (12).³⁷ The production of NO₂, HNO₃ and N₂O₅ during the NO oxidation process was proved by the FTIR spectra (Fig. 2).

NO + 2·OH → NO₂ + H₂O (8)

HNO₃ + ·OH → NO₃ + H₂O (10)

HNO₃ + ·OH → NO₂ + H₂O₂ (11)

NO₂ + NO₃ → N₂O₅ (12)

Fig. 5 Catalytic mechanism of NO oxidation by H₂O₂ decomposition over α-FeOOH (≡Fe^III:≡Fe(III), ≡Fe(III)–OH; ≡Fe^II:≡Fe(II), ≡Fe(II)–OH).

In summary, α-FeOOH can be used as an efficient catalyst for coke oven flue gas to enhance NO oxidation efficiency through the catalytic decomposition of gaseous H₂O₂. Moreover, α-FeOOH showed high selectivity for NO oxidation in the presence of SO₂. In this study, NO conversion achieved 98.8% under the following experimental conditions: H₂O₂/NO of 2.0, reaction temperature of 225 °C, catalyst dosage of 0.5 g and GHSV of 137 747 h⁻¹. The 45 h test indicated that α-FeOOH has good catalytic stability. The EPR test and radical trapping experiments revealed that as the primary reactive oxygen species was involved in the NO oxidation process together with ·OH. Furthermore, NO₂, HNO₃ and N₂O₅ were the products of NO oxidation through the catalysis of gaseous H₂O₂ over α-FeOOH within the temperature range of 100–200 °C, and NO₂ was the only oxidation product within the temperature range of 225–350 °C.

The catalyst α-FeOOH can be used to decompose gaseous H₂O₂ into radicals and oxidise NO into high-valence NO_x, and then the oxidation products can be absorbed together with SO₂ in existing industrial-scale WFGD systems.^4,5 This process can achieve the simultaneous removal of SO₂ and NO_x from coke oven flue gas.

Conflicts of interest

There are no conflicts of interest to declare.

Acknowledgements

This work was financially supported by the NSFC (51574214) and the Young Scientists Fund of the NSFC (21706264).

Notes and references

1. A. J. Sweeney and Y. A. Liu, Ind. Eng. Chem. Res., 2001, 40, 2618–2627.
2. H. Yin, W. Lü, G. Sun, G. Sun and B. Hong, Fuel & Chemical Processes, 2015, 46, 1–4.
3. Q. Ma, Z. Wang, F. Lin, M. Kuang, R. Whiddon, Y. He and J. Liu, Energy Fuels, 2016, 30, 2302–2308.
4. Z. Meng, C. Wang, X. Wang, Y. Chen and H. Li, Energy Fuels, 2018, 32, 2028–2036.
5. Z. Meng, C. Wang, X. Wang, Y. Chen, W. Wu and H. Li, Fuel, 2019, 255, 115760.
6. J. Zhang, R. Zhang, X. Chen, M. Tong, W. Kang, S. Guo, Y. Zhou and J. Lu, Ind. Eng. Chem. Res., 2014, 53, 6450–6456.
7. S. Zhou, J. Zhou, Y. Feng and Y. Zhu, Ind. Eng. Chem. Res., 2016, 55, 5825–5831.
8. B. Wu, Y. Xiong, J. Ru and H. Feng, Korean J. Chem. Eng., 2016, 33, 3407–3416.
9. Y. Zhao, B. Yuan, R. Hao and Z. Tao, Energy Fuels, 2017, 31, 7282–7289.
10. S. Yang, Y. Xiong, Y. Ge and S. Zhang, Mater. Lett., 2018, 218, 257–261.
11. S. Yang, Y. Xiong and S. He, ChemistrySelect, 2019, 4, 8829–8836.
12. B. Wu, Y. Xiong and Y. Ge, Chem. Eng. J., 2018, 331, 343–354.
13. Y. Zhao, R. Hao, B. Yuan and J. Jiang, J. Hazard. Mater., 2016, 301, 74–83.
14. Y. Zhao, R. Hao, F. Xue and Y. Feng, J. Hazard. Mater., 2017, 321, 500–508.
15. R. Hao, X. Wang, X. Zhao, M. Xu, Y. Zhao, X. Mao, B. Yuan, Y. Zhang and K. Gao, Chem. Eng. J., 2018, 333, 583–593.
16. Y. Zhao, Y. Han, T. Ma and T. Guo, Environ. Sci. Technol., 2011, 45, 4060–4065.
17. B. Wu and Y. Xiong, J. Chem. Technol. Biotechnol., 2018, 93, 43–53.
18. K. Kou, W. Zhou, Y. Wang, H. Zhao and J. Gao, Can. J. Chem. Eng., 2019, 97, 2419–2425.
19. X. Liu, C. Wang, T. Zhu, Q. Lv, Y. Li and D. Che, Chem. Eng. J., 2019, 371, 486–499.
20. B. Yang, S. Ma, R. Cui, S. Sun, J. Wang and S. Li, Chem. Eng. J., 2019, 359, 233–243.
21. H. Jin, X. Tian, Y. Nie, Z. Zhou, C. Yang, Y. Li and L. Lu, Environ. Sci. Technol., 2017, 51, 12699–12706.
22. S. He, Y. Xiong, S. Yang and Y. Ge, Chem. Ind. Eng. Prog., 1–12,  DOI:10.16085/j.issn.1000-6613.2019-0630.
23. J. Böhm, Z. Anorg. Allg. Chem., 1925, 149, 203–216.
24. A. N. Christensen, T. R. Jensen, C. R. H. Bahl and E. DiMasi, J. Solid State Chem., 2007, 180, 1431–1435.
25. G. Bbuscå and P. F. Rossi, Mater. Chem. Phys., 1983, 9, 561–570.
26. J. Jin, Master, Zhengzhou University, 2019.
27. J. Ding, Q. Zhong, S. L. Zhang, F. J. Song and Y. F. Bu, Chem. Eng. J., 2014, 243, 176–182.
28. J. Ding, Q. Zhong and S. L. Zhang, J. Mol. Catal. A: Chem., 2014, 393, 222–231.
29. C. L. Sun, N. Zhao, Z. K. Zhuang, H. Q. Wang, Y. Liu, X. L. Weng and Z. B. Wu, J. Hazard. Mater., 2014, 274, 376–383.
30. H. J. Yoon, H.-W. Park and D.-W. Park, Energy Fuels, 2016, 30, 3289–3297.
31. M. Pettersson, S. Tuominen and M. Räsänen, J. Phys. Chem. A, 1997, 101, 1166–1171.
32. S. Musić, Z. Orehovec, S. Popović and I. Czakó-Nagy, J. Mater. Sci., 1994, 29, 1991–1998.
33. K. Kandori, T. Shigetomi and T. Ishikawa, Colloids Surf., A, 2004, 232, 19–28.
34. D. Huang, J. Li, G. Zeng, W. Xue, S. Chen, Z. Li, R. Deng, Y. Yang and M. Cheng, Chem. Eng. J., 2019, 375, 121991.
35. S. R. Pouran, A. A. A. Raman and W. M. A. W. Daud, J. Cleaner Prod., 2013, 64, 1–12.
36. L. Guo, Q. Zhong, J. Ding, M. Ou, Z. Lv and F. Song, Ozone: Sci. Eng., 2016, 38, 382–394.
37. Z. Wen, PhD, Zhejiang University, 2009.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/d0ra00533a

---