Article pubs.acs.org/est
Kinetics and Mechanism of Sonochemical Degradation of Pharmaceuticals in Municipal Wastewater Ruiyang Xiao,† Zongsu Wei,† Dong Chen,‡ and Linda K. Weavers*,† †
Department of Civil, Environmental, and Geodetic Engineering, The Ohio State University, Columbus, Ohio 43210, United States Department of Engineering, Indiana University−Purdue University Fort Wayne, Fort Wayne, Indiana 46805, United States
‡
S Supporting Information *
ABSTRACT: A series of six pharmaceuticals were degraded by continuous wave (CW) and pulsed wave (PW) ultrasound at 205 kHz using deionized water, wastewater effluent, and its isolated organic matter matrices. In deionized water, we observed that hydrophobicity is superior to diffusivity (DW) for predicting degradation kinetics. Enhancements in degradation kinetics by the PW mode were greatest for the highest DW (i.e., fluorouracil (5-FU)) and KOW (i.e., lovastatin (LOVS)) compounds, indicating that a pharmaceutical with either high diffusivity and low hydrophobicity or low diffusivity and high hydrophobicity benefits from additional time to populate the bubble−water interface during the silent cycle of PW ultrasound. Degradation of 5-FU and LOVS were inhibited by wastewater effluent to a greater extent than the other pharmaceuticals. In addition, a pulse enhancement (PE) for 5-FU and LOVS was not present in wastewater effluent. Irradiating 5-FU and LOVS in hydrophobic (HPO), transphilic (TPI), and hydrophilic (HPI) fractions of effluent organic matter (EfOM) showed that the TPI fraction reduced the PE the most, followed by the HPI and HPO fractions. The smaller size of the TPI over the HPO fraction and higher hydrophobicity of TPI over HPI implicate both size and hydrophobicity of EfOM in hindering degradation of pharmaceuticals.
■
INTRODUCTION Many studies have reported the occurrence of various pharmaceuticals in different waters around the world, including tap water, wastewater, and receiving streams, ranging from parts-per-trillion to parts-per-billion concentrations. 1−6 Although the knowledge of adverse effects of chronic lowdose exposure to pharmaceutical mixtures on human health is limited, many toxicological studies have shown that exposure of fish and other aquatic organisms to these anthropogenic compounds causes reproductive and behavioral disorders.2,7−9 To minimize risk, efforts are underway to reduce human and aquatic organism exposure to pharmaceuticals in waters, especially from municipal wastewater effluent, the main route for pharmaceuticals to enter natural waters.10−13 Activated sludge treatment in wastewater treatment plants has been shown to improve the removal of some pharmaceuticals by using long sludge retention times.14,15 However, the majority of pharmaceuticals are not completely degraded, and most existing municipal wastewater treatment plants are not designed for operation using prolonged sludge retention times.16 Finding alternative technologies that effectively remove pharmaceuticals from wastewater effluents, therefore, has been of research interest.17−19 Ultrasound, an advanced oxidation process (AOP), shows potential to degrade pharmaceuticals in wastewater as a tertiary treatment technology.20−23 Ultrasound has unique advantages as compared to other technologies, such as no addition of chemicals, ease of use, and short contact times.24,25 Ultrasonic © 2014 American Chemical Society
irradiation induces chemical reactions in water from the collapse of cavitation bubbles.26 The collapse of cavitation bubbles generates localized hot spots, and these localized hot spots initiate thermolytic and oxidation reactions with contaminants.27,28 Under certain optimal conditions, pulsed wave (PW) ultrasound enhances the degradation of a compound, because it allows time (i.e., the silent cycle) for compounds to diffuse to bubble−water interfaces, the sources of reactivity.29−32 Properties of compounds, including surface excess (Γ),33,34 octanol−water partition coefficient (KOW),35,36 vapor pressure (p),36,37 Henry’s law constant (KH),38,39 second-order rate constant with •OH (k•OH),40,41 and diffusivity in water (DW),42−44 have been shown to affect degradation rates by CW ultrasound. In addition, benefits of pulsing ultrasound, measured by a pulse enhancement (PE), have been linked to larger DW.32 In surfactant solutions, PE values for surfactants were attributed to increased accumulation of surfactants on cavitation bubble surfaces.45 Although the correlation of contaminant degradation to these parameters has been reported, with DW, KOW, and KH being most commonly correlated with degradation, the question regarding which Received: Revised: Accepted: Published: 9675
April 2, 2014 June 25, 2014 July 15, 2014 July 15, 2014 dx.doi.org/10.1021/es5016197 | Environ. Sci. Technol. 2014, 48, 9675−9683
Environmental Science & Technology
Article
empty bed volumes per hour. The organic matter was backeluted from each column individually using 0.1 M NaOH. The collected wastewater effluent organic matter (EfOM) fractions were stored in the dark at 4 °C prior to use. The filtered wastewater effluent was characterized before pH adjustment (see Table S2 in the SI). The concentration of dissolved organic carbon in the solution was quantified by a Shimadzu TOC-5000A analyzer. Specific ultraviolet absorbance was measured at 280 nm (SUVA280) with a spectrophotometer (model UV-2401, Shimadzu). Concentrations of common elements were measured by inductively coupled plasma atomic emission spectroscopy, ICP-AES (Vista AX, Varian). Anion concentrations were determined by ion chromatography (IC) (DX-120, Dionex). MW distributions of the whole filtered and fractionated wastewater EfOM were determined with size exclusion chromatography (SEC) (Hewlett-Packard 1050) following the method of Chin et al. (see Table S3 in the SI).53 Sonochemical Experiments. Ultrasound at 205 kHz was emitted from a USW 51-52 ultrasonic flat plate transducer (A = 23.4 cm2) (ELAC Nautik, Inc., Kiel, Germany) into a water jacketed glass vessel with a volume of 300 mL. The temperature of the reactor was maintained at 20 °C. A SM-1020 Function/ Pulse generator (Signametrics Corp.) delivered sound waves continuously or in 100 ms on and 100 ms off pulses. The acoustic energy density to the reactor, determined by calorimetry, was 45 W L−1. During experiments, 0.5 mL samples for chemical analysis were taken from the reactor at designated times using a 1 mL glass syringe (Gastight 1001, Hamilton Corp.). The total sample volume taken during the course of sonication did not exceed 1% of the initial volume. The experiments were carried out in duplicate. A 10 μM initial pharmaceutical concentration was used for all experiments. One millimolar phosphate buffer was used to maintain the pH at 7.7. In select experiments, 2 mM of either SO2− or HCO−3 was added to solutions. The fraction of 4 degradation occurring in bulk solution was determined using 1 mM of CH3COO− as a bulk solution •OH scavenger, because CH3COO− quenches •OH in bulk solution without affecting cavitation bubbles.54 In other experiments, fractionated wastewater EfOM with a DOC of 3 mgC L−1 was added to experimental solutions. Chemical Analysis. A Hewlett-Packard 1100 HPLC equipped with a diode array detector (DAD) and a 5 μm, 150 × 2.1 mm SB-C18 column (Agilent Technologies) was used to quantify the concentration of the pharmaceuticals. Eluents consisted of 20 mM pH 3 phosphate buffer and acetonitrile with a flow rate of 0.5 mL min−1. An isocratic mobile phase of acetonitrile/phosphate buffer at pH 3.0 (2/98, 40/60, 20/80, 45/55, 55/45, 75/25, v/v) with detection at UV wavelengths of 265, 220, 220, 225, 238, and 238 nm was used for the quantification of 5-FU, IBU, CLND, ESTO, NIFE, and LOVS, respectively. Gibbs surface tension (γ) was measured by an Analite surface tension meter (2141, McVan Instruments). Molar volume was calculated at the Hartree−Fock (HF) level of theory with the 631+G* basis set55 and PCM solvation method.56 All calculations were performed using Gaussian 09 at the Ohio Supercomputer Center.57
aspect has a stronger influence on the cavitational system remains unclear. In this study, we systematically studied the roles that DW and KOW play in sonolytic degradation of pharmaceuticals. We selected six pharmaceuticals: fluorouracil (5-FU), ibuprofen (IBU), clonidine (CLND), estriol (ESTO), nifedipine (NIFE), and lovastatin (LOVS). The KOW and molecular weight of these six compounds are in increasing order and are inversely proportional to DW (Figure S1 in the Supporting Information, SI).46 In addition, the compounds were selected to minimize the potential influence of KH. Using both CW and PW modes of ultrasound, our comparative study with the six pharmaceuticals provides a means to probe the relative role of KOW and DW on their accumulation behavior to cavitation bubbles. Subsequently, we investigated the degradation of these six pharmaceuticals in a wastewater effluent to determine how the matrix alters accumulation behavior.
■
EXPERIMENTAL METHODS Materials. 5-FU (99%), IBU (99%), CLND (99%), ESTO (97%), NIFE (98%), sodium bicarbonate (99%), sodium sulfate (99%), and sodium bicarbonate (ACS Reagent grade) from Sigma-Aldrich, LOVS (98%) from TCI, methanol (HPLC grade), acetonitrile (HPLC grade), HCl (Trace metal grade), sodium acetate, and sodium hydroxide (97%) from Fisher Scientific, were used as received. Table S1 in the SI lists the physicochemical properties of the pharmaceuticals. Supelite XAD8 and Amberlite XAD4 resins were purchased from SigmaAldrich. A Chromaflex Chromatography column (1085 mL, 4.8 cm × 60 cm) with 0.20 mm of bed support was purchased from Kontes (Vineland, NJ). Water used was from a Millipore System (Millipore, MA). Wastewater Effluent Collection, Fractionation, And Characterization. Wastewater effluent was collected in December 2011 from a municipal wastewater treatment plant, located in the vicinity of Columbus, Ohio, U.S.A. The facility has a treatment capacity of 10 million gallons per day. The plant first uses traditional activated sludge treatment followed by clarification and tertiary rapid sand filters, but no primary clarification. The source of the wastewater is residential and commercial. The wastewater sample was collected at the plant effluent and poured into a 50 L polyethylene carboy with minimal headspace. The wastewater was consecutively prefiltered through prerinsed 5 and 0.45 μm groundwater filtration capsules (Pall Gelman). Filtered wastewater effluent was acidified to pH 2 with HCl, air stripped for 2 h to remove H2S, and stored in the dark at 4 °C until use. Before use, the pH of a wastewater effluent aliquot was readjusted to the field pH using NaOH. The concentrations of Na+ and Cl− in the wastewater effluent were 2.5 × 10−3 M and 2.0 × 10−3 M, respectively. The addition of Na+ and Cl− ions due to pH adjustment (ca. 0.01 M) was lower than the lowest reported concentration affecting cavitation (0.1 M).47−50 The XAD8 and XAD4 resins were cleaned and packed according to Standley and Kaplan.51 Organic matter from wastewater effluent was isolated using XAD8 and XAD4 resin columns following Quaranta et al.52 The XAD8 and XAD4 resin columns were used to retain operationally defined hydrophobic (HPO) and transphilic (TPI) fractions of organic matter, respectively. Permeate consisted of the hydrophilic (HPI) fraction of organic matter and inorganic salts. Approximately 40 empty bed volumes of filtered wastewater effluent passed through the columns with a flow rate of 15
■
RESULTS AND DISCUSSION Sonochemical Degradation in DI Water. Consistent with Langmuir−Hinshelwood kinetics, sonochemical degrada-
9676
dx.doi.org/10.1021/es5016197 | Environ. Sci. Technol. 2014, 48, 9675−9683
Environmental Science & Technology
Article
tion of the pharmaceuticals in DI water follows apparent firstorder kinetics (Figure S2), suggesting that the cavitation bubble surfaces are not saturated.58 Reported degradation rates are a product of the first-order rate constants and concentration. As shown in Figure 1(a), the degradation rates of LOVS under
opposite trend in our work may result from the set of compounds chosen in our study. By design, the set of compounds in our study are increasingly more hydrophobic as size increases and a positive trend with log KOW and DW or MW is not possible (Figure S1 in the SI). A trend with an equilibrium parameter such as log KOW is attributed to the compounds being at equilibrium with cavitation bubbles whereas a trend with increasing DW is attributed to the compounds not being at equilibrium with the cavitation bubbles.43 A trend in which the degradation rate increases with decreasing DW has not been reported to our knowledge. A logical explanation for this trend is that log KOW is more influential than DW in sonochemical degradation at these conditions. Indeed, at pH 3.5 the degradation rate of IBU increases compared to that at pH 7.7. At pH 3.5 log KOW = 3.7 for IBU but DW remains the same. The IBU degradation rate at pH = 3.5 is 0.58 ± 0.03 μM min−1, consistent with the log KOW trend but not the Dw trend. Pulsed wave ultrasound is another way to probe whether compounds are in equilibrium with cavitation bubbles. In Figure 1(a), the degradation rates for 5-FU, NIFE, and LOVS are faster under PW mode as compared to CW mode. IBU, CLND, and ESTO are similar or slightly slower under PW mode than CW mode. In order to evaluate the effect of pulsing ultrasound on degradation kinetics, PE was calculated from eq 1: PE(%) =
Figure 1. Sonochemical degradation rates of 5-FU, IBU, CLND, ESTO, NIFE, and LOVS in (a) DI water and (b) wastewater effluent (WW) as a function of log KOW (pH 7.7, 20 °C, [pharmaceutical]0 = 10 μM, and sonication power density at 45 W L−1).
d[C] dt PW
−
d[C] dt CW
d[C] dt CW
× 100 (1)
where d[C]/dtCW and d[C]/dtPW are the degradation rates of a pharmaceutical under CW and PW ultrasound, respectively. The PE of the six pharmaceuticals in DI water is shown in Figure S5. Pulsing ultrasound is beneficial for 5-FU, NIFE, and LOVS. The U-shaped curve shows that, a pharmaceutical with either small KOW and large Dw or large KOW and small Dw was positively affected by pulsing. For the compounds with moderate Dw and KOW, PW ultrasound did not increase the degradation kinetics. In previous work,32 we linked PE to the Dw of the PPCP by observing an inverse relationship between the molar volume of a PPCP and its PE. We did not explore the impact of hydrophobicity on PE explicitly. Our current work spans a wider range of both molecular size and log KOW and, because of the wider study range, reveals pulse enhancements at the extremities of the range studied. Acetate (1 mM), a bulk solution •OH scavenging agent, was irradiated with each pharmaceutical to determine the fraction of degradation occurring in bulk solution.32,54 5-FU and LOVS have a higher portion of degradation occurring in bulk solution, as compared to the other four pharmaceuticals (Figure S6 in the SI). Lower bulk solution degradation indicates that •OH is effectively trapped by the pharmaceuticals in or around the cavitation bubbles, whereas higher bulk solution degradation indicates that • OH is not entirely consumed by the pharmaceutical at or near the cavitation bubble surface. 5-FU has the highest Dw for reaching bubble−water interfaces, but also the lowest driving force to the bubble due to its hydrophilicity. On the contrary, NIFE and LOVS have lower Dw for migrating to the bubble−water interfaces, but higher driving forces to the bubble due to their higher hydrophobicities. The benefit of PW ultrasound is the
CW and PW ultrasound are 0.55 ± 0.01 and 0.67 ± 0.02 μM min−1, respectively, which is faster than those of the other compounds. 5-FU degraded slowest, with degradation rates of 0.111 ± 0.008 and 0.14 ± 0.01 μM min−1 under CW and PW ultrasound, respectively. The dependence of degradation rates of the pharmaceuticals under both CW and PW ultrasound on physicochemical properties, listed in Table S1, was evaluated by nonparametric correlations (Spearman’s rho) to analyze for any possible relationships (Figure 1(a) and Figure S3 in the SI). Figure 1(a) shows a positive relationship between the degradation rates and log KOW, indicating that a hydrophobic compound degrades faster than a hydrophilic one. Although surface excess is a more direct measure of accumulation on surfaces than logKOW, at these low concentrations, surface tension measurements of solutions did not change. In addition, the low solubility of the pharmaceuticals did not allow us to determine equilibrium surface partitioning coefficients. Diffusivity is an inverse function of the molar volume of the diffusing compound.46 Figure S3 in the SI shows a positive correlation between degradation rate and molecular weight and an inverse correlation with DW. These correlations indicate that a larger and slower diffusing pharmaceutical is destroyed faster than a smaller and faster diffusing one. The positive relationship between degradation rates and log KOW in this study is consistent with previous studies.35,36,59 However, the positive relationship between degradation rates and DW in this study contradicts previous reports.32,43,44 The 9677
dx.doi.org/10.1021/es5016197 | Environ. Sci. Technol. 2014, 48, 9675−9683
Environmental Science & Technology
Article
retardation of bubble growth during the silent cycle, allowing more time for 5-FU, NIFE, and LOVS to adsorb to bubble− water interfaces or reaction byproducts of previous bubble collapses to mass transfer away from the reactive bubbles.60 Thus, more adsorbed 5-FU, NIFE, and LOVS molecules react with •OH at cavitation bubbles in the subsequent pulse.61 For IBU, CLND, and ESTO, no benefit was gained by pulsing ultrasound. This lack of benefit of pulsing indicates that, in CW conditions, these compounds are either in effective equilibrium with the cavitation bubble surface or effectively capturing available •OH at the cavitation bubble surface. In the first case, no additional accumulation occurs with pulsing, resulting in no effect on degradation kinetics. In the second case, any additional accumulation on the cavitation bubble surface during pulsing, if it occurs, also has no effect on degradation kinetics. Estimated •OH concentrations of 4 mM in the cavity boundary, in excess compared to the pharmaceutical, suggest that these compounds are in effective equilibrium with the cavitation bubble.62 Sonochemical Degradation in Wastewater Effluent. When AOPs are investigated in water treatment, the water matrix containing inorganic and organic components greatly slows the degradation of the target contaminants due to matrix competition for •OH and/or sequestration of contaminants from AOP reactivity.50,63,64 However, little is known about which components in wastewater influence sonolytic degradation kinetics and if pulsing ultrasound changes the effect of the matrix on degradation. To evaluate the effect of wastewater effluent on the degradation kinetics, the pharmaceuticals were degraded individually in wastewater effluent (Figure 1(b)). Unlike the degradation kinetics observed in DI water, NIFE degraded faster than the other compounds with a degradation rate of 0.35 ± 0.04 μM min−1 under both CW and PW. As expected, with the higher concentration of the wastewater effluent components (see Table S2) compared to the pharmaceutical concentration in solution, the wastewater effluent slows degradation of the pharmaceuticals compared to degradation in DI water (Figure 2). The smallest but least hydrophobic and largest but most hydrophobic target compounds (i.e., 5-FU and LOVS, respectively) were inhibited between 70 and 90% whereas degradation of the other compounds was inhibited between 20 and 60% in the presence of wastewater effluent. 5-FU and LOVS have a higher portion of degradation occurring in bulk solution compared to the other compounds. The inhibitory effects of wastewater effluent on degradation of the pharmaceuticals (Figure 2) trend similarly with the portion of degradation that occurred in the bulk solution (Figure S6). The similar trend suggests that compounds reacting primarily on the bubble surface are minimally affected by wastewater effluent components but compounds reacting primarily in bulk solution are significantly affected by wastewater effluent components. To evaluate whether the wastewater effluent is simply acting as an •OH scavenger, using competition kinetics, the fraction of available •OH reacting with the pharmaceuticals, f OH, was calculated for compounds in which the second-order reaction rate constant with •OH is known (see Table S3).65 An average rate constant of wastewater effluent reacting with •OH was taken from Rosario-Ortiz et al., 2008.66 The observed degradation rates in wastewater are between 2.8 and 13.8 times larger than expected based on wastewater effluent competing for •OH (Table S3). The biggest discrepancies
Figure 2. Inhibition of wastewater effluent (WW) compared to diffusivity (a) and log KOW (b) for the six pharmaceuticals under CW and PW ultrasound.
occur for compounds that have less observed reaction in bulk solution. The overprediction of OH scavenging compared to what we observed indicates that the compounds are accumulating at or near the bubble surface, allowing them to outcompete wastewater effluent for sonolytic reactivity, and suggests a potential advantage of sonolysis over traditional AOPs. Notably, in wastewater effluent, the enhancement of degradation rates of the pharmaceuticals under PW ultrasound disappeared (Figure S5). The addition of the silent cycle allows wastewater components to interfere with pharmaceutical degradation. This effect was most prominent for 5-FU and LOVS. The loss of PE in the wastewater effluent indicates that either (1) wastewater effluent components are interacting with the pharmaceuticals, hindering pharmaceutical diffusion during the silent cycle, or (2) components of wastewater effluent are surface active. If the wastewater effluent were simply acting as a bulk solution •OH scavenger, we would expect to maintain or improve PE values for compounds with a PE in DI experiments. Although there is uncertainty in kEfOM−OH values used due to differences in EfOM at different plants, point 2 above is consistent with our competition kinetics analysis, while point 1 is not. To explore the role of the wastewater effluent matrix in sonolytic degradation, the major components of the wastewater effluent were investigated individually. Effect of Anions on Pharmaceutical Degradation. Previous studies indicated that sulfate and bicarbonate significantly reduce sonochemical degradation as compared to other anions, such as chloride and nitrate.50,67 Similar to − previous work, we expected the presence of SO2− 4 and HCO3 would either react with •OH (Table S4 in the SI) or partition to the bubble−water interface, reducing the temperature of collapsing cavitation bubbles, ultimately resulting in inhibition in degradation kinetics and a reduction of PE. 9678
dx.doi.org/10.1021/es5016197 | Environ. Sci. Technol. 2014, 48, 9675−9683
Environmental Science & Technology
Article
− The presence of SO2− 4 and HCO3 affected PW degradation of NIFE but had little to no effect on CW degradation of NIFE and the other pharmaceuticals under either condition (Figure − S8 in the SI). The negative effect of SO2− 4 and HCO3 on NIFE degradation under PW ultrasound but little effect under CW − ultrasound is puzzling. While, SO2− 4 and HCO3 may account for the loss of PE for NIFE, the presence of the anions is not the reason for inhibited degradation and PE loss of the other compounds in the wastewater effluent. Effect of EfOM on Pharmaceutical Degradation. Matrix organics have been shown to have different effects on sonochemical degradation of the target compounds.68−71 Matrix organics are substances with diverse size and polarity;66,72,73 the influence of various components on sonochemical degradation rates with CW or PW ultrasound is expected to be matrix dependent.68−71 In order to determine the role of the various matrix organics in wastewater effluent on sonolysis of pharmaceuticals, we fractionated the EfOM into operationally defined HPO, TPI, and HPI fractions. Only LOVS and 5-FU were investigated in these three fractions; they are the smallest and least hydrophobic and largest and most hydrophobic of the set, respectively, and experienced the greatest reduction in PE in the wastewater effluent. As illustrated in Table S3 in the SI, the EfOM was comprised of 5% TPI and the remaining was evenly distributed between the HPO and HPI fractions. The weight-averaged molecular weights of EfOM and its fractions were small for HPI (130 Da) and TPI (150 Da) and large for HPO (2060 Da). In addition, the SUVA280 value of the TPI fraction was larger than the corresponding HPO and HPI fractions, indicating that the TPI fraction has more aromatic character than the HPO and HPI fractions.52 These characteristics are similar to that observed for other EfOM studies.66,74 The size of LOVS (MWLOVS = 404 Da) is smaller than the HPO fraction but larger than the TPI and HPI fractions. On the other hand, the size of 5-FU (MW5‑FU = 103 Da) is smaller than the three fractions of EfOM. Figure 3 shows the sonochemical degradation rates of LOVS and 5-FU in different fractions of EfOM with CW ultrasound. The presence of the organic matrix negatively affects the degradation rates of both pharmaceuticals. The most hydrophobic fraction, the HPO fraction, had the least inhibition on LOVS degradation. This hydrophobic HPO fraction affected 5-
FU to a similar degree as the hydrophilic HPI fraction and less than the moderately hydrophobic TPI fraction. Under CW conditions, the similar degradation rates of LOVS with the similarly sized TPI and HPI fractions show that a change in hydrophobicity and aromatic character in the EfOM does not affect degradation kinetics. However, with 5-FU, the TPI fraction does reduce the degradation rate to a larger degree than both the larger and more hydrophobic (HPO) and similarsized and more hydrophilic (HPI) fractions. Figure 4 illustrates the effect of different fractions of EfOM on the PE of the target compounds. In contrast to DI
Figure 4. PE of 5-FU and LOVS in DI, wastewater effluent (WW), and different fractions of effluent organic matter (pH 7.7, 20 °C, [pharmaceutical]0 = 10 μM, and sonication power density at 45 W L−1).
experiments, both compounds had no PE or negative PE values in all fractions of EfOM. For LOVS, the TPI fraction had the largest negative PE value; a large negative PE value was also observed for the HPI fraction. The observed effects may be attributed to three different factors. First, •OH rate constants in EfOM depend on EfOM characteristics. Hydrophilicity and SUVA of EfOM have been correlated with •OH reactivity; MW and hydrophobicity have a negative correlation with •OH reactivity.66,74 The larger effect of TPI and HPI over HPO for LOVS is consistent with reported EfOM •OH reactivity. The larger effect of TPI with 5FU is consistent with the greater aromatic content in the TPI. The similar effect of HPO and HPI with 5-FU is inconsistent with reported smaller and more hydrophilic EfOM being more reactive. In addition, alone, differences in •OH reactivity in fractions do not account for negative PE values. Second, fractionated hydrophilic moieties in Suwanee River (SR) organic matter (OM) have been shown to bind more strongly with pharmaceuticals than hydrophobic moieties.75 Consistent with this factor, LOVS, with a higher KOC value, shows a stronger reduction in rate in HPI compared to HPO. However, this effect does not explain reductions in PE values under any of the conditions. Finally, the EfOM components may be migrating to cavitation bubble surfaces, consequently interfering with pharmaceutical sorption to cavitation bubble surfaces and competing for •OH at bubble surfaces. Xiao et al. observed that a smaller compound, terephthalic acid (TA, MW = 166 Da), had a 30−95% reduction in the sonolytic degradation rates of target contaminants compared to
