Catalysis Today 328 (2019) 230–234

Contents lists available at ScienceDirect

Catalysis Today
journal homepage: www.elsevier.com/locate/cattod

New insight into zinc oxide doped with iron and its exploitation to T
pollutants abatement
Maria Cristina Paganini, Alice Giorgini, Nuno P.F. Gonçalves, Chiara Gionco,
⁎
Alessandra Bianco Prevot, Paola Calza
Department of Chemistry, Università di Torino, Torino, Italy

A R T I C LE I N FO A B S T R A C T

Keywords: This study aims to investigate how the photocatalytic efficiency of zinc oxide can be improved by introducing
Iron doping doping species into its crystal lattice. New materials based on zinc oxide doped with iron, have been synthesized
Zinc oxide using different methods, characterized and tested toward the abatement of selected organic molecules. The
Pollutant abatement different synthetic strategies followed comprised sol-gel, precipitation and hydrothermal processes, in order to
Ketoprofen
identify which one is capable of guaranteeing the best photocatalytic performances.
The photoactivity of the new semiconductors was firstly tested using phenol as a model molecule subjected to
irradiation under UV-A light. Phenol abatement is particularly favoured when using ZnO prepared via hydro-
thermal method and doped with iron at 0.5%. These materials were then tested toward the elimination of
ketoprofen, an emerging pollutant substance, from water and real wastewater. Ketoprofen and its transformation
products are completely abated within 30 min in pure water or in 1 h in wastewater.

1. Introduction a better efficiency in the electrons /holes charge separation [3].

The present work aims to develop new photoactive materials, po-
Heterogeneous photocatalysis could play a fundamental role in the tentially applicable to water treatment. Several synthetic strategies, not
degradation of recalcitrant organic molecules present in environmental expensive and easy to perform, have been investigated, aimed to
matrices. So, over the years a lot of research has been devoted to the achieve the development of materials with enhanced photocatalytic
development of semiconductor materials that could be applied to the performances. Among all the possible doping we have chosen iron, a
purification of water. Among the semiconductor oxides, titanium di- transition metal that is easy to find and very abundant on the earth's
oxide and zinc oxide seem promising candidates. crust. Also zinc oxide is particularly easy to find, so the Fe/ZnO system
Transition metal doping is one of the best-known strategies for can be a sustainable and economical photocatalyst. We assessed how
improving the photocatalytic activity of a semiconductor, as these the photocatalytic efficiency of the zinc oxide can be improved by in-
elements are able to modify the oxide band structure, morphology and troducing iron in its reticular structure. The effect of the type of syn-
particle size. When added in proper quantities, the metallic sites act as thetic process (sol-gel, precipitation and hydrothermal methods) and of
traps for the photoproduced charges, so increasing the charge transfer the dopant concentration (from 0.5% to 3%) on the efficiency of pho-
and reducing the extent of charge recombination [1]. Moreover, the tocatalysis was studied.
doping with metals of d block solves the problem related to photo-
corrosion of zinc oxide which becomes more stable and thus more ap- 2. Materials and methods
pealing to be used in wastewater treatment [2].
The positive role of transition metal doping is expressed through the 2.1. Materials
introduction of additional energy levels inside the oxide band gap.
Furthermore, the addition of transition metals changes the chemical The precursors and reagents used in the synthesis of new materials
surrounding of Zn2+ ions, induces reticular distortions and increases and in subsequent analyzes for the evaluation of their photocatalytic
the fraction of defects. In particular, the oxygen vacancies, if present in activity, were purchased by Sigma-Aldrich. Suspensions and standard
optimal concentrations, can act as traps for the electrons and guarantee solutions were prepared in MilliQ water.

⁎
Corresponding author at: Università di Torino, via P. Giuria 5, 10125 Torino, Italy.
E-mail address: [email protected] (P. Calza).

https://doi.org/10.1016/j.cattod.2018.10.054
Received 27 June 2018; Received in revised form 4 September 2018; Accepted 22 October 2018
Available online 28 October 2018
0920-5861/ © 2018 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/BY-NC-ND/4.0/).
M.C. Paganini et al. Catalysis Today 328 (2019) 230–234

The real wastewater samples were provided by SMAT S.p.A. Pyrex glass cells, filled with 5 mL of model molecule (20 mg/L) and
(Società Metropolitana Acque Torino, Italy) and were sampled at the catalyst (1000 mg/L or 400 mg/L). Samples were irradiated for dif-
wastewater treatment facility of Castiglione Torinese (Italy) on 27 ferent times (from 5 min to 120 min) using a PHILIPS cleo 6 × 15 W TL-
October 2017. The samples used in this work were obtained from the D Actinic BL with maximum emission wavelength at 365 nm. The UV
outflow of the primary clarifier tank (hereafter, Primary sample). Real integrated irradiance on the cells in the 290–400 nm range wavelengths
samples were used after a rough pre-filtration step, carried out through was 90 ± 2 Wm–2 (measured with a CO.FO.MEGRA (Milan, Italy)
a grade 1 qualitative filter paper (Whatman) to remove large suspended power-meter). After irradiation, samples were filtered through a
solids and filtered using a hydrophilic 0.45 μm filter Sartolon Polyamide 0.45 μm filter and analyzed with the proper analytical technique.
by Sartorius Biolab. Ketoprofen degradation was also performed in effluent wastewater
provided by SMAT S.p.A. (Turin, Italy). Total organic carbon measures
2.2. Synthesis procedures on wastewater matrix allowed to determine a concentration of organic
component equal to 9.7 mg/L and inorganic carbon 68.4 mg/L. The
2.2.1. Synthesis via sol-gel (SG) organic nitrogen measures allowed to determine of 31.3 mg/L of total
Zinc acetate is solubilized in 250 mL of deionized water; the desired nitrogen.
stoichiometric amount of Fe(NO3)3 is added to the solution together
with 5 g of citric acid. The solution is subjected to magnetic stirring and 2.3.4. HPLC/UV
is kept for 6 h in a hot water bath at 80 °C, in order to favor the gel The concentrations of phenol and ketoprofen were measured
formation [4]. The acidic pH favors the hydrolysis process and there- through a YL9300 HPLC system equipped with a YL9330 Column
fore the gelation. The beaker containing the gel is placed in an oven at Compartment and a YL9150 autosampler. The column was a RP C18
70 °C for two days, in order to promote drying and obtaining the xer- column (LiChroCART®, Merck, 12.5 cm × 0.4 cm; 5 m packing). Phenol
ogel. This phase is followed by a further heat treatment which involves was analyzed using acetonitrile and phosphoric acid solution (1 × 10−2
calcining in a muffle at 500 °C for 2 h. M) at pH 2.8 (10: 90% v/v) as eluent at a flow rate of 1 mL/min and UV
detector set at 220 nm. Ketoprofen was analyzed using as eluent acet-
2.2.2. Synthesis via precipitation (P) onitrile and phosphoric acid solution (1 × 10−2 M) at pH 2.8 (35: 65%
The precipitation method is a simple and versatile approach for the v/v) at a flow rate of 1 mL/min and UV detector set at 260 nm.
synthesis of semiconductor materials; it is cheap, does not require
complex experimental conditions [5]. The solution of the precursors, 2.3.5. HPLC/MS
named A, is obtained by adding 2.2 g of zinc acetate dihydrate and the Samples were analyzed by HPLC/MS Shimadzu UFLC equipped with
stoichiometric amount of FeCl3 to 160 mL of deionized water. 80 mL of a Nexera XR, LC-20AD XR pump and a UV–vis detector SPD-M20A.
ethanol are then added to solution A and the system is kept under Injection volume was 10 μL and flow rate 0.4 mL/min using a column
magnetic stirring for 1 h in order to promote the solubilization of in- Phenomenex Luna C18 50 × 2 mm, 2.5 μm and a gradient mobile phase
organic salts. Then, 160 mL of NaOH 1 M (solution B) are added drop by composed of acetonitrile/water was: 0 min, 5/95; 5 min 0 /100; 7 min,
drop to the solution A. The precipitate is left to rest for two days, re- 0/100. A 3200 Q Trap Sciex mass spectrometer equipped with an at-
covered by filtration and dried in an oven at 70 °C. The powder is finally mospheric pressure interface and an ESI ion source was used. Samples
calcined at 300 °C for 30 h [6]. were analyzed using ESI negative mode operating in full-scan mode
between 50 and 300 m/z range.
2.2.3. Synthesis via hydrothermal method (H)
The doping agent (FeCl3), in the desired percentage, was added to 2.3.6. Total organic carbon
20 mL of Zn(NO3)2∙6H2O 1M solution, and the resulted solution was Total organic carbon (TOC) and total nitrogen (TN) were measured
kept under stirring. After complete dissolution, a solution of NaOH 4M using a Shimadzu TOC-5000 analyzer (catalytic oxidation on Pt at
was added dropwise until reaching the pH value of 10. The solution was 680 °C). The calibration was performed using standards of potassium
transferred to a 100 mL Teflon-lined autoclave, filled with water until phthalate.
complete 50 mL, and the autoclave was kept at 175 °C for 15 h. The
resulting precipitated product was washed with water and recovered by 2.3.7. Iron leaching
centrifugation (6000 rpm for 10 min), repeating the procedure 2 times. The determination of iron released in solution was evaluated by a
After dried at 70 °C during the night, the powder was homogenized spectrophotometric procedure. After irradiation, the 1000 ppm sus-
using a mortar and pestle. pension of catalyst was filtered with hydrophilic PTFE Millex-LCR fil-
ters (0.45 mm pore diameter). The iron released was determined by
2.3. Methods adding ascorbic acid (0.5 mM) to reduce the Fe(III) to Fe(II) and com-
plexing the Fe(II) with o-phenanthroline (5 mM). The absorption of the
2.3.1. XRD iron-thiocyanate complex was determined at 510 nm.
X-rays powder diffraction (XRPD) patterns were recorded using a Spectrophotometric analyses were performed using a Varian CARY 100
PANalytical PW3040/60 X’Pert PRO MPD, Lissone (MI) Italy (45 kV, Scan double-beam UV–vis spectrophotometer, using quartz cuvettes
40 mA) with a copper Kα radiation source (0.15418 nm). Samples were with 10 mm path length.
scanned continuously in the 2θ range between 10° and 100°. The X’Pert
High-Score software was used to identify the mineral phases present in 3. Results and discussion
the samples.
3.1. Materials characterization
2.3.2. DR-UV-Vis
The UV–vis absorption spectra were recorded using a Varian Cary 3.1.1. Iron doped materials
5000 spectrophotometer, coupled with an integration sphere for diffuse Fig. 1 shows the diffraction patterns of pure and doped samples. All
reflectance studies (DRS) and a Carywin-UV/scan software. A sample of the diffractograms present the reflections typical of wurtzite ZnO (ICDD
PTFE with 100% reflectance was used as reference. 01-089-7102 reference pattern). None of the doped samples show re-
flections related to other phases, such as iron oxides, α-Fe2O3 or γ-
2.3.3. Photocatalytic degradation experiments Fe2O3 [7]. This means either that the iron ions entered the structure of
The irradiation experiments were carried out in magnetic stirred zinc oxide substitutionally or that the iron oxide phases are too small to

231
M.C. Paganini et al. Catalysis Today 328 (2019) 230–234

Fig. 1. XRD patterns of bare (trace a) and Fe-doped ZnO with different iron contents: 0.5% (trace b), 1% (trace c) and 3% (trace d). Samples were prepared via sol-gel
(panel A), precipitation (panel B) and hydrothermal methods (panel C).

be detected with the XRD analysis. the analogous system Fe-TiO2, to the charge transfer between Fe3+
The addition of iron changes the average crystalline size of ZnO, states and the oxide conduction band. This band is present in the
calculated using the Debye-Scherrer equation, as reported in Table 1. samples prepared with the precipitation and the hydrothermal
Indeed, the XRD peaks of the doped samples result broadened with methods, while it was absent in the samples prepared through the sol-
respect to the pattern of the bare oxide. This phenomenon is quite gel method. The second feature, represents the main absorption in the
known in literature, and it is attributable to the fact that the presence of visible region and it is centered at 500 nm. This band has been attrib-
the dopant ion prevents the growth of crystallites, favoring instead uted in the past to “on-site” transitions (d-d transitions on the Fe3+ ion)
nucleation. The hydrothermal synthesis leads to the formation of ma- or to inter-site transitions (between two Fe3+ ions) [8–11].
terials with a higher degree of crystallinity (average crystallite size
around 100–150 nm) as evidenced also by the comparison of pure ZnO 3.2. Photoactivity tested on a probe molecule
samples for the different synthesis methods reported in the supporting
information (Fig. S1), while the sol-gel and precipitation synthesis both 3.2.1. Influence of the synthesis method
form crystallites with an average size around 20 nm. The photocatalytic activity of all synthesized materials was eval-
Fig. 2 shows DR-UV–vis spectra for all samples. It is clear from the uated by using phenol as a model molecule in MilliQ water. Fig. 3
spectra, and from the energy gap values calculated using the Tauc plot shows the degradation curves obtained by irradiating the probe mole-
reported in Table 1, that the addition of iron does not affect dramati- cule in the presence of pristine or doped zinc oxide, at different dopant
cally the fundamental edge transition due to the excitation between the concentrations produced through hydrothermal method, while data
material’s valence and conduction bands. On the other hand, the ad- obtained from sol-gel and precipitation methods are plotted in Figs. S3
dition of iron leads to the formation of an additional absorption and S4, respectively.
shoulder, that can be related to iron states. The intensity of this band is For a complete view, we report in Table 1 the kinetic constants
highly dependent on the material’s synthesis, the most intense being the values calculated from phenol degradation curves for all synthesized
one obtained with the precipitation method. Fig. S2 reports the com- materials. Materials prepared via sol-gel method exhibit a poor activity
parison of the DR spectra for the samples containing 3% of iron ob- and permit to abate only a small percentage of phenol (from 20 to 40%
tained with the different synthesis. Fig. S2 highlight the presence of two after 2 h of irradiation, see Fig. S3), while oxides produced through
intra-band features. The first one, centered at 420 nm, was assigned for hydrothermal method were the most efficient.
The materials prepared via hydrothermal method allowed to de-
grade the probe molecule with kinetic constants which are greater of
Table 1 one order of magnitude compared to those characteristics of the pre-
Average size of crystallites (d), energy gap values (EBG) and calculated kinetic cipitated powders. This phenomenon could be probably due to the
constant (phenol 20 mg/L, catalyst 1000 mg/L) for pure and doped zinc oxide
higher degree of crystallinity of the former as pointed out by the XRD
samples.
patterns. Iron leaching was tested as well on doped materials prepared
Synthesis % dopant d [nm]a EBG [eV]b k, min−1 via hydrothermal method and, in all case, is negligible.
SG 0 21 ± 7 3.24 0.04
0.5 20 ± 5 3.21 0.011 3.2.2. Doping effect
1 21 ± 5 3.23 0.0048 Data collected in Table 1 allowed to assess that iron doping induced
3 14 ± 4 3.24 0.0036 an enhancement in the phenol degradation rate compared to pristine
P 0 37 ± 5 3.27 0.03
ZnO, with SG method only exception. Results obtained for materials
0.5 19 ± 6 3.27 0.044
1 19 ± 5 3.27 0.025 prepared via hydrothermal or precipitation method show that ZnO
3 18 ± 4 3.25 0.026 doped with the lowest concentration of iron, namely 0.5%, promotes
H 0 143 ± 30 3.28 0.058 the phenol degradation with the highest kinetics, while the photo-
0.5 112 ± 17 3.27 0.171 activity is reduced as the dopant concentration increases.
1 91 ± 13 3.27 0.123
3 120 ± 14 3.25 0.035
In fact, the sample containing the lowest dopant concentration en-
sures the complete elimination of the phenol after only 30 min of ir-
a
As calculated from the Scherrer equation. radiation with Fe(0.5%)-ZnO-H (see Fig. 3) or 60 min with Fe(0.5%)-
b
As obtained from the Tauc plot. ZnO-P (see Fig. S4). Instead, as the amount of iron increases, the

232
M.C. Paganini et al. Catalysis Today 328 (2019) 230–234

Fig. 2. DR spectra for bare (trace a) and Fe-doped ZnO with different iron contents: 0.5% (trace b), 1% (trace c) and 3% (trace d). Samples were prepared via sol-gel
(panel A), precipitation (panel B) and hydrothermal (panel C) methods.

Scheme 1. Energetic levels for Fe-doped ZnO.

Fe2+ ion can then easily transfer an electron to the oxygen adsorbed on
the semiconductor surface, so favoring the formation of the superoxide
anion radical, as described in reactions (1) and (2).
Fig. 3. Phenol degradation (20 mg/L) under UV-A in the presence of (top) Fe- Fe3 + + eCB− → Fe2 + (1)
ZnO-H 1000 mg/L and (bottom) kinetic constants of phenol abatement in the
presence of different catalysts at 400 mg/L and 1000 mg/L. Fe2 + + O2(ads) → Fe3 + + O−∙
2 (2)
3+
Fe ion can also act as a capture center in the valence band. The
performances are reduced and become lower than those of the pristine resulting Fe4+ combines with the surface hydroxyl groups to produce
semiconductor; less than 90% of the initial phenol was eliminated in 1 h the OH radicals [13].
when the highest concentration of dopant is used.
For comparison purpose, the kinetic constants for all materials ob- Fe3 + + h+VB → Fe4 + (3)
tained via hydrothermal method are plotted in Fig. 3 (bottom), where (4)
Fe4 + + OH− → Fe3 + + OH∙
two catalysts concentration (1000 mg/L or 400 mg/L) were considered.
As expected, although the values of the kinetic constants are reduced by Therefore, the enhanced degradation of phenol obtained following
decreasing the photocatalyst concentration, the order of reactivity al- iron doping could be attributed to the increased production of reactive
ready observed for the materials is maintained. oxygen species, described through reactions (1)–(4).
To explain the reasons for such experimental evidences, it is ne- However, an excessive addition of dopant inhibits the degradation
cessary to consider the role played by the dopant within the crystalline process. The order of photoactivity of the materials is modified for the
lattice of zinc oxide. It is well-known that electron/hole recombination highest concentrations of doping. In this case, pure oxide has the
limits the photocatalytic efficiency and has therefore to be prevented. greatest photoactivity, attributable to an increase in electron/hole re-
Iron acts as a surficial trap for the photogenerated charges generated combination with a consequent reduction in the performance of doped
and, therefore, prevents or in any case reduces the electrons/holes re- semiconductors [14]. In fact, a high concentration of iron reduces the
combination [12]. The possibility for Fe3+ to undergo redox reactions degree of crystallinity of the solid and increases the fraction of reticular
introduces two discrete levels of energy into the oxide band gap, as defects, which represent preferential centers where electron/hole re-
shown in Scheme 1. combination could occur. From the collected data it can be inferred that
The oxide conduction band has an energy close to the redox po- 3% of dopant is enough to favor charge recombination rather than
tential of the Fe3+/Fe2+ pair and Fe3+ ion can, therefore, behave as a charge separation. However, some studies showed that the highest
scavenger for the photogenerated electrons and be reduced to Fe2+. photoactivity is guaranteed by the materials containing the highest
doping concentrations [5].

233
M.C. Paganini et al. Catalysis Today 328 (2019) 230–234

Fig. 4. Ketoprofen degradation (20 mg/L) in the presence of 1000 mg/L of catalyst in Milli-Q (MQ) water and wastewater (ww).

3.2.3. Phenol transformation products development of materials characterized by excellent photocatalytic

The evolution over time of the main transformation products (TPs) performances. The initial objective of increasing the photoactivity of
arose during phenol degradation was followed as well. The formation of zinc oxide is reached by using doping with iron at low concentrations
catechol, hydroquinone and resorcinol occurred in all cases, in agree- (0.5%), for which a clear improvement is observed in the abatement of
ment with literature data [15–17], and their evolution profiles are the model molecule. The transition metal ions, if present in suitable
plotted in Fig. S6. quantities, act as surface traps for the photoinduced charges and reduce
Doping with the lowest percentages of iron allows the abatement of e−/h+ recombination. Conversely, when present at high concentra-
the intermediate in times shorter than those necessary for pristine tions, charge recombination prevails on charge separation, so affecting
oxide. The disappearance of the intermediate is already observed at (and decreasing) the degradation performance.
30 min of irradiation with ZnO doped at 0.5% and 1%. Instead, 1 h of From the obtained results it is possible to deduce that the synthe-
irradiation is required for the 3% Fe-doped material. In particular, ca- sized materials are efficient, not only in the elimination of the phenol,
techol maximum concentration is achieved, for all the materials, at but also in the removal of the degradation intermediates. All the ma-
short irradiation times (10 min of irradiation for pristine ZnO and ZnO terials synthesized by hydrothermal method are able to completely
doped with the lowest concentrations of iron, and 15 min for ZnO degrade the aromatic products formed by the degradation of phenol
doped at 3%). Similarly, to what has been observed for catechol, the (catechol, hydroquinone and resorcinol) within 2 h of irradiation.
disappearance kinetics of hydroquinone is slower for samples char- The best performing materials have also been successfully applied
acterized by the highest doping content. for the abatement of ketoprofen in real wastewater.

3.3. Application to emerging contaminant degradation Acknowledgements

The best performing material, namely Fe(0.5%)-ZnO-H, was also This work is part of a project that has received funding from the
tested toward the abatement of ketoprofen, a contaminant of emerging European Union's Horizon 2020 research and innovation programme
concern, and the degradation profiles are shown in Fig. 4. Analyzing under the Marie Skłodowska-Curie Grant Agreement No 765860
data in MilliQ water, Fe(0.5%)-ZnO-H permits to achieve the complete (AQUAlity). We acknowledge support from MCA-H2020-RISE
degradation of ketoprofen in 20 min, while ZnO-H requires 45 min. (MAT4TREAT, no. 645551) as well.
The effect of matrix composition was assessed as well after added
spiking a real wastewater with ketoprofen and then by studying the References
drug degradation. The high organic load, combined with the presence
[1] R. Saleh, N.F. Djaja, Spectrochim. Acta A. Mol. Biomol. Spectrosc. 130 (2014) 581–590.
in the real matrix of inorganic ions, is the reason why the degradation [2] M. Samadi, M. Zirak, A. Naseri, E. Khorashadizade, A.Z. Moshfegh, Thin Solid Films 605
curves go to zero more slowly than those acquired in MilliQ water [18]. (2016) 2–19.
Even if rates are slightly decreased, all materials are able to completely [3] Y. Lu, Y. Lin, D. Wang, L. Wang, T. Xie, T. Jiang, Nano Res. 11 (4) (2011) 1144–1152.
[4] W. Bousslama, H. Elhouichet, M. Férid, Optik 134 (2017) 88–89.
abate the molecule within 1 h of irradiation. [5] R. Saleh, N.F. Djaja, Superlattices Microstruct. 74 (2014) 217–233.
The formation of transformation products (TPs) was assessed too [6] E. Cerrato, C. Gionco, M.C. Paganini, E. Giamello, J. Phys. Condens. Matter 29 (2017)
444001.
and their evolution profiles are plotted in Fig. S7 (ZnO-H) and S8 (Fe
[7] H. Liu, J. Yang, Y. Zhang, L. Yang, M. Wei, X. Ding, J. Phys. Condens. Matter 21 (14)
(0.5%)-ZnO-H). Ketoprofen transformation involved in all cases the (2009) 145803–146001.
formation of several hydroxylated derivatives, in a close analogy with [8] K.C. Christoforidis, A. Iglesias-Juez, S.J.A. Figueroa, M. Di Michiel, M.A. Newton,
M. Fernández-García, Catal. Sci. Technol. 3 (2013) 626–634.
TPs already detected during the treatment with titanium dioxide [19] [9] T. Umebayashi, T. Yamaki, H. Itoh, K. Asai, J. Phys. Chem. Solids 63 (2002) 1909–1920.
and in surface water [20]. In the presence of Fe(0.5%)-ZnO-H, all TPs [10] J. Zhang, X. Chen, Y. Shen, Y. Li, Z. Hu, J. Chu, Phys. Chem. Chem. Phys. 13 (2011)
are formed within few minutes (maximum at 5 min) and completely 13096–13105.
[11] J. Zhu, W. Zheng, B. He, J. Zhang, M. Anpo, J. Mol. Catal. A Chem. 216 (2004) 35–43.
disappeared within 30 min, while with ZnO-H the evolution profiles are [12] C. Adàn, A. Bahamonde, M. Fernàndez-Garcìa, A. Martìnez-Arias, Appl. Catal. B 72
delayed and a longer irradiation time is required to achieve the com- (2007) 11–17.
[13] M. Ba-Abbad, A.A.H. Kadhum, A.B. Mohamad, M.S. Takriff, Chemosphere 91 (2013)
plete TPs abatement. 1604–1611.
[14] V.C. Srivastava, P. Chandra, Ind. Eng. Chem. Res. 52 (2013) 17790–17799.
[15] T.T.T. Dang, S.T.T. Le, D. Channei, W. Khanitchaidecha, A. Nakaruk, Res. Chem.
4. Conclusions Intermed. 42 (2016) 5961–5974.
[16] Z. Guo, R. Ma, G. Li, Chem. Eng. J. 119 (2006) 55–59.
[17] K.I. Okamoto, Y. Yamamoto, H. Tanaka, M. Tanaka, A. Itaya, Bull. Chem. Soc. Jpn. 58
The sol-gel synthesis led to the design of materials with limited (1985) 2015–2022.
photocatalytic activity, while the most promising synthesis procedures [18] S. Carbonaro, M.N. Sugihara, T.J. Strathmann, Appl. Catal B: Environ. 129 (2013) 1–12.
were those obtained via precipitation and hydrothermal processes. [19] C. Martínez, S. Vilari ˜no, M.I. Fernández, J. Faria, L.M. Canle, J.A. Santaballa, Appl.
Catal. B 142–143 (2013) 633–646.
Among the various strategies of investigated synthesis, the hydro- [20] A. Jakimska, M. Śliwka-Kaszyńska, J. Reszczyńska, J. Namieśnik, A. Kot-Wasik, Anal.
thermal method proved to be the most advantageous in the Bioanal. Chem. 406 (2014) 3667–3680.

234