Fructose As a Novel Photosensitizer: Characterization of Reactive Oxygen Species and an Application in Degradation of Diuron and Chlorpyrifos

Chemosphere 144 (2016) 1690e1697

Contents lists available at ScienceDirect

Chemosphere

journal homepage: www.elsevier.com/locate/chemosphere

Fructose as a novel photosensitizer: Characterization of reactive oxygen species and an application in degradation of diuron and chlorpyrifos

* Shaila Nayak a, Juan Muniz b, Christopher M. Sales c, Rohan V. Tikekar d, a Program in Culinary and Food Science, Drexel University, Philadelphia, PA, USA b Department of Nutrition Sciences, Drexel University, Philadelphia, PA, USA c Civil, Architectural and Environmental Engineering, Drexel University, Philadelphia, PA, USA d Department of Nutrition and Food Science, University of Maryland-College Park, MD, USA highlights graphical abstract

Fructose underwent photolysis upon exposure to 254 nm UV light. Photolysis products include hydrogen peroxide, singlet oxygen and hydroxyl radicals. Fructose accelerated UV induced photocatalytic degradation of diuron and chlorpyrifos. Fructose is an attractive photosensitizer for decontamination of wastewater. article info abstract

Article history: The objective of this study was to identify reactive oxygen species (ROS) generated from the exposure of Received 15 June 2015 fructose solution to the 254 nm ultraviolet (UV) light and evaluate whether fructose can be used as a Received in revised form photosensitizer for accelerated photo-degradation of diuron and chlorpyrifos. We demonstrated that 25 September 2015 hydrogen peroxide, singlet oxygen (1O ) and acidic photolysis products were generated upon UV Accepted 18 October 2015 2 exposure of fructose. Consistent with these ﬁndings, UV induced degradation of chlorpyrifos and diuron Available online xxx was accelerated by the presence of 500 mM fructose. The average ﬁrst order photo-degradation rate constants in the absence and presence of 500 mM fructose were 0.92 and 2.07 min 1 respectively for Keywords: 1 ɸ Fructose diuron and 0.04 and 0.07 min for chlorpyrifos. The quantum yields ( ) for direct photo-degradation of Advanced oxidative process diuron and chlorpyrifos were 0.003 and 0.001 respectively. In the presence of 500 mM fructose, these Ultraviolet light values increased to 0.006 and 0.002 respectively. Thus, fructose may be an effective photosensitizer. Chlorpyrifos © 2015 Elsevier Ltd. All rights reserved. Diuron Photosensitizer

1. Introduction

Fructose has shown significant reactivity upon exposure to 254 nm UV light. A previous study reported that UV (254 nm) exposure of fructose present in fruit juices results in the formation * Corresponding author. Department of Nutrition and Food Science, 112 Skinner of furan (Bule et al., 2010). In previous studies, we demonstrated Building, College Park, MD 20742, USA. that fructose accelerates the degradation rates of patulin and E-mail address: [email protected] (R.V. Tikekar). http://dx.doi.org/10.1016/j.chemosphere.2015.10.074 0045-6535/© 2015 Elsevier Ltd. All rights reserved. S. Nayak et al. / Chemosphere 144 (2016) 1690e1697 1691 ascorbic acid during UV treatment of apple juice (Tikekar et al., (Pittsburgh, PA). Singlet Oxygen Sensor Green (SOSG) was pur- ® 2012, 2011). One study showed that pH of fructose solution chased from Life Technologies (Carlsbad, CA). decreased by 5.29% upon exposure to high intensity pulsed UV light. The lowering of pH was attributed to products formed due to 2.2. UV processing unit photolysis of fructose when exposed to UV light at 254 nm. How- ever, these products were not characterized (Orlowska et al., 2013). The experiments were carried out in a bench top batch UV In another study, by identifying the stable degradation products processing unit (Spectronics Spectrolinker XL-1500 UV Crosslinker, from the photolysis of fructose, it was hypothesized that the open Westbury, NY). The unit consisted of 5 UV lamps (15 W, Spectronics chain form of fructose (0.8% of total fructose in aqueous solution) Corporation, Westbury, NY) generating light at 254 nm. The total undergoes photolysis when irradiated with 254 nm UV light intensity of light incident on samples which were mounted over a resulting in generation of primary species such as hydroxyalkyl, shielded box (46.4 15.9 31.8 cm) was approximately 14.8 mW/ and hydroxyalkyl acyl radicals. These species, upon interaction with cm2. In order to avoid any variations in the intensity of light, the oxygen were hypothesized to form secondary species such as, hy- lamps were allowed to warm up for 5 min before starting the ex- droxyl, peroxyl, superoxide radicals and hydrogen peroxide periments. The temperature during UV treatment was 20e25 C (Triantaphylides et al., 1982). However, formation of only hydrogen and remained constant throughout the experiment. peroxide was experimentally determined and the kinetics of its formation were not elucidated. In congruence with the above 2.3. Generation of reactive oxygen species (ROS) from UV exposed findings, we previously demonstrated that UV exposure of fructose fructose generates miscellaneous reactive oxygen species (ROS), although we did not characterize the chemical nature of these species Fluorescein is commonly used as a probe to detect ROS, due to (Elsinghorst and Tikekar, 2014). Thus, although the reactivity of its sensitivity to oxidation from diverse oxidative species including fructose in 254 nm UV light is known, scarce information is avail- hydroxyl and peroxyl radicals (Ou et al., 2002, 2001). Test solutions able on the formation of reactive oxidative intermediates such as consisted of 1 mM fluorescein prepared in either distilled water or free radicals upon exposure of fructose to UV light. The one study 100 mM phosphate buffer of pH 6.7. Fructose was added to each of that investigated the photolysis of fructose upon exposure to UV these solutions at the concentrations of 10, 100 and 500 mM. light only hypothesized the formation of these reactive in- Treatments were carried out by exposing 10 mL of each solution to termediates by identifying the degradation products. To address UV light for up to 10 min. To ensure the uniformity of treatment this need, the present study gives direct evidence for formation of within each sample, the solutions were stirred throughout the some of these oxidative species and characterizes the kinetics of experiment. A 200 mL of the sample was obtained periodically and their formation using complimentary techniques. We also aimed to the fluorescence from fluorescein was measured using a Gemini evaluate whether the ability of fructose to generate ROS upon XPS fluorescence micro-plate reader (Molecular Devices, Sunny- exposure to UV light can be harnessed to accelerate photo- vale, CA). The excitation and emission wavelengths were set at degradation of recalcitrant organic compounds. To that end, we 485 nm and 510 nm respectively. Fluorescence data was fitted into investigated photo-degradation of an insecticide (chlorpyrifos) and the first order kinetics. an herbicide (diuron) using fructose. To the best of our knowledge, this is the first study that has reported the ability of fructose to act 2.4. Detection of hydrogen peroxide as a photosensitizer to accelerate the photo-degradation of xeno- biotic compounds. The concentration of hydrogen peroxide generated from UV Chlorpyrifos [O,O-diethyl-O-(3,5,6-trichloro-2-pyridinyl)phos- exposure of fructose was measured using the ferrous oxidation- phorothionate], an organophosphate insecticide and diuron [N0- xylenol orange (FOX) method. The principle of this method in- (3,4-dichlorophenyl)-N,N-dimethylurea] a polyurea herbicide, are volves oxidation of ferrous ions by hydrogen peroxide to ferric ions. declared as emerging contaminants by the US-EPA (Richardson, Ferric ions under acidic conditions form a complex with xylenol 2007). Chlorpyrifos inhibits cholinesterase, an enzyme respon- orange which can be measured through spectrophotometry (Bou sible for normal functioning of the nervous system both in insects et al., 2008). The assay solution consisted of 2 mM xylenol or- and humans (Simon et al., 1998). The herbicide, diuron, exhibits its ange, 2.5 mM ferrous sulfate,1 M sorbitol and 250 mM sulfuric acid. activity by inhibiting photosynthesis in weeds (Fuerst and Norman, A100mL of this assay solution was immediately added to the 700 mL 2011) and has been classified as a potential carcinogen by the US- of test sample consisting of UV exposed 500 mM fructose prepared EPA (Cox, 2003). Photosensitizer-induced degradation of these in distilled water or phosphate buffer. This mixture was incubated compounds has been explored before using diverse photosensi- at room temperature for 30 min with continuous gentle shaking. tizers such as hydrogen peroxide, titanium dioxide, zinc oxide and After incubation, absorbance was measured at 560 nm with a zinc sulfide (Devi et al., 2009; Djebbar et al., 2008; Fenoll et al., UVeVis Spectrophotometer (Molecular Devices, Sunnyvale, CA). 2012). However, this is the first study that involves a novel use of 1 fructose as a photosensitizer compound for the degradation of 2.5. Detection and quantification of singlet oxygen ( O2) pesticide residues. 1 Singlet oxygen ( O2) produced during UV treatment of fructose 2. Materials and methods solution was detected using Singlet oxygen sensor green (SOSG). 1 SOSG can react with O2 to produce SOSG endoperoxides (SOSG-EP) 2.1. Chemicals that emit strong green fluorescence at 531 nm (Lin et al., 2013). A 5 mM SOSG stock solution was prepared in methanol. Test solution Diuron (>98% purity), chlorpyrifos (>99.8% purity), fructose, 30% consisted of 500 mM fructose sample, prepared in phosphate buffer v/v hydrogen peroxide, xylenol orange, ferrous sulfate, sorbitol, of pH 6.7 and mixed with 5 mM SOSG added from stock solution sulfuric acid, ascorbic acid, fluorescein, sodium phosphate mono- immediately before use. This solution was exposed to UV light at basic monohydrate, sodium phosphate dibasic 7-hydrate, titanium 254 nm. Fluorescence was measured at an interval of 1 min of UV dioxide, pesticide grade acetonitrile and iso-octane were purchased exposure with excitation and emission wavelengths of 504 and ® from Sigma Aldrich (Sigma, St. Louis, MO) and Fisher Scientific 525 nm respectively. Furfuryl alcohol (FFA) degradation assay was 1692 S. Nayak et al. / Chemosphere 144 (2016) 1690e1697

1 used for quantitative measurement of O2 generated by UV expo- statistical significance between the samples. sure of fructose. A 40 mM FFA solution was prepared with a potas- sium phosphate buffer (pH 6.7). Test mixtures with fructose 3. Results and discussion concentrations of 0 mM, 300 mM, 500 mM, and 1000 mM were prepared with the 40 mM FFA buffered solution. In addition, 3.1. Generation of ROS from UV exposure of fructose 125 mM of titanium was used as a positive control dissolved in same buffered solution. These mixtures were exposed to UV light Fig. 1a shows the absorbance spectrum of 10% fructose solution. with wavelength of 254 nm for 10 min. One-milliliter samples were The absorbance of fructose solution was low at 254 nm (~0.08) and taken at one-minute intervals in order to measure the concentra- showed a peak at 278 nm that corresponds to the forbidden n to p* tion of FFA (Chinnici et al., 2003). FFA concentration was measured transition of electrons (Triantaphylides et al., 1982). Fig. 2a shows m with a Shimadzu Prominence HPLC, using a C18 column of 5 m the loss of fluorescence of fluorescein in solutions containing thickness and dimensions of 45 mm length 4.6 mm ID (Shi- various amounts of fructose prepared in distilled water and madzu, Columbia, MD), via UVeVis at 215 nm. The isocratic mobile phase flow rate was 0.6 mL/min and was composed of 84% 0.10 N Phosphoric acid and 16% acetonitrile. The column oven and UVeVis detector temperatures were set to 33 C.

2.6. Photochemical degradation of diuron

A stock solution of diuron (1000 mg/L) was prepared in acetonitrile. The treatment solution consisted of 25 mg/L of diuron in di- ionized water incorporated with 0e500 mM fructose. Experiments were performed by exposing 15 mL of this solution in a petri dish to UV light for up to 4 min. Diuron concentration was measured using reverse phase high pressure liquid chromatography (Bioanalytical systems, Inc. IN, USA). The mobile phase consisted of 70% acetonitrile and 30% 100 mM sodium phosphate buffer at pH of 3.5. The flow was isocratic and the rate was maintained at 1 mL/min throughout the analysis. Separation was carried out by a C18 column of 5 mm thickness and dimensions of 45 mm length 4.6 mm ID (Shimadzu Inc., MD, USA). Diuron was detected by SPD- 20A Prominence UV/Vis detector at 254 nm. The retention time for diuron with the above HPLC settings was 1.2 min. Data was fitted into the first order kinetics equation and rate constants were calculated.

2.7. Photochemical degradation of chlorpyrifos

Stock solution of chlorpyrifos (1000 mg/L) was prepared in acetonitrile and stored at 4 C. Working solution of chlorpyrifos was prepared by diluting the stock solution with de-ionized water to a concentration of 2 mg/L. To this, 0e500 mM of fructose was added. Experiments were performed by exposing 15 mL of this final solution in a petri dish to UV light for up to 20 min. Samples were taken periodically and chlorpyrifos concentration was measured using the following method previously developed in our laboratory. Chlorpyrifos was extracted from the treatment solution by liquid- eliquid extraction using iso-octane as the extracting solvent in a 1:1 ratio. Aqueous and organic phases were mixed gently for 15 min. The samples were then centrifuged at 5712 rcf for 5 min. Chlorpyrifos concentration in the iso-octane phase was measured using a gas chromatograph (5890 Series II GC, HewlettePackard) with a DB-5MS column (30 m length, 0.25 mm internal diameter, 0.25 mm film) (Agilent technologies, Delaware, USA). The carrier phase consisted of helium (1 mL/min) with hydrogen (12.1 mL/min) and air (10.3 mL/min). Injection temperature was 225 C, the initial column temperature was set at 150 C for 2 min, then increased to 200 C at a rate of 15 C/minute. A pulsed flame photometric detector set at 250 C was used for detection. Retention time for chlorpyrifos was 12.28 min.

2.8. Data analysis

Statistical analysis was performed using software STATA 11 Fig. 1. Absorbance spectrum of (a) 10 g/L (500 mM) fructose solution, (b) 25 mg/L (StataCorp, Inc., TX). Two-tailed t-test was used to determine diuron solution and (c) 2 mg/L chlorpyrifos solution. S. Nayak et al. / Chemosphere 144 (2016) 1690e1697 1693

fluorescence properties, the same experiment was repeated in 100 mM phosphate buffer as well (Fig. 2b). Similar to the treatment in distilled water, fluorescence decreased significantly when fluorescein solution prepared in 100 mM phosphate buffer was exposed to UV light in the presence of fructose. The reaction followed first order kinetics (r2 > 0.85) and the average rate constants for 10, 100 and 500 mM fructose were 0.04 ± 0.01, 0.19 ± 0.05, 0.54 ± 0.31 min 1 respectively. A statistical comparison of the two systems revealed that the rate constants were significantly lower in solutions prepared in 100 mM phosphate buffer versus those prepared in distilled water (p < 0.05). The results of this experiment suggest that ROS generated from UV exposure of fructose and some of the photolysis products of fructose were acidic in nature. The quantum yield for degradation of fructose by UV light was calcu- ðf Þ lated based on the amount of UV energy incident incident as well f as the amount of UV energy absorbed ( abs) by the fructose solution. The calculations are given in the supplementary information section. These values were 6.9 10 3 and 0.27 respectively. A large difference between the two values was a result of a very low absorbance of fructose at 254 nm. This is expected, since only a small amount of fructose that is present in the open chain configuration (0.8% of total sugar) is expected to absorb the 254 nm UV light (Triantaphylides et al., 1982). Based on a relatively large value f of abs , it is evident that the fraction of photons that are absorbed are efficiently utilized to degrade fructose molecules (potentially the ones that are present in the open chain configuration) and produce ROS. Due to the low specificity of fluorescein, this probe did not provide an understanding of chemical identity of the ROS. There- fore, experiments were performed to identify specific ROS. Based on the available literature, we hypothesized that hydrogen peroxide and singlet oxygen were likely to generate from photolysis Fig. 2. Relative fluorescence decay of fluorescein as a function of duration of UV of fructose. exposure (minute) in the presence of fructose (10, 100 and 500 mM) in (a) distilled water, and (b) 100 mM phosphate buffer of pH 6.7. Each data point is an average of triplicate measurement ± standard deviation. The solid lines indicate the fit for the first 3.2. Generation of hydrogen peroxide in UV treated fructose order rate equation. solution

Fig. 3 shows the concentration of hydrogen peroxide generated fl exposed to UV light. In the absence of fructose, the relative uo- by UV treatment of 500 mM fructose prepared in distilled water, or rescence upon exposure to UV light remained close to 100% for up phosphate buffer of pH 4.5 and 6.7. Hydrogen peroxide was not fl to 10 min. In the presence of fructose, the UV exposure of uores- detected in the control (distilled water, buffer solutions of pH 4.5 fl cein resulted in a gradual decrease of uorescence, indicating and 6.7) exposed to UV light. However, it was detected in UV fl generation of ROS. The loss of uorescence as a function of UV exposed fructose solution prepared in distilled water and buffers. In exposure followed first order kinetics (r2 > 0.9) and the average rate constants of fluorescein degradation in the presence of 10, 100, and 500 mM fructose in distilled water were 0.07 ± 0.0, 0.60 ± 0.14, and 1.46 ± 0.18 min 1 respectively. The rate of fluorescence quenching increased with concentration of fructose (p < 0.05). Interestingly, the pH of the 500 mM fructose solution decreased substantially, from 5.79 ± 0.01 to 3.53 ± 0.25 after 10 min of UV exposure. Similarly, 100 and 10 mM fructose solutions also showed a decrease in pH from 5.89 ± 0.01 to 3.84 ± 0.034 and 5.90 ± 0.02 to 5.64 ± 0.087 respectively. Thus, the pH change was concentration dependent with the highest decrease observed in 500 mM fructose solution. A decrease in pH indicates that UV exposure of fructose resulted in the generation of acidic species. These results are consistent with the study by Triantaphylides and others, who also showed formation of acidic species such as formic acid, glycolic acid, D-erythronic acid and D-arabinonic acid during photolysis of fructose (Triantaphylides et al., 1982). The fluorescence of fluorescein in aqueous solutions is pH dependent and its quantum yield decreases substantially with a Fig. 3. Hydrogen peroxide concentration (mM) generated as a function of duration of decrease in pH (Klonis and Sawyer, 1996). Thus, the observed loss of UV exposure by 500 mM fructose solution in DI water and 100 mM phosphate buffers fl uorescence was due to a combination of ROS generation and of pH 4.5 and 6.7. Each data point is an average of triplicate measurement ± standard change in pH. In order to determine the effect of ROS alone on deviation. 1694 S. Nayak et al. / Chemosphere 144 (2016) 1690e1697 all the samples, a rapid increase in concentration of hydrogen et al., 2007). peroxide was observed in the first minute of exposure, followed by Formation of singlet oxygen in UV exposed fructose was a more gradual increase in concentration for the next 4 min. After depicted by an increase in solution fluorescence from SOSG-EP 5 min, the hydrogen peroxide concentrations stabilized and did not (Fig. 4a). The relative fluorescence increased in control (phos- change significantly (p < 0.05). The average concentrations of phate buffer) as well as treatment (500 mM fructose) solutions 1 hydrogen peroxide produced in 500 mM fructose solution in indicating that O2 was generated in both the samples. Production 1 distilled water (initial pH 5.8) and buffers of pH 4.5 and 6.7 at the of O2 in UV-treated SOSG solutions in the absence of photosensi- end of 7 min of UV exposure were 63.7 ± 1.11, 52.91 ± 1.07 and tizers has been reported previously (Ragas et al., 2009; Lin et al., 67.08 ± 1.08 mM respectively. The stabilization of hydrogen 2013). Nevertheless, based on the comparison of increase in rela- 1 peroxide concentration can be explained by the fact that hydrogen tive fluorescence values, a significantly higher amount of O2 was peroxide undergoes photolysis under UV light to produce two hy- generated in the treatment sample compared to control (p < 0.05). droxyl radicals per photon absorbed at 254 nm (Maillard et al., In an experiment to validate that the increase in fluorescence was 1992). Thus, generation of hydrogen peroxide is accompanied by not due to an interaction between SOSG and hydrogen peroxide or its photolysis, and the stable concentration observed after extended hydroxyl radicals, SOSG was incubated with 50 mMH2O2 or a UV exposure was possibly the steady-state concentration reached mixture of 5 mM ferrous sulfate and 50 mMH2O2 respectively and after equilibration between the rate of formation and the rate of fluorescence was measured immediately. There was no significant decomposition of hydrogen peroxide. difference (p > 0.05) in the fluorescence intensity of the three solutions (Fig. S1), indicating that the change in fluorescence was not 3.3. Generation of singlet oxygen in UV treated fructose solution due to the interaction of the dye with either hydrogen peroxide or hydroxyl radicals. However, interestingly, the relative fluorescence Singlet oxygen sensor green (SOSG) is a highly selective sensor began to decrease in fructose solution after one minute of UV 1 for detecting singlet oxygen ( O2). It does not show any appreciable exposure and continued to decrease for the duration of the 1 response to hydroxyl or superoxide radicals. In the presence of O2 experiment. On the contrary, the fluorescence continued to in- it shows green fluorescence due to formation of SOSG- crease in control samples through the duration of the experiment. endoperoxide (Lin et al., 2013). SOSG has been successfully used We hypothesized that it may be due to reduction of SOSG-EP by to detect singlet oxygen in diverse systems (Lin et al., 2013; Maisch fructose present in the cyclic form (99.2% of total fructose). Cyclic form of fructose is known to act as a reducing agent (Dong et al., 2014). Therefore, it is plausible that the reduction of SOSG-EP by cyclic fructose would lower the fluorescence intensity. To test this hypothesis, ascorbic acid (AA), a mild reducing agent was added to the buffer solution of pH 6.7 after 2 min of UV exposure at concentration of 50 mg/L. As shown in Fig. 4b the fluorescence of control solution continued to increase in the absence of AA. After its addition though, the fluorescence remained at the level observed after 2 min of exposure and did not increase upon further UV exposure. The fluorescence values of the two systems were significantly different after 5 min of UV exposure (two-tailed, p < 0.05). These results provide an indication that the presence of a reducing agent (such as cyclic fructose or AA) can inhibit SOSG-EP formation, such as that observed in experiments involving fructose. However, it does not fully explain the observed decrease in fluorescence intensity in presence of fructose, and further research is needed to address this unexpected behavior. It is also possible that the decrease in fluorescence intensity was due to photo-bleaching of the dye (Ragas et al., 2009). In our study, it is also interesting to 1 observe that generation of O2 followed a similar path as generation of hydrogen peroxide, in that both products showed a swift increase in concentration in the first minute of exposure. It may be 1 that hydrogen peroxide and O2 generation were correlated. In addition to using SOSG, we used furfuryl alcohol (FFA) to 1 quantify the steady-state concentration of O2 generated from the varying concentrations of fructose (0, 100, 500, and 1000 mM) exposed to 254 nm UV light. Solutions containing 40 mM FFA and 0e1000 mM fructose were exposed at 254 nm in phosphate buffer 1 and the steady-state O2 concentration was calculated for each condition by dividing the observed FFA degradation rate by the 1 7 1 1 reaction rate between FFA and O2 (8.3 10 M s )(Latch et al., 1 2003). It was determined that the steady-state O2 concentration ranged from 21 to 56 pM for the fructose concentrations tested (see 1 Fig. S2). These concentrations of O2 are within an order of magnitude of the range seen for the vitamin riboflavin at 20 mM and Fig. 4. (a) Relative fluorescence of SOSG-EP in solution of 500 mM fructose in phos- exposure to 365 nm light (Remucal and McNeill, 2011), albeit at phate buffer, as a function of duration of UV exposure, (b) Relative fluorescence of significantly higher concentrations (0e20 uM of riboflavin vs. SOSG-EP in solution of 50 mg/L ascorbic acid (AA) in phosphate buffer, added after e 2 min of UV exposure as a function of duration of UV exposure. Each data point is an 0 1000 mM of fructose). It is important to re-iterate that only the average of triplicate measurement ± standard deviation. acyclic form of fructose is responsible for photosensitization, which S. Nayak et al. / Chemosphere 144 (2016) 1690e1697 1695 is estimated to be present at 0.8% of the total concentration of showed that the rate constant for degradation was significantly fructose. Also using the FFA assay, exposure of 125 mM TiO2 to higher in presence of 500 mM fructose compared to control (one- 1 254 nm light led to the production of a steady-state O2 concen- tailed, p < 0.05), while it was not significantly different at 300 mM tration of 52 pM. fructose (p > 0.05). The t50% (time required for 50% degradation of UV exposure of fructose results in generation of ROS including diuron) were calculated based on the reaction constants. The t50%in hydrogen peroxide, hydroxyl radicals (likely to originate from presence of 500 mM fructose was 0.334 min compared to 0.74 min 1 photolysis of hydrogen peroxide) and O2. Since UV-assisted in the absence of fructose, indicating doubling of degradation rate advanced oxidative processes typically employ hydrogen in presence of fructose. The quantum yield (ɸ) for direct photo- peroxide, hydroxyl radicals and singlet oxygen for degradation of degradation of diuron in the absence and presence of 500 mM chemicals including pesticides (Badawy et al., 2006; Coelho et al., fructose was 0.003 and 0.006 respectively. The photosensitization 2011), it is plausible that fructose can act as a photosensitizer in activity of fructose was compared with commonly used photo- accelerating the UV induced degradation of the selected contami- sensitizers such as titanium dioxide (TiO2) and hydrogen peroxide nants, diuron and chlorpyrifos. (H2O2)(Fig. 5b). The degradation rate constants for diuron in the presence of 1.5 mM H2O2 and 125 mM TiO2 were 1.58 ± 0.4 and 1 3.4. Photochemical degradation of diuron 1.47 ± 0.4 min respectively. These concentrations for TiO2 and H2O2 were selected based on preliminary studies (Supplementary Fig. 1b shows the absorbance spectrum of diuron (25 mg/L). Fig. S3 and S4). These values are comparable to the degradation rate Diuron absorbed significantly in the UV-C region with a peak constants obtained in presence of 500 mM fructose. absorbance at 248 nm and an absorbance value of 1.69 at 254 nm. Photo-degradation of diuron by UV irradiation in the absence of Fig. 5a shows the UV induced degradation of diuron in the absence photosensitizer results in the formation of N-4,5-dichloro-phenyl- and presence of fructose (300 and 500 mM). UV alone induced isocyanate, N-dichloro-phenyl-isocyanate, benzene-isocyanate, N- degradation of diuron and the degradation followed first order ki- (4-isopropyl-phenyl)-formamide, aniline and 2-methyl-propinic netics (r2 > 0.99). Presence of fructose accelerated the degradation acid (Kiss and Virag, 2009). Photodegradation of diuron by hydroxyl of diuron. The average degradation rate constants for diuron in radical arising from ferrous ions resulted in formation of 3-(3,4- presence of 0, 300 and 500 mM fructose were 0.92 ± 0.03, dichlorophenyl)-1-formyl-1-methylurea as the major product 1.55 ± 0.49 and 2.07 ± 0.61 min 1 respectively. Statistical analysis (Mazellier et al., 1997). Since UV exposed fructose was shown to generate similar oxidative species, we anticipate that diuron degradation by UV exposed fructose would follows similar path- ways. However, more research is needed to determine if fructose produced other degradation products as well and to determine their toxicity.

3.5. Photochemical degradation of chlorpyrifos

Fig. 1c shows the absorbance spectrum of chlorpyrifos (2 mg/L). Chlorpyrifos did not absorb significantly in the UV-C region and a maximum absorbance of 0.06 was observed at 230 nm. The average extraction efficiency using the solvent extraction process was approximately 85%. Fig. 6a shows UV induced degradation of chlorpyrifos in the absence (control) and presence of various concentrations of fructose. Chlorpyrifos was inactivated by UV light alone. However, the UV induced degradation of chlorpyrifos was accelerated in presence of fructose. The degradation kinetics followed first order (r2 > 0.0.85) and the average degradation rate constants for chlorpyrifos in the presence of 0, 300 and 500 mM of fructose were 0.04 ± 0, 0.05 ± 0.01, and 0.07 ± 0 min 1 respectively. Statistical analysis of the degradation rate constants showed that the presence of 500 mM fructose significantly increased the rate of degradation compared to UV exposure alone, (p < 0.05), while 300 mM fructose did not have a significant effect (p > 0.05). The t50% (time required for 50% degradation of chlorpyrifos) was calculated based on the reaction constants. The t50% in presence of 500 mM fructose was 10.04 min compared to 18.7 min in the absence of fructose. The quantum yield (ɸ) for direct photo-degradation of chlorpyrifos and in the absence and presence of 500 mM fructose was 0.001 and 0.002 respectively. Hence it can be predicted that the oxidative species produced by fructose under UV light at 254 nm are capable of degrading chlorpyrifos with the optimum concentration of fructose at 500 mM, reducing the pesticide to nearly 20% at the end of 20 min. Photosensitizer induced degradation experiments for chlorpyr- Fig. 5. (a) Relative concentration of diuron remaining as a function of duration of UV ifos were also performed using H2O2 and TiO2. Fig. 6b shows the light exposure in (a) aqueous solution containing no photosensitizer (control), 300 and relative concentration of chlorpyrifos remaining as a function of UV 500 mM fructose, and (b) 1.5 mM hydrogen peroxide or 125 mM titanium dioxide. Each data point is an average of triplicate measurement ± standard deviation. The solid exposure time in the presence of 1.5 mM H2O2 and 125 mM TiO2. lines indicate the fit for the first order rate equation. Chlorpyrifos inactivation followed first order kinetics. The average 1696 S. Nayak et al. / Chemosphere 144 (2016) 1690e1697

important role than hydroxyl radicals in the oxidation of chlorpyrifos (Remucal, 2014; Zeng and Arnold, 2013). The varying susceptibility of the two compounds to the different ROS generated by UV exposure of fructose can also explain the differences in their degradation rates. However, it is important to highlight that it is difﬁcult to control the type of ROS generated by a given sensitizer. Thus, the selection of photosensitizer can be based upon the type of contaminant and its relative susceptibility to the diverse ROS.

4. Conclusion

Exposure to 254 nm UV light causes photolysis of fructose resulting in generation of oxidizing species such as hydrogen peroxide, hydroxyl radicals, singlet oxygen and miscellaneous acidic molecules. Due to generation of these oxidizing species, fructose can act as a photosensitizer and accelerate UV induced degradation of compounds. This was demonstrated by the accelerated photo-degradation of chlorpyrifos and diuron in the presence of fructose. Thus, the results of this study highlight a novel application of fructose as a photosensitizer. This may be particularly attractive to the food and beverage processing industry as the ef- ﬂuents emanating from them typically contain high amounts of fructose. The relatively lower cost of fructose, its biodegradability and signiﬁcantly higher chemical safety can make fructose an attractive alternative to the conventional photosensitizers.

Acknowledgment

This project was supported by Agriculture and Food Research Initiative grant no. 2014-67017-21642 from the USDA National Institute of Food and Agriculture (USDA-NIFA) (grant number 2014- 67017-21642) Program in Improving Food Quality (A1361) and Drexel IExE seed grant. Authors thank Dr. Liangli (Lucy) Yu, Boyan Gao and Qiao Ding at the University of MarylandeCollege Park for Fig. 6. (a) Relative concentration of chlorpyrifos remaining as a function of duration of fi UV light exposure in (a) aqueous solution containing no photosensitizer (control), 300 their assistance in quanti cation of degradation of fructose and Dr. and 500 mM fructose, and (b) 1.5 mM hydrogen peroxide or 125 mM titanium dioxide. Saeed Keshani Langroodi and Jacob Price at Drexel University for Each data point is an average of triplicate measurement ± standard deviation. The solid their assistance in quantification of singlet oxygen using FFA assay. lines indicate the fit for the first order rate equation. Appendix A. Supplementary data degradation rate constants for 1.47 mM H2O2 and 125 mM TiO2 Supplementary data related to this article can be found at http:// 1 were 0.91 ± 0.0.5 and 0.98 ± 0.47 min respectively. Thus, the dx.doi.org/10.1016/j.chemosphere.2015.10.074. inactivation rate of chlorpyrifos in presence of H2O2 and TiO2 was fi signi cantly higher than in the presence of 500 mM fructose. A References previous study showed that UV/H2O2/TiO2 treated chlorpyrifos solution resulted in the formation of carbonyls groups such as Affam, A.C., Chaudhuri, M., 2013. Degradation of pesticides chlorpyrifos, cyper- aldehyde and carboxylic acids, and isomerization of the pesticide methrin and chlorothalonil in aqueous solution by TiO2 photocatalysis. J. Environ. Manag. 130, 160e165. http://dx.doi.org/10.1016/j.jenvman.2013.08.058. (Affam and Chaudhuri, 2013). Since the oxidative species generated Badawy, M.I., Ghaly, M.Y., Gad-Allah, T.A., 2006. Advanced oxidation processes for by UV exposed fructose are similar to those produced by UV/H2O2/ the removal of organophosphorus pesticides from wastewater. Desalination 194, 166e175. http://dx.doi.org/10.1016/j.desal.2005.09.027. TiO2, we expect formation of similar by-products of photo- Bou, R., Codony, R., Tres, A., Decker, E.A., Guardiola, F., 2008. Determination of hy- degradation of chlorpyrifos. However more research is needed to droperoxides in foods and biological samples by the ferrous oxidation-xylenol investigate whether fructose produced other degradation com- orange method: a review of the factors that influence the method's perfor- pounds and their toxicity. mance. Anal. Biochem. 377, 1e15. http://dx.doi.org/10.1016/j.ab.2008.02.029. Bule, M.V., Desai, K.M., Parisi, B., Parulekar, S.J., Slade, P., Singhal, R.S., Rodriguez, A., Based on the comparison between the rate constants, it is 2010. Furan formation during uv-treatment of fruit juices. Food Chem. 122, evident that chlorpyrifos was more recalcitrant than diuron 937e942. http://dx.doi.org/10.1016/j.foodchem.2010.03.116. regardless of the type of photosensitizer used. However, the effect Chinnici, F., Masino, F., Antonelli, A., 2003. Determination of furanic compounds in was more pronounced with fructose, possibly due to the differences traditional balsamic vinegars by ion-exclusion liquid chromatography and diode-array detection. J. Chromatogr. Sci. 41, 305e310. in the relative concentration of individual oxidizing species pro- Coelho, C., Cavani, L., Halle Ter, A., Guyot, G., Ciavatta, C., Richard, C., 2011. Rates of duced by each photosensitizer. The rate constants for photo- production of hydroxyl radicals and singlet oxygen from irradiated compost. Chemosphere 85, 630e636. http://dx.doi.org/10.1016/j.chemosphere.2011.07.007. degradation in the presence or absence of fructose degradation e fi Cox, C., 2003. Diuron. J. Pestic. reform 23, 12 20. was signi cantly lower for chlorpyrifos compared to diuron. This Devi, L.G., Murthy, B.N., Kumar, S.G., 2009. Photocatalytic activity of V5þ, Mo6þ and can be explained at least partially by a significantly lower absor- Th4þ doped polycrystalline TiO2 for the degradation of chlorpyrifos under UV/ bance of chlorpyrifos at 254 nm compared to diuron (Fig. 1). solar light. J. Mol. Catal. A Chem. 308, 174e181. http://dx.doi.org/10.1016/ j.molcata.2009.04.007. Although diuron and chlorpyrifos are both known to be susceptible Djebbar, K.E., Zertal, A., Debbache, N., Sehili, T., 2008. Comparison of diuron degra- to oxidation by hydroxyl radicals, singlet oxygen occupies a more dation by direct UV photolysis and advanced oxidation processes. J. Environ. S. Nayak et al. / Chemosphere 144 (2016) 1690e1697 1697

Manage 88, 1505e1512. http://dx.doi.org/10.1016/j.jenvman.2007.07.034. liquid foods. Part 1: effects on quality of fructose solution, apple juice, and milk. Dong, Y.Y., Li, S.M., Ma, M.G., Yao, K., Sun, R.C., 2014. Compare study cellulose/Ag Food Bioprocess Technol. 6, 1580e1592. http://dx.doi.org/10.1007/s11947-012- hybrids using fructose and glucose as reducing reagents by hydrothermal 0779-8. method. Carbohydr. Polym. 106, 14e21. http://dx.doi.org/10.1016/ Ou, B., Hampsch-Woodill, M., Flanagan, J., Deemer, E.K., Prior, R.L., Huang, D., 2002. j.carbpol.2014.02.023. Novel fluorometric assay for hydroxyl radical prevention capacity using fluo- Elsinghorst, A., Tikekar, R.V., 2014. Generation of oxidative species from ultraviolet rescein as the probe. J. Agric. Food Chem. 50, 2772e2777. http://dx.doi.org/ light induced photolysis of fructose. Food Chem. 154, 276e281. 10.1021/jf011480w. Fenoll, J., Flores, P., Hellín, P., Martínez, C.M., Navarro, S., 2012. Photodegradation of Ou, B., Hampsch-Woodill, M., Prior, R.L., 2001. Development and validation of an eight miscellaneous pesticides in drinking water after treatment with semi- improved oxygen radical absorbance capacity assay using fluorescein as the conductor materials under sunlight at pilot plant scale. Chem. Eng. J. 204e205, fluorescent probe. J. Agric. Food Chem. 49, 4619e4626. http://dx.doi.org/ 54e64. http://dx.doi.org/10.1016/j.cej.2012.07.077. 10.1021/jf010586o. Fuerst, E.P., Norman, M.A., 2011. Weed science society of america interactions of Ragas, X., Jimenez-Banzo, A., Sanchez-García, D., Batllori, X., Nonell, S., 2009. Singlet herbicides with photosynthetic electron transport interactions of herbicides oxygen photosensitisation by the fluorescent probe singlet oxygen sensor with photosynthetic electron transport1. Weed Sci. 39, 458e464. green. R. Soc. Chem. 2920e2922. http://dx.doi.org/10.1039/b822776d. Kiss, A., Virag, D., 2009. Interpretation and modelling of environmental behaviour of Remucal, C.K., 2014. The role of indirect photochemical degradation in the envi- diverse pesticides by revealing photodecomposition mechanisms. Microchem. J. ronmental fate of pesticides: a review. Environ. Sci. Process. Impacts 16, 92, 119e122. http://dx.doi.org/10.1016/j.microc.2008.12.002. 628e653. Klonis, N., Sawyer, W.H., 1996. Spectral properties of the prototropic forms of Remucal, C.K., McNeill, K., 2011. Photosensitized amino acid degradation in the fluorescein in aqueous solution. J. Fluoresc. 6, 147e157. http://dx.doi.org/ presence of riboflavin and its derivatives. Environ. Sci. Technol. 45, 5230e5237. 10.1007/BF00732054. http://dx.doi.org/10.1021/es200411a. Latch, D.E., Stender, B.L., Packer, J.L., Arnold, W.A., McNeill, K., 2003. Photochemical Richardson, S.D., 2007. In: Water Analysis: Emerging Contaminants and Current fate of pharmaceuticals in the environment: cimetidine and ranitidine. Environ. Issues, 79, pp. 4295e4324. Sci. Technol. 37, 3342e3350. http://dx.doi.org/10.1021/es0340782. Simon, D., Helliwell, S., Robards, K., 1998. Analytical chemistry of chlorpyrifos and Lin, H., Shen, Y., Chen, D., Lin, L., Wilson, B.C., Li, B., Xie, S., 2013. Feasibility study diuron in aquatic ecosystems. Anal. Chim. Acta 360, 1e16. http://dx.doi.org/ on quantitative measurements of singlet oxygen generation using singlet oxy- 10.1016/S0003-2670(97)00680-6. gen sensor green. J. Fluoresc. 23, 41e47. http://dx.doi.org/10.1007/s10895-012- Tikekar, R.V., Anantheswaran, R.C., Laborde, L.F., 2012. Patulin Degradation in a 1114-5. Model Apple Juice System and in Apple Juice during Ultraviolet Processing, Maillard, C., Guillard, C., Pichat, P., 1992. Comparative effects of the titanium pp. 1e11. http://dx.doi.org/10.1111/jfpp.12047. dioxide-UV, hydrogen peroxide-UV, hydrogen peroxide-iron (II) systems on the Tikekar, R.V., Anantheswaran, R.C., LaBorde, L.F., 2011. Ascorbic acid degradation in a disappearance rate of benzamide and 4-hydroxybenzamide in water. Chemo- model apple juice system and in apple juice during ultraviolet processing and sphere 24, 1085e1094. storage. J. Food Sci. 76, 62e71. Maisch, T., Baier, J., Franz, B., Maier, M., Landthaler, M., Szeimies, R., Baumler,€ W., Triantaphylides, C., Schuchmann, H.-P., von Sonntag, C., 1982. Photolysis of d-fruc- 2007. The role of singlet oxygen and oxygen concentration in photodynamic tose in aqueous solution. Carbohydr. Res. 100, 131e141. http://dx.doi.org/ inactivation of bacteria. Proc. Natl. Acad. Sci. 104, 7223e7228. 10.1016/S0008-6215(00)81031-9. Mazellier, P., Jirkovsky, J., Bolte, M., 1997. Degradation of diuron photoinduced by Zeng, T., Arnold, W.A., 2013. Pesticide photolysis in prairie potholes: probing pho- iron ( III ) in aqueous solution. Pestic. Sci. 49, 259e267. tosensitized processes. Environ. Sci. Technol. 47, 6735e6745. http://dx.doi.org/ Orlowska, M., Koutchma, T., Grapperhus, M., Gallagher, J., Schaefer, R., Defelice, C., 10.1021/es3030808. 2013. Continuous and pulsed ultraviolet light for nonthermal treatment of