Photochemical Formation of Hydroxyl Radical from ... - ACS Publications

Feb 21, 2012 - Mei Mei Dong and Fernando L. Rosario-Ortiz*. Department of Civil, Environmental and Architectural Engineering, 428 UCB, University of ...
5 downloads 0 Views 267KB Size
Article pubs.acs.org/est

Photochemical Formation of Hydroxyl Radical from Effluent Organic Matter Mei Mei Dong and Fernando L. Rosario-Ortiz* Department of Civil, Environmental and Architectural Engineering, 428 UCB, University of Colorado, Boulder, Boulder, Colorado 80309, United States S Supporting Information *

ABSTRACT: The photochemical formation of hydroxyl radical (HO•) from effluent organic matter (EfOM) was evaluated using three bulk wastewater samples collected at different treatment facilities under simulated sunlight. For the samples studied, the formation rates of HO• HO• (R ) were obtained from the formation rate of phenol following the hydro• xylation of benzene. The values of RHO ranged from 2.3 to 3.8 × 10−10 M s−1 for • HO the samples studied. The formation rate of HO• from nitrate photolysis (RNO ) 3 −7 −1 −1 • was determinedHOto• be 3.0 × 10 MHO• MNO3 s−10. The−1HO production rate from EfOM (REfOM ) •ranged from 0.76 to 1.3 × 10 M s . For the wastewater HO samples studied, REfOM varied from 1.5 to 2.4 × 10−7 MHO• MC−1 s−1 on molar carbon basis, which was close to HO• production from nitrate photolysis. The a •) was apparent quantum yield for the formation of HO• from nitrate (ΦNO 3−HO determined as 0.010 ± 0.001 for the wavelength range 290−400 nm in ultrapure a •) water. The apparent quantum yield for HO• formation in EfOM (ΦEfOM−HO −5 −5 ranged from 6.1 to 9.8 × 10 , compared to 2.99 to 4.56 × 10 for organic matter (OM) isolates. The results indicate that wastewater effluents could produce significant concentrations of HO•, as shown by potential higher nitrate levels and relatively higher quantum yields of HO• formation from EfOM.



HO• is via the oxidation of ferrous iron in the presence of H2O2, known as the Fenton reaction.10−13 Although the mechanism by which nitrogen oxides form HO• is well understood,14 the fundamentals of the photochemically driven formation of HO• from OM are still being evaluated.15−17 The proposed mechanism includes the formation of an excited triplet state from the absorbance of specific chromophores within OM followed by a hydrogen abstraction from water to form HO•. The basis for this mechanism comes from studies examining the formation of HO• from the absorption of light by quinone-type compounds in aqueous systems.15,18−20 Recent evidence has shown that these processes form HO• in addition to other low level hydroxylating species, which could be characterized as HO•-complexes.16 Formation rates of HO• on the order of 10−12 to 10−11 M s−1 have been observed for many surface waters,12,21 and these values are on the same magnitude as ones reported for coastal and open ocean water (10−12 to 10−11 M s−1).22 When nitrate to OM ratios are less than 3.3 × 10−5 molNO3 mgC−1, it has been established that OM sensitization could dominate the production of HO•.21 In terms of the capacity of OM to produce HO•, formation rates around 1.1 × 10−11 M s−1 have been reported for river water,12 and they ranged from 0.27 to

INTRODUCTION One of the key areas of interest in environmental engineering is understanding processes that impact the fate and transport of organic contaminants in the environment. In the past decade, particular emphasis has been placed on a new group of organic contaminants, specifically organic compounds that are of human use.1−5 These compounds include pharmaceuticals, personal care products, and other synthetic organic compounds. Wastewater effluents constitute one of the main routes by which these compounds enter the environment. Once released into the environment, these widely occurring compounds were found to impose adverse effects on ecosystem health.6 The removal mechanisms of organic contaminants in surface waters include sorption, biodegradation, and photolysis.7 There are two general mechanisms for the photolytic removal of organic compounds: direct and indirect photolysis. Direct photolysis refers to excitation of a compound upon absorption of a photon, resulting in its transformation.7,8 Indirect photolysis refers to the degradation of a compound through a photochemically generated intermediate. These reactive intermediates include hydroxyl radical (HO•), singlet oxygen (1O2), peroxyl radicals, and hydrogen peroxide (H2O2).9 Among these species, HO• is of particular importance due to its high reactivity toward organic contaminants. For surface waters, the overall formation rate of HO• will depend on the photolysis of nitrite, nitrate, and sensitization of organic matter (OM).7 Another pathway for the formation of © 2012 American Chemical Society

Received: Revised: Accepted: Published: 3788

December 5, 2011 February 7, 2012 February 21, 2012 February 21, 2012 dx.doi.org/10.1021/es2043454 | Environ. Sci. Technol. 2012, 46, 3788−3794

Environmental Science & Technology

Article

3.4 × 10−11 M s−1 22 for open ocean and coastal water (not including the contribution of nitrate). Though HO• formation rates through OM sensitization have been studied, the apparent quantum yield for HO• formation from OM sensitization a • ) are infrequently reported. One study found that (ΦOM−HO a ΦOM−HO• ranged from 1.1 to 3.0 × 10−4 for three bay waters and 7.5 × 10−5 for Suwannee River fulvic acid (SRFA) when irradiating at wavelength of 320 nm.17 The same study also examined SRFA at nine wavelengths in the range of 290−360, a • was 5.4 × 10−5. and the average ΦOM−HO Although the formation rates of HO• from OM have been reported in surface and ocean waters, no detailed studies have been conducted on wastewater-derived effluent organic matter (EfOM) or to quantify the apparent quantum yield for a • ). the formation of HO• from EfOM sensitization (ΦEfOM−HO The main objective of this work was to investigate the HO• a • when irradiating EfOM under formation rate and ΦEfOM−HO simulated sunlight. Given that wastewater-impacted streams are dominated by EfOM, detailed information is needed on its capacity to form HO• and its impact on the removal of organic contaminants.

Figure 1. Photon irradiance as a function of wavelength obtained using the experimental setup (xenon lamp in Oriel solar simulator (model no. 81171)) taken using Ocean Optic spectrometer.



glass absorption which yielded 2.1× 10−8 einstein s−1 cm−2. Samples were spiked with •3 mM benzene directly to quantify the HO• production (RHO ) through the formation of phenol via Dorfman reaction.24 The second order reaction rate constant between HO• and benzene has been reported as 7.8 × 109 M−1 s−1.25 Addition of 3 mM of benzene yielded a scavenging rate of approximately 2 × 107 s−1, at least 2 orders of magnitude greater than the total scavenging rate of 105 s−1 estimated from the water matrix.26,27 Therefore, benzene was the main scavenger of HO• under these conditions, and HO• formation was quantified by determining the product phenol concentration. The rate of phenol formation was corrected with the reaction yield between benzene and HO• via eq 1 to HO• obtain the value of R . Previous studies have shown that the yield of phenol from the reaction between benzene and HO• is between 0.9528 and 0.75.29 An average yield of 0.85 was implemented for this study, similar to what was done elsewhere (eq 1).10 Due to the very low molar absorption coefficient from 290 to 400 nm for benzene, the concentration used had minimum effect on overall light screening. Previous studies have reported nitro-phenol formation using benzene as a hydroxyl radical probe.21 However, this was not of concern in these wastewaters for two reasons: the highest nitrate concentration was 0.89 mM, lower than the 5 mM reported in that study; and benzene present at high levels would preferentially react with HO• which would minimize the phenol nitration pathway.

MATERIALS AND METHODS Sample Collection. Three final secondary effluents (samples A, B, and C) were collected at the discharge outfall from three municipal wastewater facilities that employ an activated sludge process in Colorado. All three facilities incorporate nitrification and partial denitrification. Samples were filtered through a 0.7 μm glass fiber filter, stored in the dark and kept at 4 °C prior to the experiments. Samples were analyzed within one month of collection. Chemicals. OM isolates were obtained from the International Humic Substance Society (IHSS), and included Suwannee River humic acid (SRHA), SRFA, Suwannee River natural organic matter (SRNOM), and Pony Lake fulvic acid (PLFA). The OM isolate solutions were prepared in 1 mM carbonate buffer (pH 7.5 ± 0.1). Anhydrous benzene (99.8% Alpha Aesar), sodium bicarbonate (Mallinckrodt, ACS grade), phosphoric acid (JT Baker), and acetonitrile (HPLC grade, Honeywell) were obtained from VWR and used as received. Bovine liver catalase (Sigma-Aldrich) was also used as received. Irradiation Condition. An Oriel sunlight simulator (model 81171, Newport Corporation, California) equipped with a 1000 W xenon lamp and an air mass (AM) 1.5 global filter was used for all experiments. The irradiance was 71 W m−2, which corresponded to a photon irradiance of 2.2 × 10−8 einstein s−1 cm−2 in the range of 290−400 nm (Figure 1). The lamp scan was taken using a spectrometer USB2000 (Ocean Optics Inc., Florida). The lamp irradiance was almost twice the value of 44 W m−2 reported for AM 1.5 global irradiance under cloudless sky condition when summing irradiance from 290 to 400 nm,23 indicating that the solar simulator was equivalent to two sun power. Samples were held in 2 mL clear glass vials free of headspace and exposed to the collimated beam lying flat inside a water bath where the temperature was kept at 20 °C. The irradiated samples were collected at 30 min increments for a period of two and half hours. The effective water depth within the 2 mL clear glass vial was 0.84 cm, which was determined using the Beer−Lambert law. A clear vial filled with ultrapure water was used as blank, and absorbance for 60 mgN/L nitrate was measured at 304 nm which was used to back calculate the cell depth. The photon irradiance was corrected for the clear

• d[Phenol] = 0.85 × RHO dt

(1)

To minimize the contribution of the Fenton process to the formation of HO• in the wastewater samples studied, 20 unit/mL of catalase was added into each sample 30 min before irradiations to quench the H2O2 produced during the irradiation either through the photolysis of EfOM or as a byproduct of the hydroxylation of benzene. Addition of 20 unit/mL catalase added 6 mg/L carbon into the solution, which yielded a scavenging rate of 2 × 105 s−1 (assuming a typical reaction rate constant of 3 × 104 L mg−1 s−1 between added carbon and HO•). The resulting HO• scavenging rate was 2 orders of magnitude lower than that from benzene (107 s−1). This concentration of catalase exhibited minimum absorb3789

dx.doi.org/10.1021/es2043454 | Environ. Sci. Technol. 2012, 46, 3788−3794

Environmental Science & Technology

Article

Table 1. Water Quality for Three Wastewaters Samples Collected sample

DOC (mgC/L)

NO3− (mM)

SUVA254 (m−1mg−1L)

FI

alkalinity (mM)

total Fe

pH

REfOMHO• × 107 (MHO•/MC s−1)

A B C

7 6.4 6

0.89 0.29 0.71

2.2 2.2 1.9

2.06 1.94 1.85

1.2 1.4 1.8

1.04 0.67 0.84

7.4 7.5 7.1

1.9 ± 0.5 2.4 ± 0.2 1.5 ± 0.4

ance from 290 to 400 nm, therefore catalase addition did not interfere with the matrix light absorption. All irradiations were performed in triplicate and results are presented as the average values of these triplicate samples with error bars corresponding to two standard deviations. Analytical Methods. Dissolved organic carbon (DOC) (method detection level (MDL) 0.2 mgC/L) was measured using a TOC-VSCH (Shimadzu Corp., Japan) analyzer. Nitrate (MDL = 0.010 mgN/L) was determined using ion chromatography (IC) (DX-500 Dionex Corp., California). Ultraviolet absorbance at 254 nm (UV254), and pH were measured using standard procedures.30 An Agilent 1200 LC system (Palo Alto, CA) with a Zorbax SB- C18 5 μm 46 × 250 mm column (Agilent, Germany) was used with an injection volume of 100 μL for phenol analysis. A variable wavelength detector from Agilent was used (model 1200 Palo Alto, CA) monitoring at 210 nm. The mobile phase flow was set at 1 mL/min and consisted of a mixture of phosphate buffer (pH 2.8) and methanol (60/40). The method run times were 10 min, and phenol eluted at approximately 8.6 min. Fluorescence analysis was completed with a Flouromax-4 fluorometer (Jobin Yvon Horiba, CA). Routine lamp, Raman, and cuvette checks were completed daily. Excitation was from 240 to 450 at 10 nm increments. Emission scans were collected from 350 to 550 at 2 nm increments. Slit widths were set at 5 nm for both excitation and emission. The fluorescence index (FI) was calculated as the ratio of the emission intensity at 470 and 520 nm at an excitation of 370 nm.31 Total iron (MDL = 0.016 mg/L) was measured using ICP-OES (ARL model 3410+) (Thermo Scientific) for all three wastewater samples.

Figure 2. Phenol concentration in samples A, B, and C (catalase added) as a function of irradiation time under simulated sunlight. The solar simulator outputs irradiance of 71 W m−2 summing photons from 290 to 400 nm. Error bars represent two standard deviations of triplicate analysis.

could produce significant amounts of H2O2, catalase was added to minimize the HO• production from Fenton reaction. Figure 2 presents the formation of phenol as a function of irradiation time for the three wastewaters studied with added catalase. The • obtained values for RHO were between 2.3 and 3.8 × 10−10 M s−1. These values were in general higher than those reported in waters from different lakes (range of 0.69−5.9 × 10−11 M s−1) with low nitrate (