H2O2 Advanced Oxidation

Oct 17, 2013 - fluence including wavelengths from 200 to 300 nm was calculated using the ... (Agilent Technologies, Santa Clara, CA). Antibacterial Ac...
1 downloads 0 Views 891KB Size
Article pubs.acs.org/est

Degradation of Antibiotic Activity during UV/H2O2 Advanced Oxidation and Photolysis in Wastewater Effluent Olya S. Keen†,‡ and Karl G. Linden*,† †

Civil, Environmental and Architectural Engineering, University of Colorado - Boulder, Boulder, Colorado 80309, United States Civil and Environmental Engineering, University of North Carolina, Charlotte, North Carolina 28223, United States



S Supporting Information *

ABSTRACT: Trace levels of antibiotics in treated wastewater effluents may present a human health risk due to the rise of antibacterial activity in the downstream environments. Advanced oxidation has a potential to become an effective treatment technology for transforming trace antibiotics in wastewater effluents, but residual or newly generated antibacterial properties of transformation products are a concern. This study demonstrates the effect of UV photolysis and UV/H2O2 advanced oxidation on transformation of 6 antibiotics, each a representative of a different structural class, in pure water and in two different effluents and reports new or confirmatory photolysis quantum yields and hydroxyl radical rate constants. The decay of the parent compound was monitored with HPLC/ITMS, and the corresponding changes in antibacterial activity were measured using bacterial inhibition assays. No antibacterially active products were observed following treatment for four of the six antibiotics (clindamycin, ciprofloxacin, penicillin-G, and trimethoprim). The remaining two antibiotics (erythromycin and doxycycline) showed some intermediates with antibacterial activity at low treatment doses. The antibacterially active products lost activity as the UV dose increased past 500 mJ/cm2. Active products were observed only in wastewater effluents and not in pure water, suggesting that complex secondary reactions controlled by the composition of the matrix were responsible for their formation. This outcome emphasizes the importance of bench-scale experiments in realistic water matrices. Most importantly, the results indicate that photosensitized processes during high dose wastewater disinfection may be creating antibacterially active transformation products from some common antibiotics.



INTRODUCTION

in bench-scale experiments also exhibited resistance to silver ion.22 Several studies also present findings with implications for water reuse. For example, soil irrigated with treated wastewater effluent showed higher presence and broader spectrum of antibacterial resistance than soil watered with groundwater.19 Another study demonstrated that plants irrigated with recycled water can uptake some of the antibiotics.23 Some antibiotics present in reclaimed water used for irrigation accumulate in soils and persist for months after the irrigation season.24 These results indicate that antibiotic-resistant human pathogens can be introduced to food crops which may have health implications and decrease acceptance of water reuse practices for crop irrigation. Concern about the spread of antibacterial resistance in the environment due to trace antibiotics in WWTP effluents led scientists to an evaluation of potential treatment technologies. Although apparent attenuation occurs during the wastewater treatment process (mainly due to the partitioning of the compounds into sludge), trace levels capable of inducing

Antibiotics have been detected in wastewater treatment plant (WWTP) effluents worldwide.1−6 Selective pressure by subinhibitory levels of these compounds could cause proliferation of antibacterial resistance among human pathogens. Indeed, several studies have been conducted documenting the connection between wastewater treatment plant effluent discharge and presence of antibiotic resistant micro-organisms in the environment. These trends have been reported downstream of an antibiotic manufacturer discharge,7 facilities treating hospital wastewater8,9 and pharmaceutical manufacturing wastewater,10 conventional WWTPs,11−15 and a WWTP using tertiary treatment.16 Surveys of wider watersheds also indicated that wastewater treatment plants along with runoff from animal husbandry are significant sources of antibacterial resistance genes in the environment.17,18 In addition to an increased presence of antibiotic resistant organisms downstream of WWTPs, there is also a shift in microbial representation7 and a broader spectrum of resistance.9 The resistance genes found in wastewater impacted streams have been shown to be transferrable to nonresistant bacteria species introduced to samples in the lab.19,20 In the samples impacted by WWTP effluents, bacteria have shown resistance not typical for the species.21 Bacteria exposed to trace levels of amoxicillin © 2013 American Chemical Society

Received: Revised: Accepted: Published: 13020

June 3, 2013 October 12, 2013 October 17, 2013 October 17, 2013 dx.doi.org/10.1021/es402472x | Environ. Sci. Technol. 2013, 47, 13020−13030

Environmental Science & Technology

Article

Table 1. Summary of Antibiotics Used in the Study

catalysis,28 and ozonation29−31 were all shown to be very effective for a number of antibiotics, but some compounds were more amenable than others because of the selective nature of these processes. Oxidation by hydroxyl radicals has shown nonselective effectiveness for a number of antibiotics tested.29 Limited studies exist evaluating the effects of UV photolysis on degradation of the antibacterial activity of antibiotics.32

antibiotic resistance in the environment can still make it through the treatment works.25 Because of the high solubility of most antibiotics, compared to other pharmaceuticals, oxidative or other transformative processes (e.g., photolysis) rather than sorption-based processes have been considered the most promising technology. Oxidation by potassium permanganate,26 Fenton advanced oxidation,27 titanium dioxide photo13021

dx.doi.org/10.1021/es402472x | Environ. Sci. Technol. 2013, 47, 13020−13030

Environmental Science & Technology

Article

state HO· concentration and therefore the efficiency of advanced oxidation processes for contaminant oxidation. Bacterial stocks were purchased from ATCC. Bacillus subtilis Marburg strain (ATCC 6051) was used as a gram-positive surrogate and Escherichia coli B strain (ATCC 11303) was used as a gram-negative surrogate. Both strains have no known resistance to the antibiotics selected. Difco nutrient broth was used for B. subtilis and was purchased from BD (Sparks, MD) and consisted of 3.0 g/L of beef extract and 5.0 g/L peptone. Difco nutrient agar contained the same ingredients plus 15 g/L of agar. Nutrient broth used for E. coli was prepared with 10 g/ L Bacto tryptone (BD, Sparks, MD), 5 g/L Bacto yeast extract, and 5 g/L sodium chloride (BDH, West Chester, PA). Nutrient agar for E. coli was prepared by adding 15 g/L Bacto agar (BD, Sparks, MD) to the nutrient broth. Phosphate buffered saline (PBS) was prepared with 8 g/L sodium chloride (BDH, West Chester), 0.2 g/L potassium chloride (Fisher Scientific, Rochester, NY), 1.81 g/L of dibasic sodium phosphate dihydrate (Na2HPO4·2H2O) and 0.24 g/L of potassium phosphate monobasic KH2PO4 (EMD, Gibbstown, NJ) with final pH 7.4. Antibiotic concentrations used in the experiments for quantum yield and kHO determination were between 0.68 and 1.72 μM, which was about 100 times above the detection limit of the instrument for the compounds. Concentrations used in the antibacterial assays were higher to make sure that the antibiotic was present at its inhibitory level before the serial dilution and at its subinhibitory level after the serial dilution. Concentrations of 20−30 times the LD50 (determined in preliminary experiments) were selected to achieve the result in 10-member 2-fold serial dilution. The corresponding concentrations for the compounds were 13.6 μM erythromycin, 22.5 μM doxycycline, 4.67 μM clindamycin, 29.9 μM penicillin-G, 6.04 μM ciprofloxacin, and 68.9 μM trimethoprim. All stock solutions were prepared in pure unbuffered water. Photolysis and Advanced Oxidation. Both low-pressure mercury vapor and medium-pressure mercury vapor UV sources were used in the study. The medium-pressure lamp system was manufactured by Calgon Carbon Inc. (Pittsburgh, PA) and consisted of a 1-kW lamp emitting a polychromatic spectrum above 200 nm. The lamp was collimated with a 10cm-long, 6.4-cm-diameter cylindrical tube. The low-pressure lamp system consisted of four 15-W lamps (ozone-free, General Electric G15T8) collimated by two 10-cm diameter apertures 1.2 cm apart. The low-pressure lamp emitted monochromatic UV at around 253.7 nm. Spectral irradiance of the lamps was measured with an Ocean Optics USB2000 spectrometer (Ocean Optics, Dunedin, FL), and incident irradiance at the sample surface was measured with a NIST calibrated IL-1700 radiometer (International Light, Peabody, MA). Average UV fluence including wavelengths from 200 to 300 nm was calculated using the appropriate factors and was not germicidally weighted.42 To generate hydroxyl radicals, samples were spiked with hydrogen peroxide (30%, Sigma-Aldrich, St.Louis, MO) so that the final concentration of H2O2 was ≈0.3 mM (10 mg/L). Hydrogen peroxide concentration was measured spectrophotometrically (Hach DR5000, Hach Corporation, Loveland, CO) using the triiodide method.43 Incremental UV fluences up to 2000 mJ/cm2 were used in treatment. Irradiations were carried out in batch reactors using a continuously stirred crystallization dish 50-mm in diameter. Sample depth was 1.9 cm, which was over 10 times smaller than the distance from the lamp, indicating that light path divergence

In recent years, researchers have begun to evaluate the performance of transformative treatment processes based on toxicity and biological activity end points rather than the disappearance of the parent compound. Studies have monitored the changes in estrogenic activity of the transformation products,33 toxicity,34 or biodegradability of the products of recalcitrant compounds.35 With antibiotics, the antibacterial potency of the products is monitored. 28−30,32,36 Some researchers also evaluated changes in toxicity to Vibrio f ischeri for several antibiotics during transformation by UV and UV/ H2O2.37 This study evaluated the effectiveness of UV/H2O2 advanced oxidation as well as direct photolysis by low-pressure and medium-pressure mercury vapor UV sources for transforming antibiotics into inactive products. Antibiotics selected for this study have been regularly detected in WWTP effluents at concentrations up to low μg/L at facilities employing a wide range of secondary and tertiary treatment processes.1,−3,38 Each antibiotic examined is representative of a different structural class. The experiments were performed in both pure water and treated wastewater effluent.



MATERIALS AND METHODS Reagents. Antibiotics used in the study were reagent grade: erythromycin (MP Biomedicals, Solon, OH), clindamycin (Enzo Life Sciences, Farmingdale, NY), doxycycline, penicillin-G, ciprofloxacin, and trimethoprim (all four manufactured by Sigma-Aldrich, St .Louis, MO). Most of the selected antibiotics are used exclusively for humans so wastewater treatment plants would be their main routes into the environment. The only two exceptions are erythromycin which, in addition to humans, is also used on farm animals, and clindamycin which is used for humans and pets.39 Table 1 summarizes some of the relevant details of the selected compounds. Experiments were performed in ultrapure water (arium 611VF, Sartorius Stedim, Bohemia, NY) and in secondary effluent from two wastewater treatment plants. The effluents were collected prior to disinfection and filtered through a 0.2-μm nylon filter (Millipore, Billerica, MA). Plant 1 uses solids contact activated sludge followed by nitrifying trickling filters, followed by denitrifying trickling filters. Plant 2 uses Modified Ludzack−Ettinger biological treatment (activated sludge with an anoxic zone at the influent portion of the tank for the denitrification of the return flow). Relevant water quality parameters for both effluents are listed in Table 2. Hydroxyl Table 2. Effluent Water Quality alkalinity, mg/L as CaCO3 123 88

pH

nitrite, mg-N/L

nitrate, mg-N/L

dissolved organic carbon, mg-C/L

6.84 6.58