Simultaneous Degradation of Hexazinone and Diuron Herbicides by H 2 O 2 / UV and Toxicity Assessment

Diuron (DR) e hexazinona (HX) são agrotóxicos da classe dos herbicidas muito utilizados na agricultura, sendo que no Brasil, a sua formulação mista é utilizada principalmente na cultura de cana-de-açúcar. Esses compostos são tóxicos aos organismos aquáticos, sendo potencialmente cancerígenos. Os processos oxidativos avançados (AOP) são uma alternativa para o tratamento de DR/HX em ambientes aquosos. Neste estudo, avaliou-se a degradação simultânea de DR/HX via H2O2/UV e fotólise direta utilizando um planejamento experimental do tipo composto central. As concentrações iniciais de HX e DR foram 7 e 20 mg L-1, respectivamente. No sistema, o planejamento indicou que a concentração de H2O2 tem maior influência do que o pH. As condições ótimas de degradação (7 mmol L-1 de H2O2 e pH 2,8) proporcionaram uma remoção de carbono orgânico total de 96,4%, enquanto que o processo de fotólise direta removeu apenas 17,2%. Análises cromatográficas indicaram a remoção completa dos dois agrotóxicos a partir de 2 min de reação, o que impossibilitou a diferenciação da cinética de degradação entre DR e HX. Após o tratamento, a toxicidade foi testada utilizando bactérias Vibrio fischeri bioluminescentes, com uma diminuição com a utilização de H2O2/UV. A degradação via H2O2/UV foi empregada com sucesso, mostrando excelente desempenho devido ao aumento da taxa de mineralização.


Introduction
Cultivation of sugarcane in Brazil, specifically in São Paulo state, 1 the bioethanol production has been related with the high demand for renewable fuels economically and environmentally sustainable in domestic and foreign markets.
As a consequence of this intense agricultural activity, the groundwater and surface water are contaminated due to diffuse sources arising from the intense use of fertilizers and pesticides.Diuron (DR), (3-(3,4-dichlorophenyl)-1,1dimethylurea), and hexazinone (HX), 3-cyclohexyl-6dimethylamino-1-methyl-1,3,5-triazine-2,4-dione, are the most potent and known herbicides applied in various stages of the sugarcane production. 2 Vol. 25, No. 11, 2014   DR exhibits moderate water solubility (42 mg L -1 at 20 °C), therefore, it is highly persistent (one month to one year) and can contaminate diverse environments such as soil, sediments and water.According to the Environmental Protection Agency of the United States (USA-EPA), DR is in the list of carcinogenic contaminants to humans. 3HX is slightly soluble in water (33 mg L -1 ) and highly mobile in soil. 4Although it is not highly toxic to humans, it causes irritation to the eyes, nose and throat. 5Recently, HX has been detected in cane sugar plantation areas in Australia, in concentrations up to 5 µg L -1 . 6It presents an effective concentration (EC 50 ) of 3.0 to 3.6 µg L -1 . 7In surface water, it is toxic to primary producers affecting the reducing food availability. 8In this context, the detection and degradation of theses contaminants have been the subject of many studies.Magnusson et al., 9 tested the toxicity of many herbicides using tropical microalgae (Navicula sp.Cylindrotheca and closterium (Ochorophyta) and Nephroselms pyriformis (Chlorophyta)) which DR and HX have been showed the most toxic.According to Chen et al., 10 DR has been detected in the water in California, USA.The study suggests that DR may be a precursor for the formation of carcinogen compound, nitrosodimethylamine (NDMA).
The conventional wastewater treatment is not effective for the complete degradation of herbicides.In this context, the advanced oxidation processes (AOP's) have been developed to degrade these refractory contaminants in surface water and industrial effluents.The AOP's are constituted by a combination of oxidizing agents such as UV radiation and hydrogen peroxide (H 2 O 2 ). 11H 2 O 2 is a strong oxidizing agent (oxidation potential of 1.8 V and 0.8 V, 14 and 0 pH's, respectively). 12However, hydrogen peroxide alone is not strong enough to oxidize most of the herbicides.When combined with other physical and chemical agents, such as ozone, ferric ions and UV radiation, the formation of radicals is facilitated.][16] In comparison to Cl 2 and O 3 , the use of H 2 O 2 as the oxidizing agent presents advantages, such as commercial feasibility, thermal stability, high water solubility, absence of problems related to mass transfer and no formation of halogenated hydrocarbons and bromide ions. 17,18Regarding effectiveness, an important requirement of AOP's is evaluate the presence of coproducts and derivatives compounds from pollutants that have biological activity and toxicity.According to Fatta-Kassinos et al., 19 only few studies have dealt with this topic.
The present paper reports on a study of the simultaneous degradation of DR and HX herbicides by H 2 O 2 /UV process and photolysis.The efficiencies of these processes were evaluated by high-pressure liquid chromatography (HPLC), ion chromatography (IC) and total organic carbon (TOC) assay.The influence of pH and H 2 O 2 concentration was studied using a central composite design.The toxicity was measured using a LUMIStox test using bioluminescent bacteria Vibrio fischeri (V.fischeri).

Experimental
Materials DR and HX (analytical standard; 99.5% and 99.9% purity, respectively) were used for calibration curves and were obtained from Sigma-Aldrich (St. Louis, MO, USA; product number 330-54-1 and 51235-04-2).The mixed formulation (JUMP; Milenia Agrociências SA, Londrina, PR, Brazil) containing 53.3% w/w of DR and 6.7% w/w of HX was obtained from an agricultural store and was used for experiments degradation.The mixed formulation was used in the experiments due to representation of applications in agricultural cultures and the influence of inert compounds in the degradation.Methanol, acetonitrile and sulphuric acid were commercially obtained from Mallinckrodt (Xalostoc, Edomex., Mexico).Sodium sulphite and ammonium metavanadate were acquired from Sigma Aldrich (St. Louis, MO, USA) and the 30% (w/w) solution of hydrogen peroxide (reagent grade) was from Ecibra (São Paulo, SP, Brazil).Sodium carbonate and sodium bicarbonate were acquired from JT Baker (Phillipsburg, NJ).Purified water (18.2MΩ cm resistivity; 0.039 mg C L -1 ) was prepared using a Millipore (Eschborn, Germany) Milli-Q water purification system.
The photodegradation experiments were carried out on a laboratory scale using a UV photoreactor (Figure 1) thermostatically controlled (25 °C) and irradiated by a Philips (Amsterdam, The Netherlands) 125 W UVC lamp (254 nm emission peak, 22,874.83mW m -2 ).In each experiment, the reactor was filled with 0.2 L of the test solution and operated with a constant magnetic stirring.

Evaluation of pH and H 2 O 2 concentration in the degradation process
The levels of hydrogen peroxide and pH required for the most efficient H 2 O 2 /UV degradation of DR and HX herbicides were determined using a central composite design (CCD).Experimental calculations design were performed using Matlab 2011a software (Mathworks, Natick, MA, USA).The central point concentrations for CCD, represented by the coded values 0, were established as 7 for pH and 7 mmol L -1 for hydrogen peroxide.The coded values (Cv) for further levels of the two independent variables tested (Table 1) were obtained by the equation: where Rv is the real value, Rv +1 is the real value for +1 level and Rv -1 is the real value for -1 level.The equation that relates the content of total organic carbon removed with the process parameters (pH and H 2 O 2 concentration) was obtained by a linear regression between the parameters of the experiments and the experimental responses.The response surfaces were constructed in order to achieve the best degradation performance.The statistical model was evaluated by analysis of variance (ANOVA) at 95% confidence level.

Analytical procedures
The DR and HX concentrations were determined by HPLC using a Shimadzu (Kyoto, Japan) Prominence LC 20 AT modular system comprising two CBM-20 A pumps, a CTO-10AS oven, an SIL 20A autosampler, an SPD-20A variable wavelength detector and an LC-10 Workstation Class data processor.Separations were carried out on a Supelco (Bellefonte, PA, USA) Supelcosil C-18 column (250 mm × 4.6 mm i.d.; 5 µm), protected by a Supelcosil C-18 column guard column (20 mm × 4.6 mm i.d.; 5 µm) and eluted with mixtures of water:methanol (70:30).The chromatographic conditions were oven temperature of 35 °C, flow rate of 0.8 mL min −1 ; injection volume of 20 µL (Rheodyne loop); and UV detection at 254 nm (retention time 5.7 and 7.4 min for HX and DR, respectively).
A concentration of inorganic ions formed during degradation was detected by IC using a Metrohm model 850 Pro-IC unit combined with a conductivity detector and fitted with a Metrosep A Supp 5 column.The chromatographic conditions were mobile phase-aqueous solution of sodium carbonate 3.2 × 10 -3 mol L -1 and sodium bicarbonate 1.0 × 10 -3 mol L -1 and elution-isocratic at a flow rate of 0.7 mL min -1 .
The decrease in the organic material during chemical degradation was monitored by a Shimadzu PC-Controlled TOC Analyzer model TOC-VCPN.The hydrogen peroxide consumed was estimated by absorption at 450 nm (Cary-50 Scan UV-VIS spectrophotometer; Varian Inc, Lake Forest, USA) following the addition of ammonium metavanadate to the reaction mixture. 22he remaining H 2 O 2 was removed by the addition of sodium bisulfite (0.1 g) immediately after collection sample.For monitoring, samples were filtered on 0.45 µm cartridges and immediately analysed the DR, HX concentrations, total organic carbon and inorganic ions.A toxicity bioassay on bacteria luminescence was carried out with a LUMIStox 300 (Dr.Lange, Duesseldorf, Germany).Tests were performed using gram negative marine bioluminescent bacteria of the V. fischeri (GLX8400 lyo 5) species.The samples were treated with a NaCl solution of 20 g L -1 and brought to 50 mS cm -1 conductivity before analysis.Starting from the concentration of the sample, eight consecutive dilutions were tested (dilution factor 1:2); the inhibition of bioluminescence was measured at a wavelength of 490 nm, with readings after 5 and 15 min of incubation at 15 °C.

Direct UV photolysis
The UV radiation promotes the degradation of photolabile organic compounds by direct incidence. 23,24owever, few photolysis studies have results obtained with mixtures of xenobiotics wherein the competition effects can alter the rate of degradation. 25The assay performed in the photoreactor (Figure 1) by incidence UV radiation through a Hg-vapor lamp demonstrated that DR was rapidly degraded while the HX proved to be more resistant to photolysis. Figure 2 shows concentration of mixed formulation containing the DR and HX during the photodegradation process.The first one shows falling detection at limit of the HPLC/UV after 8 min and the HX curve has the same aspect with less pronounced delay compromising a considerable degradation.As long the herbicides are photodegraded, it is observed in the chromatograms (Figure 3) the formation of intermediates.In addition, the intermediates formation is also observed by the low total organic carbon removal, only 17.2%.This information may be related with the chemical composition of mixed, which suggests the presence of inert and recalcitrant compounds at the end of the process.Considering the high efficiency degradation during 30 min, comprised by the total removal of DR and HX (Figure 3), it is expected a low kinetic after this time, due to low concentration of remaining degradable compounds.Thus, at the end of the experiment, intermediates still are detected but in reduced concentration.Similar results for DR photodegradation are also described by Sanchez et al.. 26 With respect to hydrolisys, the DR and HX, do not suffer degradation during the 30-minute experiment.
The central composite design was performed for the evaluation of the influence of H 2 O 2 concentration and pH during the mixed formulation degradation.In most experiments, the DR and HX concentrations were below the detection limit of the HPLC/UV at the end of the 30 min of reaction.From these results, only TOC percentage removal could be modeled.The experimental conditions for all experiments and the corresponding results are showed in Table 1.The experimental data were analyzed for the determination of the predictive quadratic model which describes the response TOC (y) as a function of process parameters H 2 O 2 (x 1 ) and pH (x 2 ).The quadratic model is shown in equation 2 and the standard deviations of the coefficients are given in brackets.
(R 2 = 0.61) (2)   Due to the lack of fit of the quadratic model, two more cubic terms were added to the model.Equation 3 describes this model.
As a result of the lack of fit, the quadratic and cubic models failed to predict the complex surface of empirical answers.The complexity of the empirical response surface is related to a simultaneous degradation of two herbicides, which each herbicide respond differently to experimental condition of degradation.They have different degradation kinetics, making this reaction difficult to model.Extra experiments planning (Table 1, experiments 10-13) were performed to improve the fit of the models, but no satisfactory results were obtained.Within the experimental domain, it is possible to observe an irregular influence of the concentration of hydrogen peroxide only at -1.41 and -0.4 (0.66 and 5.2 mmol L -1 H 2 O 2 ) presented TOC percentage removal below 81% (experiments 6 and 12 of Table 1, respectively).Concerning the pH, the experiments indicate that under high alkalinity, the process loses efficiency.This fact was confirmed by extra experiments planning (experiments 10 and 12), which was observed a 30% drop in efficiency degradation when the concentration of H 2 O 2 remained constant and the pH was changed from 2.8 to 11.2.These results are in agreement with that provided by Catalkaya et al., 27 who obtained an increase of 87% to 97% removal of DR when the pH was reduced from 11 to 3 units.
Figure 4 shows the empirical results as function of pH and H 2 O 2 concentration.It also shows the existence of several maxima and minima within the experimental domain (indicated by the double arrows), justifying the lack of fit of the models.The planning indicates an optimum condition for degradation at 7 mmol L -1 H 2 O 2 and pH 2.8, corresponding to 96.4% removal of total organic carbon (Figure 5a), and this value is obtained within 30 min of degradation, comprising the high efficiency of process.Hydrogen peroxide was completely consumed within 50% of the total reaction time (Figure 5b).After 8 min of treatment, DR and HX were not detected by HPLC/UV (Figure 6).The inorganic ions formation comprises the degradation of herbicides.In this respect, at the end of the process, it was observed 8.1 mg L -1 of chloride, 2.1 mg L -1 of nitrate and 0.49 mg L -1 of nitrite, which approximate to the theoretical values (5.21 mg L -1 of nitrogen and 9.89 mg L -1 of chloride).These agreements show the high mineralization of herbicides and no formation   of organochlorine and nitrogen intermediates.Thus, these results suggests that after 30 min, the degradation reaction will follow a low kinetic that will results in the small variation of overall TOC.
In order to evaluate the biological potency and toxicity of coproducts formed during the H 2 O 2 /UV process, the V. fischeri test was performed.The assays were conducted at the initial and finished stages of the optimum condition for degradation.The initial sample presented effective concentration to 20% of test organisms (EC 20 ) of 33% and the sample collected after treatment presented EC 20 41%.The increase in EC 20 after treatment points indicated the reduction of toxicity.The H 2 O 2 /UV process promotes a decrease in toxicity according to the mineralization of DR and HX.Noteworthy, the present work shows more promising than related studies of area involving degradation of herbicides HX and DR.

Conclusions
The DR and HX are potent and toxic herbicides applied to diverse cropping, mainly sugar cane cultivation.Concerning effluents treating methods, we have demonstrated the important applicability of APO's to DR/HX simultaneous degradation.The H 2 O 2 /UV method has proved more efficient than DR/HX photolysis degradation.For the H 2 O 2 /UV, the removal of organic carbon rate for 30 min of degradation was 96.4% while for the photolysis had less efficacy and accuracy (17.2%).This fact is related to the formation of more polar byproducts than DR/HX herbicides during the photolysis degradation.The planning to H 2 O 2 /UV shows that the quadratic and cubic models failed to predict the complex response surface in the empirical process.This feature is explained by overlapping degradation mechanisms of herbicides, as consequence a highly complex surface can not be requested.However, an optimum condition could be found for the simultaneous degradation (7 mmol L -1 H 2 O 2 and pH 2.8).In addition, the ion chromatography showed the formation of nitrate, chloride and nitrite after the process, comprising the molecules breaking.The acute toxicity assay using a microbe V. fischeri showed a toxicity decrease in both methods.All these results indicate that H 2 O 2 /UV methods can be utilized for an efficient decontamination of wastewaters containing DR/HX.Furthermore, H 2 O 2 /UV process was more effective than only the photolysis process.Considering the radiation use and the experimental acid condition, the process may be an economic and viable method of herbicide removal, since it does not generate subproducts with the effective degradation of HX and DR without phase transference of contaminants.Therefore, it appears to be a promising technology for the removal of aqueous herbicides.

Figure 2 .
Figure 2. Monitoring of herbicides HX and DR during photolysis using the mixed formulation.

Figure 3 .
Figure 3. Chromatograms obtained from the analysis of HPLC/UV during the simultaneous photolysis of DR and HX using mixed formulation.

Figure 4 :
Figure 4: Empirical results from the central composite design, whose independent variables were H 2 O 2 (mmol L -1 ) and pH.

.
Chromatogram obtained by the analysis of HPLC/UV during the H 2 O 2 /UV process (experiment 8, Table1).

Table 1 .
Degradation of diuron and hexazinone and the removal of total organic carbon by H 2 O 2 /UV reaction in central composite design experiments in which hydrogen peroxide (x 1 ) and pH (x 2 ) were independent variable a Experiments extra planning.