Abstract
Introduction
A current shift in the paradigm of the intensification of pest control in agriculture has increased the demand for pesticides because these have become fundamental tools in the productivity of the agro-industrial sector. 1 –6 This industry produces close to 150 million tons of wastewater during the year worldwide, 7 mainly by the use of herbicides, fungicides, and insecticides, approximately 3.3 × 106 tons per year worldwide, of which 0.42 × 106 tons per year are consumed in Europe (data 2010–2014 of FAOSTAT). 8
These wastewaters show very different characteristics because of the variety of pesticides (herbicides, insecticides, fungicides, rodenticides, nematicides, microbicides, and regulators of plant and insect growth) that have been identified in concentrations of up to 500 mg L−1. 7,9 The excessive amount of pesticide residues in the soil, surface water, and groundwater causes serious contamination of soil and water; due to their high toxicity and persistence, these residues have been declared as priority pollutants by the Environmental Protection Agency of the United States. 10 –12 On the other hand, European legislation for drinking water sets a limit of 0.1 µg L−1 for a single active ingredient of pesticide and 0.5 µg L−1 for the sum of all the individual active substances detected. 7 In Colombia, Resolution No. 2115 of June 22, 2007 establishes that the maximum permissible limit value for the active ingredients of pesticides of toxicological category II 13 is 0.1 mg L−1. 14 Cypermethrin (Class II, moderately dangerous) causes human health problems related to immunosuppression, hormonal alteration, diminished intelligence, reproductive anomalies, and cancer. In the studies focused on these problems, only 0.1% of the pesticide used reaches the target pest and the remaining 99.9% scatter in the environment through routes such as runoff and erosion, leaching, drainage, and discharges of pesticide production plants. 15 In wastewater treatment plants, biological treatment does not eliminate toxic, polar, or nonbiodegradable contaminants, although it is fostered as the best accessible technology. To comply with strict environmental laws, significant research studies are being conducted to develop better and more affordable wastewater treatment schemes, and improvements in processes that can degrade all bio-recalcitrant organic pollutants, minimizing their accumulation. 16 Advanced oxidation processes (AOPs) are characterized by the “in situ” production of hydroxyl radicals (OH●), which are capable of oxidizing and mineralizing any organic molecule, producing CO2, H2O, and inorganic ions. Due to the reactivity of the hydroxyl radicals, its attack is nonselective, which is useful for the treatment of wastewater containing different pollutants. AOPs can be executed with solar irradiation since the sun provides photons with the required wavelength for these processes. 9,10,17 –19
Heterogeneous photocatalysis is an advanced oxidation technology for the degradation of persevering organic pollutants, such as pesticides, which produces by-products that are more biodegradable than the original compounds. Titanium dioxide (TiO2) is a standard photocatalyst used for this process when applying ultraviolet/ultraviolet–visible (UV/UV-vis) spectroscopy. Commercial TiO2 is manufactured by the traditional sulfate process or high-temperature chloride process; the sulfate process involves digesting ilmenite or titania slag with concentrated sulfuric acid to produce titanium sulfate which is hydrolyzed to precipitate a hydrous titanium oxide compound and calcined at 650–1000°C to form either anatase or rutile-type TiO2. 20 On the other hand, emerging technologies such as green chemistry aim to reduce the use of high toxicity reagents and achieve products with high performance and low environmental impact. 21 A further procedure to expand the photo-response of TiO2 in the visible light region of the spectrum is the doping of this material with nonmetallic elements, 22 –24 such as N, C, B, P and F, with noble metals such as Au and transition metals such as Cr, iron (Fe), Ni, V, Mn, and Cu. 25 –30 In the photocatalytic degradation of chlorpyrifos, a composite material of CoFe2O4-TiO2/reduced graphene was used in a batch reactor under UV radiation, reaching a degradation efficiency of 95% for an initial concentration of 5 mg L−1, pH 5.8, a catalyst dose of 0.6 g L−1, and 60 min of reaction. 16,31,32 In a similar study, the batch degradation of atrazine was evaluated using B-TiO2 nanomaterials, reaching efficiencies close to 85% in the photocatalytic process. 33 Additionally, heterogeneous photocatalysis has been integrated with other AOPs such as ozonation. The O3/UV/TiO2 treatment has been proposed for the elimination of cypermethrin, malathion, and dichlorvos, through which a degradation percentage of 83% was obtained for cypermethrin. 9,34 Also, a wide variety of common pesticides such as bromoxynil, diuron, o-phenylphenol, 2-methyl-4-chlorophenoxyacetic acid, and terbuthylazine have been mineralized through heterogeneous photocatalysis using B-TiO2 and Cs-TiO2 catalysts, which exhibit degradation efficiencies greater than 75% for these pollutants. 35,36
In the present research, the photocatalytic degradation of cypermethrin was evaluated using Fe-TiO2 nanoparticles in slurry and supported on coconut palm spathe (
Materials and methods
Materials
Fresh lemongrass plant leaves (
Plant extract and green synthesis of Fe-TiO2 nanoparticles
Lemongrass aqueous extract and TiO2 nanoparticles using green chemistry were obtained through a previously reported procedure. 37 The lemongrass leaves were washed with tap water. Then, they were dried for 6 h in an air circulation oven at 60°C (Esco Isotherm®, OFA 32-8, Changi South, Singapore, Singapore) and crushed with a manual mill. The dried and crushed biomass (100 g) was placed in cloth bags and subjected to a solvent extraction process using a Soxhlet extractor for 6 h in approximately 600 mL of distilled water. 38,39 The extract was cooled down to 4°C so that it could later be used to synthesize TiO2 nanoparticles. 40 In a typical experiment, the reaction was carried out in a 250 mL beaker, which was introduced in an ultrasound processor (WiseClean WUC-A06 H, Acinterlab (Miami, Florida, USA) 60 Hz). Then, 20 mL of the precursor agent (C12H28O4Ti) was added along with 100 mL of the aqueous extract of lemongrass, which was contained in a burette, at a rate of 1 mL s−1; additionally, the reaction lasted approximately 30 min with constant agitation, making use of a stirring rod. 25,40 The nanoparticles were washed with 70 vol% ethanol and then with distilled water, using separation by centrifugation (Universal Centrifuge PLC-012E, GEMMY INDUSTRIAL CORP., Taipei, Taiwan) for 15 min at 5000 r min−1. The synthesized TiO2 nanoparticles were calcined at 550°C for 3 h in a Thermo Scientific FB1415M-1450 W-50/60 Hz muffle. 17 For wet impregnation, 1 g of TiO2 and the amount of FeCl3·6H2O were weighed to obtain Fe3+:Ti molar ratios of 0.05, 0.075, and 0.1 and dissolved in 20 mL of double distilled water with constant magnetic stirring for 1 h at 500 r min−1. Then, the precipitate was washed with distilled water and separated by centrifugation for 15 min at 5000 r min−1; this process was repeated three times. Next, the precipitate was calcined at 350°C for 3 h, and finally, the catalyst was sieved in a 300 mesh. 26,41
Immobilization of Fe-TiO2 nanoparticles on coconut palm spathe
The immobilization of the nanoparticles on the coconut palm spathe was carried out through the doctor blade technique using a water-based emulsion composed of styrene copolymer and 50% solids acrylic ester (RECOL® CRYL, Bogotá, Cundinamarca, Colombia). 42 –46 For this, the coconut palm spathe (10 × 10 cm2) was placed on a flat surface. Then, 4 g of emulsion was used to create a thin film on which 0.15 g of nanopowder was uniformly deposited to ensure that most of the nanomaterial remained on the surface of the composite. Finally, it was homogenized using a metal spatula.
Characterization
The samples were studied by XRD using a Panalytical X’Pert Pro Alpha1 instrument Malvern Panalytical (Madrid, Spain), which is equipped with a primary fast X’Celerator detector operating at 45 kV and 40 mA, and fitted with a primary curved Ge 111 monochromator, to obtain Cu
Solar photocatalysis experiment
A photocatalysis experiment was executed in a lab-scale pilot plant in Cartagena, Colombia (latitude 10° 25′30″ N, longitude 15° 32′25″ W) using natural sunlight irradiation during July–September 2018. Radiation levels were measured with a portable photoradiometer (Delta Ohm HD 2102, Padova, Veneto, Italy) with LP-UVB (300–600 nm) probe. The solar lab-scale pilot used in this experiment is based on FPSR technology (Figure 1). This small prototype consists of one photoreactor module (100 cm2), where the water flows directly from a tube to the flat plate and finally to the reservoir tank (1 L), and a centrifugal pump (12 V DC, 0.7A, 1/2-inch suction/discharge diameter and 8 L min−1 maximum flow). Storage tank, flowmeter (Arduino flow sensor by effect hall YF-S401, working voltage: DC 5–24V, flow rate range: 0.3–6 L min−1), pipes (1/4-inch), and fittings completed the installation. In photocatalysis assay, 250 mL of solution (50 mg L−1 of cypermethrin) was prepared; this will be exposed to sunlight in the flat plate photocatalytic reactor without (photolysis test) and with addition of a catalyst (powder and composite) until it reaches a cumulative solar radiation value of 12,000 J m−2. For the molecular adsorption test, 250 mL (50 mg L−1) was placed in contact with the synthesized composite or 0.15 g of Fe-TiO2 (Fe:Ti = 0). The solution was maintained in constant recirculation at 1.3 L min−1 in a closed wooden box, taking samples at the beginning and after 120 min, separating the solution by centrifugation when this is required.

Flat plate photoreactor design.
Cypermethrin concentration in aqueous solution was determined by GC-MS. Gas chromatographic analysis was carried out on an Agilent 7820A gas chromatography equipped with an HP-5 MS capillary gas chromatographic column (30 m × 0.25 mm ID × 0.25 µm) and a 5975-mass spectrometry. Helium was used as the carrier gas in a constant flow of 53.3 mL min−1 at an input temperature of 250°C. The temperature program of the oven was set as follows: initial temperature at 80°C for 3 min; then ramped at 20°C min−1 to 130°C and 30°C min−1 to 190°C, held for 10 min; then ramped at 40°C min−1 to 280°C, held for 7 min. 34 On the other hand, chromatographic data were analyzed using the ChemStation software (Agilent, Santa Clara, California, USA).
Results and discussion
Characterization
XRD
XRD patterns of synthesized TiO2 nanoparticles using the leaves extract of lemongrass (
where

XRD patterns of Fe-doped and bare TiO2 nanoparticles. XRD: X-ray diffraction; Fe: iron; TiO2: titanium dioxide.
Particle sizes obtained by Debye–Scherrer equation.
Fe: iron; Ti: titanium.
FTIR spectroscopy
FTIR spectra were also registered between the wavenumber of 400 cm−1 and 4000 cm−1 for all samples. The FTIR spectrum of all these samples is shown in Figure 3, showing nearly similar patterns between them. The functional groups of the commercial catalyst (Evonik P-25) prepared TiO2 and Fe-TiO2 were determined using FTIR. The synthesized TiO2 and Fe-TiO2 absorption band at 3400 cm−1 can be assigned to the vibration of OH groups including the Ti–OH group. The band at 1620 cm−1 corresponds to the Fe-TiO2 bond and H–O–H bending vibration of adsorbed water. 17 The absorption peak between 400 cm−1 and 690 cm−1 corresponds to the Ti–O–Ti bond vibration and the symmetrical vibration of Fe–O–Fe bonds, for the case of doped nanomaterials. All four synthesized materials have a high intensity of Ti–O–Ti bond, including the reference material (P-25). 55

Typical FTIR spectra of as-prepared different Fe-doped and bare TiO2. Fe: iron; TiO2: titanium dioxide; FTIR: Fourier-transform infrared.
TEM
The microstructure and morphology of undoped and Fe-doped TiO2 nanoparticles were observed by TEM. Figure 4(a) and (b) shows the TEM images of the undoped TiO2 and Fe-doped TiO2, respectively. It could be found that the shape of all the nanoparticles is generally spherical and quite similar to each other with size varying from 13.73 ± 1.69 nm to 15.60 ± 1.95 nm which are in close with the XRD result.

TEM images of (a) undoped and (b) Fe-doped TiO2 (Fe:Ti = 0.1) nanoparticles. Fe: iron; TiO2: titanium dioxide; TEM: transmission electron microscopy.
Surface area and porosity analysis
N2 physical adsorption/desorption studies were carried out to determine the surface area and the pore distribution of the synthesized TiO2 and Fe-TiO2 samples. The isotherms and their relative Barret–Joyner–Halender (BJH) pore size distributions gathered from the isotherm’s adsorption branch of the catalysts are shown in Figures 5 and 6, respectively.

Nitrogen adsorption–desorption isotherms of bare TiO2 and Fe-TiO2. Fe: iron; TiO2: titanium dioxide.

Cumulative pore volume from BJH adsorption isotherm. BJH: Barret–Joyner–Halender.
In Figure 5, it was observed that the prepared TiO2 and Fe-TiO2 possess mesoporous surface as the isotherms are of type IV and with typical H3 hysteresis loop according to the International Union of Pure and Applied Chemistry (IUPAC) convention; it indicates that TiO2 sample contain nonrigid aggregates of plate-like or slit-shaped pores with a contribution of micropores and mesopores.
56
–59
Furthermore, Figure 6 shows broad pore size distributions, ranging from 1.7 nm to 280 nm, further confirming the existence of micropores, mesopores, and macropores in TiO2 and Fe-TiO2 samples according to the BJH adsorption pore distribution.
57
Additionally, from BET data, the particle size,
Textural properties of pure TiO2 and Fe-TiO2 nanoparticles.
Fe: iron; TiO2: titanium dioxide.
UV-Vis-DRS
The optical properties of TiO2 and Fe-TiO2 were separately detected by ultraviolet–visible diffuse reflection spectrometer. According to Figure 7, the absorption edge of the Fe-TiO2 in comparison to undoped TiO2 is extended greatly toward the visible light. The absorption edge is more extended to the visible light with the increasing of the Fe concentration, which is induced was attributed to the electron transition from Fe3d orbitals to the TiO2 conduction band (CB).
25
Doping Fe3+ causes structural defects of crystal lattice to introduce impurity or defect energy level and induces the local states below the CB edge, then results in this redshift and narrows the band gap.
26
Doping of the Fe ions in the TiO2 introduces new energy levels (Fe3+/Fe4+) of the transition Fe ions into the band gap of the TiO2. The direct band gap energy (

Typical UV-vis spectra and Tauc plot analysis (inset) of bare TiO2 and Fe-TiO2 and band gap energy. Fe: iron; TiO2: titanium dioxide; UV-vis: ultraviolet–visible.
SEM
Figure 8 shows the microscopy of coconut palm spathe and coated by the doctor blade technique. Samples prepared with water-based emulsion composed of styrene copolymer and 50% solids acrylic ester (Figure 8(c) to (e) and (f)) present a coarse nonhomogeneous layer of TiO2 or Fe-TiO2 and some cracks are visible, similar to that reported by Mejía et al. 63

SEM micrograph of (a and b) coconut palm spathe, (c) Fe:Ti = 0/spathe, (d): Fe:Ti = 0.05/spathe, (e) Fe:Ti = 0.075/spathe, and (f) Fe:Ti = 0.1/spathe. Fe: iron; Ti: titanium; SEM: scanning electron microscopy.
Photocatalytic activity and mechanism
Results obtained for the photocatalytic degradation of cypermethrin using nanomaterials suspended and immobilized are shown in Figure 9 and the GC-MS analysis for the slurry Fe:Ti = 0.05 sample is illustrated in Figure 10. The best results were achieved in slurry with the use of TiO2 Evonik P-25, Fe:Ti = 0 and Fe:Ti = 0.05, with rates of degradation of cypermethrin of 99.84%, 99.62%, and 100%, respectively. However, the molar ratios higher than Fe:Ti (Fe:Ti = 0.075 and Fe:Ti = 0.1) had a negative effect on the percentage degradation, like what was reported recently by Moradi et al. 64 In this study, the authors evaluated the photocatalytic degradation of methyl orange (20 mg L−1) using TiO2 doped with different concentrations of Fe3+ (0.25, 0.5, 1, 5, and 10 Fe:Ti mol% ratio). These authors found that it is only possible to achieve degradation percentages lower than 5% with the use of catalysts doped with high concentrations of Fe3+ (5–10 Fe:Ti mol% ratio), which is attributed to the formation of layers composed of iron oxide, with photocatalytic reactions generating pollution on the surface of the catalyst particles, which decreases the available active sites, where they occur. 64 For this reason, the authors proposed an acid treatment with HCl (pH 2 solution) to remove impurities that cause inhibition of the photocatalytic activity, achieving a significant improvement in the percentage of degradation.

Results for the cypermethrin solar photocatalytic degradation using GC-MS. GC-MS: gas chromatography–mass spectrometry.

GC-MS analysis for the Fe:Ti = 0.05 sample using HP-5 MS capillary gas chromatographic column, 5975-mass spectrometry, helium as carrier gas (53.3 mL min−1) at an input temperature of 250°C. Fe: iron; Ti: titanium; GC-MS: gas chromatography–mass spectrometry.
On the other hand, the phenomenon of inhibition caused by high concentrations of the dopant has also been explained by other authors. Ali et al. 53 studied the photocatalytic activity of TiO2 doped with Fe3+ ions for the elimination of methylene blue (10 mg L−1), finding high concentrations of dopant (5, 7, and 10 Fe:Ti mol% ratio) produces accumulation on the surface of the catalyst, which reduces the depth of light penetration. Besides, such saturation covers part of the surface of the photosensitive, thus reducing the number of surface active sites for the effective degradation of organic molecules. 53
The generalized mechanism for degradation of cypermethrin is shown in Figure 11. Photocatalytic degradation is initiated by electron/hole pairs, and the recombination of these carrier species decreases the photocatalytic efficiency of TiO2. Fe3+ ions can act as trap for photo-generated carriers (equations (4) to (8)). However, when the doping agent concentration is very high, Fe3+ ions become recombination centers for electron/hole pairs, resulting in reduced photocatalytic activity (equations (9) to (13)). 25,64 From Figure 9, it can be seen that Fe:Ti = 0.05 (molar ratio) is the optimal Fe3+ doping content for the degradation of the cypermethrin under visible light irradiation. When the Fe:Ti molar ratio was higher than 0.05 (Fe:Ti = 0.075 and 0.1), the photocatalytic activity decreased. It means that the Fe ions will become recombination centers for high doping levels 25

Generalized mechanism for degradation of cypermethrin.
Many authors have studied the behavior of Fe-doped TiO2 materials through photoelectrochemical and electrochemical phenomena; most research has focused on determining the optimal Fe concentration in applications such as photocatalysis, solar cells, and photoanodes. The most commonly used electrochemical techniques are photoelectrochemical cells (PECs), electrochemical impedance spectroscopy (EIS), intensity-modulated photocurrent spectroscopy, and Mott–Schottky analysis. 65 –73 In a current research, Chakhari et al. 74 studied the electrochemical properties of Fe-TiO2 nanorods deposited on fluorine-doped SnO2 conductive glass for application as photoanodic materials. The Fe+3 concentration was varied between 0 mM and 10 mM during material synthesis, and electrochemical properties were determined by experiments in PECs using Ag/AgCl as a working electrode under illumination, and EIS in open circuit in dark and under illumination with an AC perturbation of 10 mV. Results showed that high Fe concentrations decrease photocurrent density as a result of the increase in the recombination rate because photo-generated electron or hole trapping sites will disappear, and more oxygen vacancies act as recombination centers; on the other hand, the EIS (Nyquist plots) for TiO2 and Fe-TiO2 under illumination revealed a significant increase in impedance due to a poor charge transfer for materials with excess of doping.
Based on the results shown in Figure 9, it can be inferred that the immobilization of nanoparticles on the spathe of coconut palm biomaterial generates a decrease in the percentages of elimination of cypermethrin, although it is possible to achieve efficiencies of degradation close to 85% for reference Evonik P-25, Fe:Ti = 0 and Fe:Ti = 0.05, while the doping made with high molar ratios of Fe:Ti (Fe:Ti = 0.075 and Fe:Ti = 0.1) only allowed to achieve efficiencies of 67% and 48%, respectively, due to above reasons. Also, there is evidence of the appearance of a yellowish coloration, which is attributed to nonselective degradation of the resin and the biological material used for the immobilization of the powder materials by the hydroxyl radicals generated (OH●). 75
There is also a clear decrease in the efficiency of degradation of the supported systems in comparison to the evaluated in suspension (slurry). Recent studies support the low-efficiency for immobilized catalysts, which is associated with a decrease in the surface area exposed to the solution. For example, Cerrato et al. 76 studied the immobilization of TiO2 on glass materials for the degradation of Ibuprofen under UV-C radiation, finding that the application in suspension allows to achieve the best results for the degradation of the studied drug.
The cypermethrin elimination or degradation percentages were determined using equation (14). The results showed values lower than 5% and 10% for the photolysis and dark adsorption tests, respectively, which correspond to commonly published in research related to targets of molecular adsorption and photolysis of organic compounds, such as pesticides, dyes, and pharmaceuticals, among others. 55,76,77
where
Conclusions
The immobilization of nanoparticles on the spathe of coconut palm biomaterial generates a decrease in the percentages of elimination of cypermethrin, although it is possible to achieve efficiencies of degradation close to 85% for reference Evonik P-25, Fe:Ti = 0 and Fe:Ti = 0.05, while the doping made with high molar ratios of Fe:Ti (Fe:Ti = 0.075 and Fe:Ti = 0.1) only allowed to achieve efficiencies of 67% and 48%, respectively. However, the immobilization of nanoparticles on coconut palm-based biomaterial using the resin applied in this study brings two important advantages from an economic and environmental point of view, that is, it eliminates operating costs associated with the separation of the material after the treatment when applied in suspension, and also prevents potential contamination of the effluent treated by TiO2 or Fe-TiO2 nanoparticles, which could impact negatively on the aquatic ecosystems and human health.
