Photocatalytic degradation of humic acid using a novel photocatalyst: Ce-doped ZnO

Abstract

This study aimed at investigating the photocatalytic degradation of humic acid (HA) as a representative of natural organic matter (NOM) by using Ce-doped ZnO as a novel material. Following photocatalysis, HA degradation was characterized by specified UV-vis and fluorescence spectroscopic parameters as well as by the dissolved organic carbon (NPOC) content. Excitation–Emission Matrix (EEM) fluorescence features were also evaluated by using advanced techniques. Comparison of Ce-doped ZnO photocatalysis to TiO₂ P-25 photocatalytic treatment of the HA samples was elucidated under similar experimental conditions. Kinetic modeling of the photocatalytic removal of HA expressed promising results indicating that Ce-doped ZnO could serve as an efficient catalyst for the degradation of NOM.

---

Introduction

A big effort has been done to orient scientific research towards the exploitation of solar light in various ways. This has occurred also within the catalytic community, which is currently more and more oriented towards studies directed to exploit light energy in various kinds of applications.¹ Among photochemical applications in catalysis, one has to mention quite mature approaches like photocatalytic reactions for pollutant abatement.² The basic component of a photocatalytic system is a semiconductor capable of generating a phenomenon of charge separation by absorbing light. The electron (excited in the conduction band, CB) and the hole (consequently formed in the valence band, VB) are potentially the agents of a reduction and of an oxidation reaction respectively.³ To achieve high performances of the whole photocatalytic system it is necessary to develop semiconductors having excellent electronic properties and, in parallel, co-catalysts having high efficiency. Recent interest has been directed to the use of novel photocatalysts as well as to the development of third generation photocatalysts (TGP) that are active both in the UV and in the visible light region. The third generation of photocatalysts tries to go beyond titanium dioxide and was initially a prediction proposed by Serpone and Emeline.^4,5 This was based on the idea of a wide band gap semiconductor containing extra electronic levels at intermediate energy in the band gap capable of allowing the transition of electrons from the VB to the CB with a double excitation. This was based on the idea that a semiconductor containing extra electronic levels in the wide band gap could allow transition of electrons from VB to CB with a double excitation. These extra electronic levels can be obtained at the interfaces formed between crystallites of different oxides during the synthesis process.⁶ Though uncommon, the presence of rare earth ions in photocatalytic systems is not totally new. In recent years, for instance, Zaleska and coworkers have reported the photocatalytic properties of titanium dioxide doped with various rare earth ions.⁷ Very recently in the literature a growing number of papers based on doped ZnO with lanthanides and in particular with cerium, has occurred. In most of these cases Ce-doped ZnO showed interesting properties in abating organic pollutants.^8–14 Calza and colleagues focused on obtaining different materials (i.e. ZnO doped with cerium) by changing both the precursors of the bare oxide and the synthesis process. The efficiency of the thus obtained photocatalysts was tested in the degradation of acesulfame K which is a persistent emerging contaminant under various conditions e.g. river water expressing the organic content as 4.8 mg L⁻¹.¹⁵

The presence of natural organic matter (NOM) in natural waters deserves prime importance to be investigated for maintaining safe drinking water. It is very well documented that TiO₂ photocatalysis has been applied extensively for the degradation of NOM and NOM analogs as humic acids (HAs), and fulvic acids for decades.^16–19 Application of the second generation of photocatalysts for the degradation of natural organic matter has received recent interest.²⁰ Besides TiO₂ photocatalysis, Oskoei and colleagues reported efficient removal of HA by ZnO nano-photocatalysis.²¹ Complementary to the reported findings, the role of TGP would be of interest to the researchers working in the field of photocatalysis primarily for effective removal of NOM and for elucidation of the role of NOM in the degradation of persistent pollutants. In this respect, the major aim of this work is devoted to use Ce-doped ZnO as a TGP for the photocatalytic treatment of humic acids selected as the model compound of NOM. Moreover, due to the polydispersity character of NOM, further interest was directed to the understanding of the molecular size fractionation of the humic matter displaying similar organic carbon contents. It should also be indicated that activity testing of a novel photocatalyst should also be extended to higher molecular weight recalcitrant organics for assessment of the real performance.²² As indicated above, TiO₂ photocatalytic treatment of NOM had been systematically investigated by Bekbolet and colleagues, therefore TiO₂ was also used for comparison purposes.

Results and discussion

Characterization of the prepared materials

Physico-chemical properties of the prepared Ce-doped ZnO and of the TiO₂ P-25 (used for comparison purposes) are presented in Table 1. Ce-doped ZnO displayed a mixed phase composed of CeO₂ (1%) and ZnO (99%). The surface area was considerably smaller (<10 m² g⁻¹) than TiO₂, (55 m² g⁻¹).

Table 1 Physico-chemical properties of Ce-doped ZnO and TiO₂

Sample	Phase	E bg ^a, eV	mol, %	Surface area m^2 g^−1	d , nm
a Band gap. b Crystallite size. c Anatase 70% and rutile 30%.
Ce-doped ZnO	ZnO/CeO2	3.26	99/1	<10	299/9
TiO2	Mixed^c	3.2	100	55 ± 15	22.3

---

HA solutions i.e. 0.45 μm ff HA and 100 kDa HA displayed almost similar organic carbon contents of NPOC_i: 4.80 mg L⁻¹ and 4.67 mg L⁻¹ respectively that could be considered as resembling the natural water organic matter content investigated by Calza and colleagues.¹⁵ Molecular size distribution (fractions passing through 100 kDa, 30 kDa, 10 kDa and 3 kDa filters) profiles considered as NPOC expressed variations with respect to the prepared HA samples. The 0.45 μm ff HA sample contained the organic content with a molecular size greater than 100 kDa of 65.6%. Further fractionation displayed the 53.1% 30 kDa fraction, 18.1% 10 kDa fraction and 8.8% 3 kDa fraction. On the other hand, the 100 kDa HA sample displayed the 62.0% 30 kDa fraction, 31.7% 10 kDa fraction and 6.3% 3 kDa fraction. The UV-vis and fluorescence spectroscopic properties of the samples as well as NPOC and SUVA were also presented (Table 2).

Table 2 UV-vis and fluorescence spectroscopic properties, NPOC and SUVA of 0.45 μm ff HA and 100 kDa HA samples

	Color436	UV365	UV280	UV254	NPOC^a	SUVA^b	FI
a NPOC: mg L^−1. b SUVA: L m^−1 mg^−1.
0.45 μm ff HA
0.45 μm	0.079	0.162	0.379	0.454	4.56	9.98	1.09
100 kDa	0.025	0.051	0.129	0.156	1.41	11.1	1.25
30 kDa	0.017	0.038	0.102	0.125	1.32	9.47	1.26
10 kDa	0.003	0.007	0.025	0.031	1.08	2.87	1.22
3 kDa	0.001	0.003	0.011	0.015	0.913	1.65	1.53
100 kDa HA
100 kDa	0.066	0.146	0.374	0.445	4.67	9.53	1.17
30 kDa	0.035	0.085	0.236	0.287	2.97	9.66	1.21
10 kDa	0.015	0.038	0.119	0.147	2.28	6.45	1.27
3 kDa	0.004	0.008	0.023	0.029	1.14	2.54	1.44

---

The UV-vis spectroscopic parameters as well NPOC contents decreased in accordance with the decreasing molecular size for both of the HA samples. SUVA > 4 indicating aromaticity could be visualized for molecular size fractions greater than the 30 kDa fraction of the 0.45 μm ff HA sample and the 10 kDa fraction of the 100 kDa HA sample. EEM contour plots displayed the presence of humic-like and fulvic-like fluorophores in accordance with the SUVA values (Fig. 1a and b).

Fig. 1 EEM fluorescence contour plots of molecular size fractions (30 kDa, 10 kDa and 3 kDa) of 0.45 μm ff HA (a) and 100 kDa HA (b) samples.

Preliminary experiments

Preliminary experiments were conducted under conditions representing, i. adsorptive interactions: the absence of light and the presence of a photocatalyst, and ii. photolytic reactions: the absence of a photocatalyst and the presence of light.

Pre-adsorption. Adsorptive interactions were assessed by time dependent NPOC variations in the absence of light and the presence of Ce-doped ZnO. The photocatalyst dose dependent initial adsorption pattern of 0.45 μm ff HA was followed in the range of 0.1–1.0 mg mL⁻¹ Ce-doped ZnO for reaction periods up to 60 min with 10 min intervals. Under all conditions <15% NPOC adsorption was attained. The adsorption extent of 0.45 μm ff HA onto 0.25 mg mL⁻¹ Ce-doped ZnO was <10% NPOC for a period of 10 min and remained unchanged throughout the extent of the reaction period. 100 kDa HA displayed slightly higher initial adsorption of 14%. Since the basic mechanism of photocatalysis relies on pre-adsorption of the substrate onto the photocatalyst, the adsorbed amount of HA onto the photocatalyst fulfills the initial requirement of surface coverage. Jiang and colleagues reported that a wide array of NOM and NOM analogs displayed very small adsorption onto ZnO particles.²³ On the other hand, initial adsorptions of 0.45 μm ff HA and 100 kDa HA onto TiO₂ (0.25 mg mL⁻¹) were 49% and 46% respectively. The results were found to be in good agreement with the surface area properties of the photocatalysts (Table 1).

Photodegradation. Neither of the HA solutions displayed any significant NPOC degradation (<5%) under exposure to light in the absence of a photocatalyst as expected and reported previously.²⁰

Photocatalytic degradation experiments and kinetics

Ce-doped ZnO photocatalytic degradation of HA samples was followed with respect to irradiation time (t) by using humic spectroscopic parameters and the mineralization extent was followed by NPOC. All of the parameters displayed an irradiation time dependent logarithmic decreasing profile irrespective of the photocatalyst type. Therefore, with respect to the prevailing degradation mechanism based on the non-selective oxidation via reactive oxygen species, a pseudo first order kinetic model was successfully applied to the degradation data expressed by all humic parameters (Table 3).

Table 3 Pseudo first order kinetic constants

Photocatalyst	Rate constant k, min^−1				NPOC
	UV-vis spectroscopic parameters
	Color436	UV365	UV280	UV254	k × 10^−2, min^−1	Rate, mg L^−1 min^−1
0.45 μm ff HA, NPOCi: 4.80 mg L^−1
Ce-doped ZnO	5.06 × 10^−2	6.18 × 10^−2	6.54 × 10^−2	5.79 × 10^−2	1.80	0.0864
TiO2	0.123	0.126	9.22 × 10^−2	9.09 × 10^−2	3.87	0.186
100 kDa HA, NPOCi: 4.67 mg L^−1
Ce-doped ZnO	6.76 × 10^−2	8.47 × 10^−2	9.18 × 10^−2	8.17 × 10^−2	2.18	0.102
TiO2	0.166	0.174	0.190	0.193	4.74	0.221

---

Photocatalytic degradation rates expressed as the reaction rate constants of 0.45 μm ff HA using Ce-doped ZnO displayed a decreasing sequence as UV₂₈₀ > UV₃₆₅ > UV₂₅₄ > Color₄₃₆. Decolorization of the humic fractions was mainly due to the removal of chromophoric groups composed of conjugated π systems and centers possessing a lone pair of electrons. On the other hand, the highly condensed core of humic acid fractions displayed a comparatively higher degradation rate constant due to the high aromatic character as expressed by the rate constants attained for UV₂₈₀ and UV₂₅₄. Although a similar trend of removal for UV-vis parameters was observed for 100 kDa HA with respect to 0.45 μm ff HA, considerably higher first order rate constants were attained. The reason could be attributed to the more uniform molecular size distribution profile of 100 kDa HA in comparison to 0.45 μm ff HA although both HA samples contained almost equal NPOC contents as presented previously (Table 2).

Photocatalytic degradation extents expressed as the reaction rate constants of 0.45 μm ff HA using TiO₂ displayed a different trend in a decreasing sequence as UV₃₆₅ > Color₄₃₆ > UV₂₈₀ > UV₂₅₄ with respect to the trend that was attained previously.^20,24 On the other hand, 100 kDa HA expressed an increasing trend as Color₄₃₆ < UV₃₆₅ < UV₂₈₀ < UV₂₅₄. The prevailing mechanism could be visualized by electrostatic attractions between negatively charged deprotonated humic functional groups and positively charged photocatalyst surface groups as opposed by the repulsive interactions. Moreover, the denser aromatic rich inner core displayed considerably higher removal tendency through hydrophobic interactions. Therefore, it could be deduced that electrostatic attractions were suppressed over hydrophobic attractions as well as the van der Waals forces.

Photocatalytic mineralization (NPOC) extents could be expressed in terms of first order rates (R, mg L⁻¹ min) of 0.0864 and 0.102 for 0.45 μm ff HA and 100 kDa HA respectively. On the other hand, the rates (R, mg L⁻¹ min) of 0.186 and 0.221 were attained for TiO₂ photocatalytic degradation of 0.45 μm ff HA and 100 kDa HA respectively. Ce-doped ZnO photocatalytic treatment of 0.45 μm ff HA, expressed half-life values in the range of 11–14 min for the UV-vis spectroscopic parameters, whereas the NPOC half-life was 39 min. In a similar manner, Ce-doped ZnO photocatalytic degradation of 100 kDa HA expressed half-life values for the UV-vis spectroscopic parameters in a range of 7–10 min whereas the NPOC half-life was 32 min. On the other hand, in accordance with the observed higher degradation rate constants, for the TiO₂ photocatalytic degradation of 0.45 μm ff HA, half-life values for the UV-vis spectroscopic parameters were found to be in the range of 6–8 min whereas the NPOC half-life was 18 min. The TiO₂ photocatalytic degradation of 100 kDa HA expressed half-life values for the UV-vis spectroscopic parameters as 3–4 min whereas the NPOC half-life was 15 min. NPOC removal rates were found to be in good agreement with the initial adsorption conditions of both of the photocatalysts onto 0.45 μm ff HA and 100 kDa HA respectively.

Based on the mineralization extent in comparison to UV₂₅₄ removals, the SUVA (L m⁻¹ mg⁻¹) changes were found to be decreasing from 9.50 to 1.43 for Ce-doped ZnO and 8.47 to 1.34 for TiO₂ representing significant elimination of aromaticity. Since SUVA > 4 indicates aromaticity and the presence of hydrophobic moieties, Ce-doped ZnO was found to be more effective on the removal of the aromatic core possibly occurring through a degradation pathway by the semi-selective action of reactive oxygen species. Further evaluation could be related to the diverse stability and aggregation properties of ZnO nanoparticles depending on the solution matrix conditions.²⁵ On the other hand, the stabilizing effect of NOM significantly affected the stabilization and aggregation of TiO₂ particles.²⁶ Due to the differences in the primary particle sizes of the photocatalysts, through electrostatic attractions the role of the humic sub-fractions could be related to the disturbance of colloidal stability leading to possible aggregation conditions most probably excluding sweep co-precipitation through surface adsorption (Table 1). The resulting non-homogeneous reaction medium could possibly lead to the diminished light harvesting capacity of Ce-doped ZnO. The overall effect could be visualized by slower reaction rates in comparison to TiO₂.

Fluorescence spectroscopic elucidation of the photocatalytic degradation of HA

EEM fluorescence contour plots were presented for the photocatalytic degradation of 0.45 μm ff HA and 100 kDa HA using Ce-doped ZnO (Fig. 2a and b respectively).

Fig. 2 Time dependent EEM fluorescence contour plots of 0.45 μm ff HA (a) and 100 kDa HA (b) upon photocatalytic treatment by using Ce-doped ZnO.

Following the initial adsorption (t = 0) of 0.45 μm ff HA onto Ce-doped ZnO, the presence of humic-like and fulvic-like fluorophores was still evident up to an irradiation period of 20 min. Upon further irradiation conditions, the EEM contour plots were found to be devoid of any fluorescence signatures. In a similar manner, following the initial adsorption (t = 0) of 100 kDa HA onto Ce-doped ZnO, the presence of humic-like and fulvic-like fluorophores was also still evident up to an irradiation period of 20 min. However, the presence of fulvic-like fluorophores displayed considerably lower fluorescence intensities in comparison to 0.45 μm ff HA. Upon further irradiation conditions, the EEM contour plots were also found to be devoid of any fluorescence signatures.

EEM fluorescence contour plots were presented for the photocatalytic degradation of 0.45 μm ff HA and 100 kDa HA using TiO₂ (Fig. 3a and b respectively).

Fig. 3 Time dependent EEM fluorescence contour plots of 0.45 μm ff HA (a) and 100 kDa HA (b) upon photocatalytic treatment by using TIO₂.

Following the initial adsorption (t = 0) of 0.45 μm ff HA onto TiO₂, the presence of humic-like and fulvic-like fluorophores was still obvious up to an irradiation period of 20 min. Upon further irradiation conditions, the EEM contour plots were devoid of any fluorescence signatures. Following the initial adsorption (t = 0) of 100 kDa HA onto TiO₂, humic-like and fulvic-like fluorophores were still present up to an irradiation period of 20 min. However, humic-like fluorophores displayed considerably higher fluorescence intensities in comparison to 0.45 μm ff HA. Upon further irradiation conditions (t = 30 min) the EEM contour plots were also devoid of any fluorescence signatures.

EEM contour plots of the treated HA samples displayed close similarities to the 3 kDa fraction of untreated HA being more apparent for 100 kDa HA upon 30 minutes of Ce-doped ZnO photocatalytic treatment. On the other hand, a similar result could be addressed upon TiO₂ photocatalysis for 20 minutes. These results significantly demonstrated the transformation of higher molecular weight fractions to lower molecular weight fractions via successful mineralization of the total organic carbon contents irrespective of the starting humic material composition.

Since Ce doped ZnO exhibited buffer properties at pH 7.4–7.6, dissolution of the photocatalyst was prevented. However, in the presence of HA, through chelation by ligand–metal attractions, leaching of metal ions could be expected. It was also reported by Jiang and colleagues that in the presence of HA, ZnO dissolution kinetics were positively correlated with SUVA and the molecular weight of the studied NOM and NOM analogs.²³ However, neither Ce nor Zn leaching in considerable amounts could be determined under the specified experimental conditions. On the other hand, the possible fluorescence quenching conditions could be attributed to the nonexistence of humic-like and fulvic-like fluorophores rather than achievement of complete mineralization.

Experimental

Preparation and characterization of Ce-doped ZnO photocatalysts

All reactants were purchased from Aldrich and have purity higher than 99.9%. All reactants were used without any further purification treatment. The bare ZnO sample was synthesized starting from a 1 M water solution of Zn(NO₃)₂·6H₂O. Then a 4 M NaOH solution was added dropwise until the pH was 10–11, and finally the solution was transferred into a Teflon lined stainless steel 100 mL autoclave (filling 70%), and then treated at 175 °C for 16 hours. The Ce-doped ZnO sample (Ce molar concentration 1%) was prepared by adding a stoichiometric amount of CeCl₃·7H₂O in the starting solution, and then the same procedure was followed. The sample was labelled as Ce-doped ZnO. Detailed information regarding the properties of the thus prepared Ce-doped ZnO is presented in comparison to TiO₂ P-25 (Table 1).

Powder X-ray diffraction (XRD) patterns were recorded with a PANalytical PW3040/60 X'Pert PRO MPD using a copper K_α radiation source (0.15418 nm). The intensities were obtained in the 2θ range between 20° and 70°. The X'Pert High-Score software was used for data handling. Rietveld refinement was performed on the diffraction patterns to determine the crystallite size and relative abundance of phases, using the MAUD 2.26 software and a NIST Si powder to determine the instrumental function.²⁷ The UV-Vis absorption spectra were recorded using a Varian Cary 5000 spectrometer, coupled with an integration sphere for diffuse reflectance studies (DRS), using Carywin-UV/scan software. A sample of PTFE with 100% reflectance was used as the reference. Surface area measurements were carried out on a Micromeritics ASAP 2020/2010 using the Brunauer–Emmett–Teller (B.E.T.) model on the N₂ adsorption measurement. Prior to the adsorption run, all the samples were outgassed at 160 °C for 3 hours.

Preparation and characterization of HA samples

HA (20 mg L⁻¹) was prepared upon dilution of the stock solution (1.0 g L⁻¹) and filtered through a 0.45 μm membrane filter (designated as 0.45 μm ff HA). HA solution (50 mg L⁻¹) was subjected to ultrafiltration through a 100 kDa membrane filter and thus the obtained sample was designated as 100 kDa HA. The molecular size distribution of the prepared HA solutions was performed by a stirred cell ultrafiltration system applied in sequential mode using membrane filters with nominal molecular weight cut-offs of 100 kDa, 30 kDa, 10 kDa and 3 kDa. The NPOC mass balance was 95%. The reproducibility of molecular size fractionation experiments was reported previously.²⁴ Millipore MilliQ water, resistivity of 18.2 MΩ cm at 25 °C was used in the preparation of all solutions.

Following photocatalytic treatment, all samples were immediately filtered through 0.45 μm membrane filters to remove the photocatalyst from the reaction medium. Thus, the prepared clear solutions were subjected to further analysis. NPOC was determined by using a Shimadzu TOC-VWP Total Organic Carbon Analyzer. UV-vis and fluorescence spectroscopic measurements were performed by using a Perkin Elmer lambda 35 UV-vis Spectrophotometer and a Perkin Elmer LS 55 Luminescence Spectrometer respectively.

Humic acid quantification and related parameters. The humic acid organic matter content was expressed by NPOC (mg L⁻¹) and the specified UV-vis spectroscopic parameters (m⁻¹) were Color₄₃₆, UV₃₆₅, UV₂₈₀ and UV₂₅₄.²⁴ The specific UV parameter i.e. SUVA as NPOC normalized UV₂₅₄ (L m⁻¹ mg⁻¹) was also calculated.²⁸ EEM fluorescence contour plots and the fluorescence Index were derived from the fluorescence data.²⁹ The EEM contour plots were evaluated according to the regional description of Coble.³⁰ Leaching of Ce was determined by using an inductively coupled plasma optical emission spectrometer (ICP-OES), Perkin-Elmer Optima 2100DV.³¹ Zn leaching was determined by using (Perkin-Elmer AAnalyst 300) atomic absorption spectroscopy.

Photocatalytic degradation experiments

Photocatalytic degradation experiments were performed by using an Atlas Suntest CPS+ simulator equipped with an air cooled Xenon Lamp (wavelength range: 300–800 nm and intensity, I: 250 W m⁻²). Irradiation periods were 0–60 min with 10 min intervals. Related information on the reaction conditions was reported in detail by Birben and colleagues.²⁰ The photocatalyst dose was 0.25 mg mL⁻¹ for Ce-doped ZnO as well as TiO₂. Furthermore, ZnO was also used as a photocatalyst and the attained results are presented in S1 ESI.†

Conclusions

Based on the presented data, photocatalytic degradation kinetics indicated the possible use of Ce-doped ZnO for the degradation of humic acid. The initial adsorption extent affected the Ce doped ZnO photocatalytic degradation efficiency of the HA samples in comparison to TiO₂. The humic molecular size distribution profiles as well as the EEM fluorescence features displayed the role of a denser aromatic core. Further assessment on the influence of operational parameters, such as catalyst loading, light intensity and aqueous medium constituents should also be performed. Moreover, Ce leaching into aqueous medium and possible complexation with oxidized and non-oxidized humic molecular fractions should also be investigated in detail. With respect to the attained results, further studies should be designed for the elucidation of the photocatalytic reaction mechanism using ZnO and/or doped ZnO specimens. Based on the above given information as well as the presented ESI,† it could be deduced that using humic acid and its sub molecular size fractions as representatives of NOM should be cautiously interpreted in the elucidation of the photocatalytic activity of a novel photocatalyst. Furthermore, direct evaluation of the experimental results could lead to erroneous interpretation of the photocatalytic activities of the novel materials in comparison to TiO₂.

Acknowledgements

Financial support provided by the Research Fund of Bogazici University through Project Number 11081 is gratefully acknowledged.

References

1. K. Maeda and K. Domen, J. Phys. Chem. Lett., 2010, 1, 2655–2661.
2. S. L. Suib, New and Future Developments in Catalysis: Solar Photocatalysis, Elsevier, Amsterdam, The Netherlands, 2013.
3. A. Fujishima, X. Zhang and D. A. Tryk, Surf. Sci. Rep., 2008, 63, 515–582.
4. A. V. Emeline, V. N. Kuznetsov, V. K. Ryabchuk and N. Serpone, Environ. Sci. Pollut. Res., 2012, 19(9), 3666–3675.
5. N. Serpone and A. V. Emeline, J. Phys. Chem. Lett., 2012, 3(5), 673–677.
6. C. Gionco, M. C. Paganini, E. Giamello, R. Burgess, C. Di Valentin and G. J. Pacchioni, J. Phys. Chem. Lett., 2014, 5(3), 447–451.
7. J. Reszczynska, T. Grzyb, J. W. Sobczak, W. Lisowski, M. Gazda, B. Ohtani and A. Zaleska, Appl. Catal., B, 2015, 163, 40–49.
8. M. Ahmad, E. Ahmed, F. Zafar, N. R. Khalid, N. A. Niaz, A. Hafeez, M. Ikram, A. Khan and Z. Hong, J. Rare Earths, 2015, 33(3), 255–262.
9. O. Bechambi, L. Jlaiel, W. Najjar and S. Sayadi, Mater. Chem. Phys., 2016, 173, 95–105.
10. S. G. Kumar and K. S. R. K. Rao, RSC Adv., 2015, 5, 3306–3351.
11. J.-C. Sin, S. M. Lamb, K. T. Lee and A. R. Mohamed, J. Mol. Catal. A: Chem., 2015, 409, 1–10.
12. C. Yu, K. Yang, Y. Xie, Q. Fan, J. J. Yu, Q. Shu and C. Wang, Nanoscale, 2013a, 5, 2142–2151.
13. C. Yu, K. Yang, W. Zhou, Q. Fan, L. Wei and J. C. Yu, J. Phys. Chem. Solids, 2013b, 74, 1714–1720.
14. C. Yu, W. Zhou, H. Liu, Y. Liu and D. D. Dionysiou, Chem. Eng. J., 2016, 287, 117–129.
15. P. Calza, C. Gionco, M. Giletta, M. Kalaboka, V. A. Sakkas, T. Albanis and M. C. Paganini, J. Hazard. Mater., 2016 DOI:10.1016/j.jhazmat.2016.03.093.
16. M. Bekbolet and G. Ozkosemen, Water Sci. Technol., 1996, 33(6), 189–196.
17. M. Bekbolet, A. S. Suphandag and C. S. Uyguner, J. Photochem. Photobiol., A, 2002, 148, 121–128.
18. C. S. Uyguner and M. Bekbolet, Desalination, 2005a, 176, 167–176.
19. C. S. Uyguner-Demirel and M. Bekbolet, Chemosphere, 2011, 84(8), 1009–1031.
20. N. C. Birben, C. S. Uyguner-Demirel, S. Sen-Kavurmaci, Y. Y. Gurkan, N. Turkten, Z. Cinar and M. Bekbolet, Catal. Today, 2015, 240, 125–131.
21. V. Oskoei, M. H. Dehghani, S. Nazmara, B. Heibati, M. Asif, I. Tyagi, S. Agarwal and V. K. Gupta, J. Mol. Liq., 2016, 213, 374–380.
22. W. Y. Teoh, J. A. Scott and R. Amal, J. Phys. Chem. Lett., 2012, 3, 629–639.
23. C. Jiang, G. R. Aiken and H. Hsu-Kim, Environ. Sci. Technol., 2015, 49, 11476–11484.
24. C. S. Uyguner and M. Bekbolet, Catal. Today, 2005b, 101(3), 267–274.
25. D. Zhou and A. A. Keller, Water Res., 2010, 44, 2948–2956.
26. I. Chowdhury, D. M. Cwiertny and S. L. Walker, Environ. Sci. Technol., 2012, 46, 6968–6976.
27. L. Lutterotti, Nucl. Instrum. Methods Phys. Res., Sect. B, 2010, 268, 334–340.
28. J. K. Edzwald, W. C. Becker and K. L. Wattier, J. AWWA, 1985, 77, 122–132.
29. M. Bekbolet and S. Sen Kavurmaci, Photochem. Photobiol. Sci., 2015, 14, 576–582.
30. P. G. Coble, Mar. Chem., 1996, 51, 325–346.
31. APHA, AWWA and WPCF, Standard Methods for the Examination of Water and Wastewater, ed. E. W. Rice, R. B. Baird, A. D. Eaton and L. S. Cercei, American Water Works Association, Washington, DC, USA, 22nd edn, 2012.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6pp00216a

---