Greener Method for the Application of TiO2 Nanoparticles to Remove Herbicide in Water

TiO2 nanoparticles have emerged as a great photocatalyst to degrade organic contaminants in water; however, the nanoparticles dispersed in water could be difficult to be recovered and potentially become contaminant. Herbicide like 2,4-dichlorophenoxyacetic acid (2,4-D) used in agriculture usually ends up with a large fraction remaining in water and sediment, which may cause potential risk to human health and the ecosystem. This study proposes a greener method to utilize TiO2 as photocatalyst to remove 2,4-D from water. Accordingly, TiO2 nanoparticles (10–45 nm) were synthesized and grafted on lightweight fired clay to generate a TiO2-based floating photocatalyst. Experimental testing revealed that 60.2% of 2,4-D (0.1 mM) can be decomposed in 250 min under UV light with TiO2-grafted lightweight fired clay floating on water. Degradation fits well into the pseudo-first-order kinetic model. The floating photocatalysts can degrade approximately 50% 2,4-D in 250 min under sunlight and the degradation efficiency is stable for cycles. The results revealed that the fabrication of floating photocatalyst could be a promising and greener way to remove herbicide contaminants in water using TiO2.


Introduction
2,4-dichlorophenoxyacetic acid (2,4-D) is an herbicide agent that has been widely used to control broadleaf weeds in agriculture and urban landscape practices [1]. Tis chemical has been registered as an active ingredient in approximately 1500 herbicide formulations, with a large amount being produced and consumed worldwide every year [2]. In China, for example, 2,4-D production reached 40,000 tons/year in 2010 [3]; meanwhile, the consumption in the USA was about 13,000-15,000 tons annually in 2001 [2]. Herbicides are usually applied onto soil or sprayed over crops; therefore, they can reach superfcial water and sediments [4]. It is estimated that 91.7% of 2,4-D ends up in surface and ground water due to its high solubility in water [5]. 2,4-D is a moderately persistent chemical, which can be decomposed by both photodegradation and microbial degradation at a very slow rate, with a half-life between 20 and 312 days [6]. 2,4-D contamination could be the source of health hazard to exposed animals and human, which may cause the endocrine disruption, reproductive disorder, genetic alterations, and carcinogenic efects [1]. Because of these environmental and health concerns, it is necessary to eliminate 2,4-D from water.
Floating photocatalyst-based water treatment technology (FPWT) that uses sunlight to breakdown pollutants has recently attracted great attention because of its potential for large-scale application, particularly to treat water resources that are contaminated with persistent organic pollutants (POPs) such as herbicides, pesticides, and antibiotics [25,26]. Tese pollutants, which usually originated from agriculture, aquaculture, livestock, medicine, and chemical industries enter water reservoir and could not be removed by normal water treatment plants [1,27,28]. FPWT breaks POPs under sunlight using photocatalysts that are grafted on a foating substrate. When the foating photocatalysts (FPC) are dispatched to a water reservoir, it will foat and continue degrading POPs under sunlight irradiation without any requirement of external intervention.
Previous studies revealed that TiO 2 can efectively decompose 2,4-D in water [46,47]. Te decomposition efciency may vary with the TiO 2 composition; pure anatase TiO 2 can remove 68.2-70.5%, but it increased to 92.7% as TiO 2 containing 8% rutile [46]. Obviously, TiO 2 powder can efectively remove 2,4-D; however, it could not be directly dispersed into water resources since it will require very large amount and could not be recovered. TiO 2 powder, consequently, is not feasible for the direct use to treat organic contaminants in large water reservoirs. Tus, the fabrication of FPC could be a greener and more feasible measure for the removal of 2,4-D and other organic pollutants from water using TiO 2 . To the best of our knowledge, the application of FPC for the degradation of 2,4-D has not been investigated; therefore, the objective of this work is to graft TiO 2 onto LFC for TiO 2 -based FPC production and investigate its potential for 2,4-D removal.  Table 1 were purchased from a local supplier. Rice husk was collected from a local source in Hai Duong, Viet Nam. Clay samples were dried and ground to the particle size ≤63 μm while rice husk was crushed until the size of ≤0.5 mm. Tetraisopropyl orthotitanate (TTIP, 97%), isopropanol (IPA, 99.5%), acetyl acetone (ACAC, 99%), and 2,4-dichlorophenoxyacetic acid (2,4-D) were purchased from Sigma-Aldrich and used without further purifcation.

Preparation of Lightweight Fired
Clay. LFC was prepared in accordance with previous publications [48,49]. In a typical preparation process, desired amounts of clay and rice husk with a mass ratio of 1 : 1 were weighed and mixed well prior to the addition of water. Te water quantity was sufciently adjusted to ensure the plasticity of the clay mixture. Te clay mixture was pelletized into spherical-like granules, which were then dried under sunlight for 2-3 days before being fred in a furnace. Te fring process was conducted in two steps from the room temperature to 1200°C. Te frst step related to the temperature increment from the room temperature to 200°C at the ramping rate of 15°C/min and then to 1200°C at the ramping rate of 20°C/ min. Te temperature remained constant for 20 min and 10 min at the end of the frst and second steps, respectively. After cooling down to room temperature, the LFC sample was stored for further characterization and experiments.
2.3. TiO 2 Synthesis. TiO 2 was prepared by a hydrothermal method adapted from [50] using tetra-isopropyl orthotitanate as a titanium precursor. Typically, a mixture of TTIP : ACA : IPA with a molar ratio of 1 : 1 : 30 was prepared by slow addition of TTIP into a 500 mL beaker containing ACA and IPA, followed by the introduction of a solution of 15 wt % water in IPA. Te mixture was continuously stirred at room temperature for 30 min, transferred to a 500 mL hydrothermal reactor made of Tefon-lined stainless steel. Te mixture was then hydrothermally treated by placing the reactor in an oven at 160°C for 9 h. Solid TiO 2 was separated and washed with plenty of ethanol and water by centrifugation. Te obtained TiO 2 was dried at 90°C for 24 h for later characterization and fabrication of FPC.

Preparation of Floating
Photocatalyst. FPC that is TiO 2modifed lightweight fred clay (TiO 2 /LFC) was prepared according to a procedure described elsewhere [39]. First, 5 g TiO 2 was dispersed into 150 mL ethanol in a 500 mL beaker, followed by the adjustment of pH to ∼3.5 with dilute HNO 3 . Te mixture was sonicated for 30 min to generate a homogenous slurry, which was then gently mixed with 20 g of

Photocatalytic Degradation Tests.
Photocatalytic degradation was tested under UV light by a batch-wise method in an experimental chamber consists of a 6-place magnetic stirrer at the bottom and 10 fuorescent UV lamps (G8 W T5 from Sylvania producer with λ max � 365 nm to 8 watt) mounted on the top. Over the magnetic stirrer, the energy density of 6.5 mW/cm 2 was determined by using a UVA-B light meter and an ILT 1400-A Radiometer Photometer. Te chamber was constructed mainly by aluminium material and was completely covered by aluminium foil during testing. A similar experimental setup was used for the sunlight test; however, the chamber with UV light was removed for sunlight irradiation.
In a typical experiment, 0.5 g FPC and 50 mL of 2,4-D 0.1 mM solution (catalyst dose: 10 g/L) were added into a 250 mL beaker, stirred on a magnetic stirrer in the experimental chamber, and then the UV light was turned on. To follow the degradation progress, samples were extracted after a certain duration, fltered, and analyzed for 2,4-D concentration. Two FPC granule samples with average sizes of 5 mm and 8 mm were tested to evaluate the potential efect of granular size on its catalytic degradation activity. Blank and control experiments were conducted in the same procedure with no catalyst, pure TiO 2 (0.6 g/L) or LFC substrate (10 g/L). To study the recyclability of photocatalyst, FPC (8 mm) was recovered after the experiment, slightly washed with water, dried at 120°C for 2 h, and then reused in another cycle to examine any possible decrease in photocatalytic activities.
To conduct radical scavenging experiments, three radical scavengers, i.e., benzoquinone, EDTA, and isopropanol were captured to capture O − 2 , h + , and OH, respectively. Accordingly, each scavenger was added to a beaker containing 2,4-D solution with FPC (8 mm), which was then placed in a UV chamber for 300 min and samples were taken for analyses. To further confrm the 2,4-D degradation, experiments were conducted 2,4-D solution (5 ppm) without scavenger and samples were collected for the analyses of total organic carbon.

Characterization.
Te specifc surface areas of samples were analyzed by nitrogen adsorption/desorption method using Micromeritics TriStar II Plus. Samples were degassed at 250°C for 5 h prior to analysis and the surface area was determined by the BET method. X-ray difraction (XRD) patterns were collected on the XRD D8 Advance Bruker using a Cu-Kα source. Scanning electron microscopy (SEM) images were observed on the JEOL 7500F coupled with energy-dispersive X-ray spectroscopy. Fourier transformer infrared spectroscopic studies were conducted on the FTIR 6300 spectrometer (Jasco). 2,4-D concentration was analyzed on an HPLC 5890 series II, Shimadzu using a UV detector at 285 nm, a Zipax SAX (duPont) C18 column, and solvent system including CH 3 CN (A, 60%) and H 2 O with 0.15% acetic acid (B, 40%) at a fow rate of 1 ml/min and an injection volume of 20 μl. Total organic carbon was analyzed on TOC Veolia/Suez Sievers M5310C Laboratory.

Results and Discussion
3.1. Material Characterization. TiO 2 photocatalyst and LFC foating substrate were prepared separately, and then TiO 2 was grafted on the LFC surface by an adsorption-calcination procedure without the addition of any binder. TiO 2 was prepared by a hydrothermal technique using TTIP as a titanium precursor. Tis method allows one to synthesize anatase or anatase/rutile mixed TiO 2 particles at relatively mild condition [50,51]. Te coexistence of the anatase/rutile phase reduces the band gap that enhances the photocatalytic activity of TiO 2 in the range of visible light [52,53]. After hydrothermal treatment, TiO 2 nanoparticles were obtained with the particle size ranging from 10 to 45 nm. Tese particles tended to agglomerate into mesoporous powder, as shown in Figure 1 Representative SEM images of the LFC surface are shown in Figure 1(b). LFC has a porous structure in which large pores can reach a size of ≈100 μm. Its highly porous structure gives it a low bulk density (<1 g/cm 3 ). Higher magnifcation (Figure 1(b) inset) revealed that LFC constitutes of laminar structure of silicate that were interconnected into a highly porous network, similar observation in previous studies [54,55]. Tis type of materials shows relatively good adsorption performance [54,56]. Tus, the LFC surface was almost completely covered by TiO 2 nanoparticles as soon as it was contacted with TiO 2 slurry (Figure 1(c)). After calcination at, more and larger cracks appeared on the surface of the layer; however, the microstructure of TiO 2 was unchanged (Figure 1(d)). Te interconnected TiO 2 nanoparticles percolated into pores and were deposited onto the LFC surface to form a porous layer. Te degree of TiO 2 nanoparticle aggregation in the porous layer looks similar to that in the original TiO 2 powder. Tis restricted the accessibility to pores in the LFC structure, which resulted in a signifcant reduction in the surface area of LFC from 37.7 to 1.2 m 2 /g.
Te presence of TiO 2 on LFC was asserted by the XRD study, as shown in (Figure 2). Te XRD pattern of LFC exhibited peaks at 20.6, 26.5, 36.5, and 40.2°, which could be attributed to the difraction of quartz. Difraction at 30.9 and 40.8°corresponds to the mullite phase, which was upon the calcination. Most of these peaks decreased when LFC adsorbed TiO 2 slurry and calcined, except for the peak at 26.5°that belongs to the stable quartz phase. In addition, a novel and distinct peak emerged at 25.3°and some minor peaks at 37.9, and 48.1°that could be assigned to the characteristic difraction of anatase TiO 2 . Tis indicated that the TiO 2 was successfully grafted onto the LFC. Moreover, several additional minor peaks were observed at 27.7°on the calcined samples, which suggest the possible transformation of anatase to rutile TiO 2 during calcination.
Te addition of TiO 2 onto LFC was further observed on the FTIR spectra of the samples. As shown in Figure 3   layer on the surface of the LFC. Elemental analyses by EDX indicated that the content of Ti increased from 0.82 wt% to 6.86 wt% after TiO 2 was grafted on the LFC surface (Figures 4(a) and 4(b)). Elemental mapping analyses revealed that TiO 2 distributed throughout the surface of the LFC substrate (Figures 4(c) and 4(d)). Tis is very meaningful to a foating catalyst that helps the catalyst stay active irrespective of the catalyst surface that receives the sunlight. Figure 5. A negligible decrease in 2,4-D concentration was detected after 250 min UV irradiation without catalyst. Te test with LFC substrate showed a 4% reduction in the frst 30 min and after that no considerable change was recorded. Tese suggested that the photolysis of 2,4-D occurred at a relatively slow rate and that the reduction in the presence of LFC substrate was due to its adsorption on LFC. Adsorption was also observed on FPC as the tests were conducted in the dark with a 4.5% and 9.6% reduction in 2,4-D concentration after 30 and 120 min, respectively. Te adsorption of 2,4-D has very important role in the performance of FPC. Tis allows FPC to continuously attract pollutants from the water volume onto its surface for photocatalytic decomposition while foating on the surface without vigorous mixing.

Photocatalytic Degradation toward 2,4-D. 2,4-D degradation efciency by photocatalysts is exhibited in
Te degradation efciency increased sharply and reached 79.91% in the initial stage of 60 min as TiO 2 powder was used. Te degradation occurred at slower rate in the later stage and reached 99.87% after 250 min. Te slow degradation in the later stage is mostly due to the low 2,4-D concentration remained in the solution. Tis result revealed that the synthesized TiO 2 efectively decomposed the 2,4-D under UV radiation. Te photocatalytic degradation was sustained as TiO 2 was grafted onto the foating structure of the LFC, however, at a slower rate. As seen in Figure 5(a), the 2,4-D removal efciency reached only 21.7% after 60 min and 60.4% in 250 min. In this study, the quantity of TiO 2 in FPC used (0.057 g) is almost double that of TiO 2 powder (0.03 g), thus, the slow degradation rate is likely due to the less accessibility to photocatalytic sites in the foating catalyst compared with the TiO 2 powder. As added into water, TiO 2 particles in powder form can disperse throughout the water phase under mixing condition, thereby, 2,4-D can approach to TiO 2 particles instantly and then easily decomposed as TiO 2 particles are exposed to UV light. Meanwhile, the foating catalyst appears on water surface only, it takes time for 2,4-D molecules to migrate from bulk water to the surface of catalyst. Tis migration induces by 2,4-D concentration gradient and is rate-limiting process. Te migration rate could be enhanced by the application of external forces, i.e., stirring or air bubbling; however, it could not be occurred instantly because of the long distance. Moreover, as FPC granules foat on water, only about half of their surface area exposes to the light, which further limits the activity of foating catalyst. Even though the removal efciency achieved by FPC was lower than that achieved by TiO 2 powder, it could be used to develop a sustainable water treatment technology. Tis method could considerably reduce the risk of secondary contamination and be particularly suitable for the treatment of large water resources, aquaculture, and agriculture water.
2,4-D degradation kinetics was investigated using the pseudo-frst-order kinetic model, as given in the following equation: where r is reaction rate, C is 2,4-D concentration, t is reaction time, and k is pseudo-frst-order rate constant. Solving equation (1) with the boundary conditions of t � 0, C t � C 0 , an integration form was obtained as the following equation: Te rate constant, k, can be determined by a linear plot of Ln (C t /C 0 ) vs. time (t), as shown in Figure 5(b). R square and k values received from linear ftting are exhibited in Table 2. Rate constants were very small, only 1.98 × 10 −6 min −1 and 1.34 × 10 −4 min −1 , in the case no catalyst and LFC were used in the experiments, respectively. Besides, the regression is very bad for those two cases with the correlation coefcients (R 2 ) are −0.1967 and 0.32 only. Meanwhile, the degradation rate constant for TiO 2 powder was relatively high, reached 0.023 min −1 with R 2 of 0.9589. Rate constants were 0.0036 min −1 and 0.0038 min −1 for FPC with granular sizes of 5 mm and 8 mm, respectively. Te very close rate constants revealed that the variation in granule size from 5 to 8 mm caused no signifcant infuence on their catalytic effciency. Te correlation coefcients reached 0.9878 and 0.9967 for FPC with granular sizes of 5 mm and 8 mm, respectively, indicated that the 2,4-D degradation on FPC fts well to the pseudo-frst-order kinetic model. Simulation on 2,4-D degradation efciency vs. time based on the pseudo-frst-order kinetic model is presented in Figure 6(a). 2,4-D degradation trend resulted from the model is correlated well with that obtained from experiment. Accordingly, 90% of 2,4-D is expected to be decomposed in 640 min, equivalent to less than two sunny days depending on the location.
For large water resources such as agriculture, aquaculture, or reserve water resources, they may not require a signifcantly rapid treatment but rather a sustainable treatment method, and therefore, the application of the FPC could become suitable. However, to apply for this purpose, FPC must be active under the sunlight instead of UV light in the laboratory. In a previous work conducted by Shavisi et al., a foating catalyst based on P25 TiO 2 grafted lightweight expanded clay aggregates proved to efcient candidate for NH 4 + degradation under solar radiation with 96.5% NH 4 + removal [39]. By grafting TiO 2 synthesized from TTIP by the sol-gel method on palm trunk, Sboui et al. received a foating catalyst that can remove 98.2% Congo red after 210 min under solar radiation [38]. Several others demonstrated that the efcient degradation of organic compounds under sunlight can be achieved by grafting TiO 2 on a foating Journal of Analytical Methods in Chemistry substrate for foating catalyst production [33, 35-37, 41, 44, 57].
To investigate the catalytic activity of FPC prepared in this study under sunlight, experiments has been conducted in the same protocol in laboratory except the light source was changed to natural sunlight with the measured radiation power of 6.71 mW/m 2 . Te result revealed that over 50% of 2,4-D was decomposed after 250 min. To evaluate the recyclability, photocatalysts were recovered, slightly washed, and dried before dispersing on water for another testing cycle. Te performance of photocatalysts was assessed based on the change in degradation efciency against 2,4-D after each cycle. Results obtained revealed that a negligible reduction in degradation efciency (∼7.2%) was observed after 5 cycles (Figure 6(b)). Tis indicated that FPC is stable in experimental conditions in the laboratory. Tis work    provided solid evidence to further confrm that FPWT could become a promising technology for water treatment.
It is well known that the radicals such as OH, O − 2 , h + generated during UV light irradiation are responsible for the photodegradation of 2,4-D. To elucidate the role of those radicals on the photodegradation, radical scavengers-benzoquinone, EDTA, and isopropanol-were used as scavengers to capture O − 2 , h + , and OH, respectively. Experimental results showed that 2,4-D degradation efciency slightly changes when benzoquinone was added, while the efect was signifcant as EDTA and isopropanol were used (Figure 7(a)). Te 2,4-D degradation efciency was reduced from 66% to 53.3% and 47.3% with the addition of EDTA and isopropanol, respectively. Tese results imply that h + and OH radicals are the most infuential radical on the 2,4-D degradation. Tis observation is in good agreement with a previous study where the contribution of •OH is dominant after 50 min irradiation on TiO 2 /activated carbon system [21]. Tis suggests a mechanism for 2,4-D degradation over FPC, as described in equations (3)- (9). Under UV light, TiO 2 generates electrons and holes, which subsequently react with H 2 O and O 2 , to produce •OH and •O − 2 radicals. Te radicals and h + can oxidize 2,4-D molecules. Total organic carbon contents in the samples decreased signifcantly after UV irradiation (Figure 7(b)) indicating that 2,4-D was mineralized to CO 2 and H 2 O.

Conclusion
A foating photocatalyst was successfully prepared by grafting TiO 2 on lightweight fred clay. In this study, TiO 2 nanoparticles were synthesized from TTIP by a hydrothermal process and grafted on foating substrate by an adsorption/calcination method. Te resulting foating photocatalyst showed a great catalytic activity against 2,4-D with the degradation efciency of 60% and 50% in 250 min under a UV radiation and sunlight, respectively. Te photocatalytic degradation of 2,4-D on the foating catalyst is ftted well to the pseudo-frst-order kinetic model with  Journal of Analytical Methods in Chemistry correlation coefcient (R 2 ) of 0.9878 and 0.9967 and rate constant (k) of 0.0036 min −1 and 0.0038 min −1 for catalyst granule with size 5 mm and 8 mm, respectively. Calculation based on the pseudo-frst-order kinetic model indicated that 90% of 2,4-D can be treated within two days by sunlight using the foating catalyst. Tis demonstrated that a foating photocatalyst-based water treatment technology could be a feasible technology to degrade 2,4-D in water. Te success of the work recommended a green approach to treat organic contaminants in large water resources, where they could not be treated efciently by conventional technologies.

Data Availability
Te data used to support the fndings of this study are available within the article.

Conflicts of Interest
Te authors declare that there are no conficts of interest regarding the publication of this paper.

Authors' Contributions
Hoang Hiep conceptualized the study, provided project administration and funding acquisition, revised the manuscript, and contributed to the fnal approval of the manuscript. Pham Tuan Anh performed experimental design, performed experimental conduction, contributed to data acquisition, and revised the manuscript. Dao Van-Duong contributed to work design, performed data interpretation, devised the work, and contributed to the fnal approval of the manuscript. Dang Viet Quang conceptualized the study, performed experimental design, contributed to data analysis, contributed to frst draft, and contributed to the fnal approval of the manuscript.