Abstract: Background and Objective: The sonochemical degradation of seven estrogen hormones (17α-estradiol, 17β-estradiol, estrone, 17α-dihydroequilin, 17α-ethinyl estradiol, estriol and equilin) in water and wastewater was investigated. Methodology: The effects of pH, presence of oxidants such as peroxide and persulfate were examined. Statistical 3D modeling was used as a tool to identify the correlation between the reaction rate constants, sample pH and estrogen molecular weight. Statistical analysis was performed using STATISTICA 10. Results: The results obtained in this study indicated that the rate of ultrasonic degradation in water and wastewater is influenced by the sample pH, oxidants and the dissolved organic content. More than 90% removal of all the estrogen hormones was achieved under basic pH conditions. The degradation process followed a pseudo-first order kinetic model with reaction rate constants ranging from 0.01-0.9 min–1. The degradation rate constants were observed to follow the order of deionized water>final effluent>secondary effluent>primary effluent. Conclusion: Optimization of the process conditions led to a 50-90% enhancement in the degradation kinetics, thereby making the process attractive for field applications.
INTRODUCTION
The discharge of endocrine disruptors from wastewater effluents1,2, animal and agricultural wastes3 has become a subject of many studies in the scientific community. Conventional wastewater treatment processes are not efficient in completely removing the estrogenic compounds. Complete removal of estrogen hormones is highly desired from wastewater effluents, especially for water reuse consideration, as estrogenic compounds are known to disrupt the endocrine system4. Exposure to low concentrations of estrogenic compounds, such as 0.1 ng L–1 of ethinyl estradiol have been reported to cause feminization of male fish5,6. Typically, estrogen hormones are excreted by humans in conjugated form, which are less potent compared to free estrogens7. These conjugated hormones convert to free estrogens during biological treatment processes and impart higher estrogenicity to the receiving waters. A survey conducted by Belgiorno et al.1 reported the presence of high concentrations of pharmaceuticals, such as carbamazepine, clofibrate, phenazone, ibuprofen, ketoprofen and naproxen in the effluents of municipal sewage treatment plants. Several treatment technologies, such as adsorption, ozonation, photolysis, sonication, membrane filtration and advanced oxidation processes (AOPs) have been examined for the removal of estrogenic compounds from water8-13.
AOPs generate highly reactive hydroxyl radicals that have the capability to destroy many organic contaminants that may be recalcitrant to conventional treatment processes. Sonolysis/sonication is an AOP that involves the passage of high intensity ultrasound waves through liquid medium, which creates cavitation. The acoustic energy causes physical and chemical reactions that can degrade contaminants present in the liquid medium. Implosion of cavitation bubbles create extreme pressures (100-50,000 bar) and temperatures (1000-5000 K) at microscopic points, which last for a few microseconds (2-10 μsec). Ultrasonic irradiation causes degradation of organic compounds by various mechanisms, including thermal decomposition, radical oxidation and supercritical oxidation. Cavitation also produces high mechanical shear stresses that are exerted on the organic chemicals in the liquid. The sonication process involves reactions in the cavity, at the interface and bulk solution14,15. Thermal breakdown of volatile substances occur in the gaseous phase and in the interfacial region. There is a release of free hydroxyl radicals due to pyrolysis of water, which cause chemical transformation in the bulk solution16,17.
Sonication was shown to be effective for wastewater disinfection18,19. There are limited numbers of studies showing the effectiveness of ultrasound for the removal of pharmaceuticals from wastewater20,21. However, there are no studies that show the effectiveness of sonication for removal of estrogen hormones from wastewater. This study fills the void by evaluating sonochemical oxidation of seven estrogen hormones in the presence of oxidizing agents (peroxide and persulfate) and at different solution pH. The seven estrogen compounds included in this study were 17α-estradiol, 17β-estradiol, 17α-ethinyl estradiol, 17α-dihydroequilin, estriol, estrone and equilin. This study builds upon our previous study that reported the influence of process conditions, such as alkalinity, salinity14 and reactor parameters such as power density, power intensity and reactor configuration22,23 on the removal of estrogen hormones. The presence of alkalinity showed an inhibitory effect, while the ionic strength favored the degradation process. A full-scale treatment system has been developed and implemented at a pharmaceutical manufacturing site for estrogen degradation. This study examined the effect of introducing oxidants such as persulfate and peroxide on sonochemical degradation of seven estrogen hormones. A detailed study on the effect of sample pH (2, 5, 7 and 10) was examined and the underlying mechanisms were explored. To the best of our knowledge, no other study has examined the effect of pH on estrogen hormone removal. The effect of pH was correlated with molecular weight of the estrogen hormones. The impact of oxidants at different pH was examined with different water matrices (Milli-Q, municipal wastewater primary influent, secondary effluent and final effluent).
MATERIALS AND METHODS
This study was carried out from August, 2008-December, 2010.
Chemicals, reagents and equipment: Estrogen hormones were obtained from Sigma Aldrich and Steroids Inc. The hormones (minimum purity) were as follows: 17α-estradiol (98%), 17β-estradiol (97.1%), estrone (100%), 17α-dihydroequilin (99.4%), 17α-ethinyl estradiol (99.1%), estriol (100%) and equilin (99.9%). The 3-O-methylestrone (98%) was used as an internal standard. The structure and properties of the estrogen hormones are shown in Table 1. Solvents (methanol and toluene, HPLC grade), concentrated HCl (ACS grade, 37.1%), NaOH (solid, ACS grade) and NaCl (anhydrous) were obtained from Fisher Scientific. Hydrogen peroxide (H2O2, 30% solution by weight) and sodium persulfate (Na2S2O8, >99.0%) were obtained from Sigma Aldrich and Fluka Chemicals, respectively. Varian Bond 3 mL/500 mg solid-phase extraction (SPE) cartridges were obtained from Varian Inc. Millipore nitrocellulose filters (0.45, 0.8 and 8.0 μm) and amber glass bottles were also obtained from Fisher Scientific.
| Table 1: | Structure and properties of estrogen hormones |
| Table 2: | Wastewater characteristics |
| ND: Non detect | |
Sample collection and storage: Wastewater influent and effluent samples were collected from a local municipal wastewater treatment plant (grab samples). The plant serves nine municipalities, two hospitals and six manufacturing industries, with 60% of the wastewater coming from household waste. Primary effluent was collected after the primary clarifier; secondary effluent was collected after the secondary clarifiers and final effluent was collected after the chlorination process. The samples were collected in silanized glassware (deactivated with 5% di-methyl-di-chloro-silane in toluene) and stored at 4°C prior to use. The wastewater characteristics are shown in Table 2.
Experimental setup: Batch ultrasonic irradiation experiments were performed using a 2 kW (20 kHz, 100% amplitude setting) sonication unit with immersed titanium alloy probes and horn attachment. The power output from the probe was displayed on the control unit of the sonicator. Sample volume used was 500 mL. During the sonication process, the effective probe area was completely immersed in the solution. Power output into the solution was approximately 325 W resulting in a power intensity of 135 kW m–2 and a power density of 0.65 W m–3. The reactor was operated at a constant temperature of 15-20°C by using a coolant bath. All of the experiments were conducted in duplicate unless specified (RSD<10%).
Estrogen extraction and analysis methods: All glassware, such as sample bottles, beakers, test tubes, disposable glass pipettes and gas chromatography inserts were silanized prior to use, following the method described in Andaluri et al.23. Analysis of estrogen hormones in aqueous phase was performed using solid phase extraction (SPE), followed by Gas Chromatography-Mass Spectrometry (GC/MS). The GC/MS analysis was performed using an Agilent 6890N GC equipped with an Agilent 5973MS (Agilent Technologies, Santa Clara, CA, USA) and a Pursuit DB-225 MS capillary column (30 m×0.25 mm ×0.25 μm). Details of the analytical method are described in Suri et al.24. Briefly, SPE was performed using the Varian Bond Elut 3 mL/500 mg C-18 adsorbent cartridge (Varian Inc.) followed by elution into methanol. The methanol eluent was completely dried using a Genevac EZ-2 evaporator and then derivatized using bis-(trimethylsilyl)-trifluoroacetamide.
Rate constants: The pseudo first order reaction rate constants for the removal of estrogen hormones were calculated using Eq. 1, where, Ct is the estrogen concentration at time t, C0 is the initial concentration, k is the first order reaction rate constant and t is sonication time:
| (1) |
Statistical analysis: To correlate the rate constants with solution pH and estrogen molecular weight, statistical analysis was performed using STATISTICA 10 software (TIBC StatisticaTM, Palo Alto, CA, USA).
RESULTS
Effect of solution pH: The selected estrogen hormones in this study have an acid dissociation constant (pKa) around 10.5 (Table 1) and predominantly exist in ionic form at pH greater than 10.5. To examine the effect of solution pH on estrogen degradation, sonolysis experiments were conducted at different solution pH (2, 5, 7, 8.5 and 10.5). Control experiments were performed at different solution pH without sonication and no significant change in estrogen concentration was observed for 30 min. In control tests, the relative standard deviation of estrogen concentrations was <8%, confirming that changing the pH alone (no sonication) does not influence the estrogen concentration.
The pseudo first order reaction rate constants (k, min–1) for the removal of estrogen hormones as a function of pH are shown in Fig. 1. The order of destruction under acidic, neutral and basic pH was 17α-dihydroequilin>equilin>17α-ethinyl estradiol>17α-estradiol>17β-estradiol>estrone>estriol. The rate constants were relatively higher at both acidic and basic pH when compared to the neutral pH. For example, the rate constants for the removal of equilin at pH 2, 7 and 10 were 0.3, 0.1 and 0.4 min–1, respectively. The rate constants for estrogen removal were generally higher at pH 10 when compared to acidic pH 2, except for 17α-dihydroequilin, which had a higher removal rate under acidic conditions (Fig. 1). Approximately 67-98% enhancements in the destruction rate of the estrogen hormones was observed at pH 10 compared to the neutral pH. At neutral pH, 45-95% removal of all the estrogens was observed after 30 min of sonication, however, at pH 10, a removal of >95% of all the estrogen hormones were observed within 10 min of sonication.
The relationship between solution pH, estrogen molecular weight and the degradation rate constant is shown in Fig. 2. The observed correlation is shown in Eq. 2, where MW is the molecular weight (g mol–1) and K is pseudo first order rate constant (min–1):
| (2) |
This correlation is valid for the conditions of this study and may not be applicable to other contaminants or reaction conditions. Under acidic conditions, the compounds with lower molecular weight showed a faster removal rate and under basic conditions the compounds with higher molecular weight showed faster removal as shown in Fig. 2.
Effect of hydrogen peroxide: Addition of hydrogen peroxide (H2O2) can have significant enhancement or inhibitory effects on the sonochemical destruction of organic compounds depending on dosage25-26. Prior to sonication, control experiments were performed to study the effect of peroxide on estrogen hormones in the absence of sonication. Control experiments were conducted by adding hydrogen peroxide (100 and 500 mg L–1) to the samples and were analyzed after 30 min of reaction time. In control tests, the relative standard deviation of estrogen concentrations was <5%, confirming that addition of peroxide (no sonication) does not oxidize the estrogen hormones. The degradation rate constants of estrogen hormones at pH 2, 7 and 10 in the presence of H2O2 are shown in Table 3. Experimental results show up to a 2-fold enhancement in the reaction rates in the presence of 100 mg L–1 of peroxide at pH 2 and 7, except for dihydroequilin at pH 7.
| Fig. 1: | Rate constants for the removal of estrogen hormones as a function of pH, Sample volume: 500 mL, Power density: 0.64-0.68 kW L–1, Power intensity: 135 kW m–2 |
| Fig. 2: | Degradation rate constant vs. sample pH and estrogen molecular weight (g mol–1), Sample volume: 500 mL, Power density: 0.64-0.68 kW L–1, Power intensity: 135 kW m–2 |
| Fig. 3: | Rate constants (min–1) for the removal of estrogen hormones in the presence of persulfate (S2O82–), Sample volume: 500 mL, Pd: 0.62-0.68 kW L–1, Pi: 135 kW m–2 |
| Table 3: | Rate Constants (min–1) and regression coefficients (r2) for the removal of estrogen hormones in the presence of H2O2 |
However, a 50% reduction in the reaction rate was observed at pH 10 in the presence of 100 mg L–1 peroxide. For example, the degradation rate constants for 17α-estradiol were 0.14 and 0.47 min–1 at pH 10 with and without 100 mg L–1 of peroxide, respectively (Table 3). Further addition of 500 mg L–1 of peroxide did not provide any enhancement in the reaction rates at neutral pH (Table 3).
Effect of persulfate: The degradation rate constants for the removal of estrogen hormones in the presence of persulfate (S2O82-) are shown in Fig. 3. At neutral pH (7), the addition of persulfate increased the removal of estrogen compounds. A reaction rate enhancement of 11-89% was observed in the presence of 10 mg L–1 of S2O82– and 52-94% enhancement was observed in the presence of 1000 mg L–1 of S2O82–. For example, the degradation rate constant for 17α-estradiol was 0.04 min–1 in the absence of S2O82– and 0.09 and 0.22 min–1 in the presence of 10 and 1000 mg L–1 of S2O82–, respectively (Fig. 3). Experimental results show a 12-23% reduction in the reaction rates at pH 10 in the presence of 10 mg L–1 S2O82–. For example, the degradation rate constant for 17α-estradiol was 0.47 and 0.38 min–1 at pH 10 in the presence of 0 and 10 mg L–1 of persulfate, respectively. A statistical analysis was performed for the variation of rate constant with respect to molecular mass and persulfate concentration at neutral pH 7. The 3D surface plot for rate constant with respect to estrogen molecular weight and persulfate concentrations is shown in Fig. 4. Faster degradation kinetics were observed for compounds with higher molecular weight, with increasing persulfate concentration (Fig. 4).
| Fig. 4: | Statistical analysis for the variation of rate constant with respect to molecular weight (g mol–1) and persulfate concentration, pH 7, Sample volume: 500 mL; Pd: 0.62-0.68 kW L–1, Pi: 135 kW m–2 |
Application of ultrasound for municipal wastewater treatment: The characteristics of the wastewater samples are shown in Table 2. All the samples were at neutral pH range (7.0-7.5) and contained low concentrations of estrogen hormones (7-157 ng L‾1). All the effluent samples were spiked with 10 μg L‾1 of estrogen hormones to obtain the degradation rate constants under various process conditions. Initial control experiments were performed to confirm the stability of estrogen hormones in wastewater. The presence of H2O2 or S2O82– in the absence of ultrasound did not show any significant removal of estrogen hormones from municipal wastewater.
The removal rate constants of estrogen hormones from primary, secondary and final effluents are shown in Table 4. The order of degradation rates was deionized water>final effluent>secondary effluent>primary effluent. Experimental results show a 20-82% enhancement in the reaction rate for the removal of estrogen hormones from wastewater matrices when the pH was increased from 7-10. Approximately 20-80, 30-70 and 50-82% enhancement in the reaction rates were observed in primary, secondary and final effluent, respectively (Table 4). Approximately 16-47, 41-65 and 63-91% removal of all the estrogens were observed at pH 10 in primary effluent, secondary effluent and final effluent, respectively, after 20 min of sonication.
Batch experiments were conducted for the removal of estrogen hormones from wastewater effluents in the presence of oxidants (H2O2 and S2O82–). The degradation rate constants of estrogen hormones in the presence of H2O2 and S2O82– from wastewater effluents are shown Table 4. A 20-70% enhancement in the reaction rates of estrogen hormones were observed in primary effluent, except for 17α-ethinyl estradiol, which showed a 50% reduction in the removal rate. A 12-50 and 35-78% enhancement in the reaction rates of estrogen hormones were observed in secondary (except for 17α-ethinyl estradiol) and final effluents (except 17α-dihydroequilin), respectively. Approximately, 6-71% reduction of reaction kinetics was observed in secondary effluent in the presence of persulfate (except for 17α-estradiol and equilin). However, an 80-90% enhancement was observed in the removal rate of estrogen hormones from final effluent in the presence of persulfate.
| Table 4: | Rate constant (regression coefficients, r2) for removal of estrogen hormones from primary, secondary and final effluents of municipal wastewater |
| -na-: Not available | |
DISCUSSION
This study focused on the application of ultrasound based advanced oxidation process for the removal of seven steroid hormones from water and wastewater. Sample pH plays an important role in the removal of estrogen hormones from water and wastewater. Higher removal rates were observed at both acidic and basic pH. All the estrogen hormones have phenolic group in the A-ring and are susceptible to higher degradation with ultrasound under acidic pH conditions. Estrogen hormones exist in undissociated form below pH 10 and have higher hydrophobicity. This would facilitate the molecules to diffuse onto the interface zone of the cavitation bubble, resulting in thermal destruction upon cavity collapse. These results are similar to previously reported studies, where phenolic compounds showed higher destruction under acidic conditions26,27. However, at pH>10, estrogen hormones exist in ionic form and exhibit higher solubility in aqueous phase and the destruction would be primarily due to radical oxidation as opposed to reaction at the interface only. Since most of the reaction is radical mediated in the bulk region, the estrogens undergo higher degradation kinetics at alkaline pH 10. Similarly, Serna-Galvis et al.28 reported 45% enhancement in the degradation rate of the pharmaceutical, fluoxetine at higher pH (pH 11). This was attributed to the higher presence of hydroxyl radicals. However, Villaroel et al.29 have reported lower efficiencies for the removal of acetaminophen at basic pH (pH 12). This was due to the partial ionization of OH* radicals and low accumulation of hydrogen peroxide at pH 1229.
Presence of oxidizing agents also plays a significant role in the reaction kinetics of the ultrasound process. Addition of hydrogen peroxide results in the formation of highly reactive hydroxyl radicals (OH*) and result could result in reaction enhancement. However, the addition of excess H2O2 results in OH* radical scavenging. The generation and scavenging of radicals are shown in Equations 3 and 430,31, respectively:
| (3) |
| (4) |
Peroxide is a moderate oxidizer and is known to be stable under acidic conditions. As the pH increases, peroxide stability decreases and is known to be highly unstable under alkaline pH (safety and handling technical data sheet). Under alkaline conditions, peroxide dissociates into water and oxygen as shown in Eq. 5. The excess oxygen formed diffuses in to the cavity during the sonication process and cushions the adiabatic collapse, resulting in lower efficiency:
| (5) |
The results obtained in this study were similar to the previously reported studies30,31. Lim et al.30 reported enhancement in the removal of phenol and bisphenol-A in the presence of 0.01-10 mM of peroxide. Similarly, Dukkanci et al.31 reported up to a 2-fold rate enhancement in the presence of 1-5 mM hydrogen peroxide. However, further increasing the peroxide concentration to 6 mM resulted in a significant reduction of the reaction rate due to scavenging nature of the hydrogen peroxide towards OH* radicals when present in higher concentrations31.
Persulfate is a strong oxidizing agent with a redox potential of 2.01volts and also a stable chemical. It can remain in the environment for a long time. The persulfate ions produce sulfate radicals and these free radicals react with water to form OH* under sonication32, as shown in Eq. 6 and 7:
| (6) |
| (7) |
Under basic pH conditions, the reaction in Eq. 6 will be pushed to the left, thereby slowing down the oxidation of hormones. This inhibition was similar to when peroxide was added to the samples at pH 10. Li et al.33 reported a reduction in the removal rate of 1,1,1-Trichloroethane (TCA) in the presence of excess persulfate at neutral pH and suggested that an optimal dosage is necessary for enhanced removal rates. Presence of excessive amounts of persulfate could inhibit the generation of OH* radicals resulting in a reduction in overall removal efficiency33.
Application of the optimized process conditions to the wastewater matrix enhanced the degradation rate kinetics. However, the degradation rates did not increase in the same ratio as that in de-ionized water. The rate constants were higher for final effluent when compared to secondary and primary effluent. This indicates that the background matrix plays a significant role in the ultrasound-induced degradation of estrogen hormones. The background matrix consumes hydroxyl radicals thereby reducing the number of radicals available to react with the estrogen hormones. Similar results were reported by Ifelebuegu et al.34, where the mass degradation rates of 17β-estradiol and 17α-ethinyl estradiol reduced with increasing concentrations of dissolved organic carbon. It was also reported that the presence of low alkalinity does not affect the removal of estrogen hormones, however, alkalinity higher than 120 mM HCO3‾ can slowdown the kinetics of estrogen hormone removal during sonolysis14.
The results of this study could help to determine the optimum conditions for the sonochemical removal of estrogen hormones. Additionally, the results of this study could be used for designing ultrasound systems for wastewater treatment that could help solve the current issues of the existence of estrogen hormones in wastewater and simultaneously provide disinfection. Pilot scale application of this treatment technology remains an area for future study.
CONCLUSION
Ultrasound process was successfully developed for the removal of estrogen hormones from water and wastewater. Removal of estrogen hormones from wastewater appears to be primarily driven by hydroxyl radical oxidation. Optimization of the process conditions resulted in up to a 5-fold enhancement in the reaction rate kinetics in both DI water and wastewater. Statistical 3D models were generated to estimate the reaction rate kinetics with respect to process conditions. This model may be useful to estimate the removal kinetics of the estrogen hormones. This study provides valuable insights on the effect of pH and size of the estrogen molecules on the degradation rates.
SIGNIFICANCE STATEMENTS
This study discovers the possible synergistic effects of ultrasound and oxidizing agents that can be beneficial for the removal of steroid hormones from wastewater effluents. This study will help researchers to address the critical issue of presence of emerging contaminants in wastewater effluents that was not explored. Public utilities and water/wastewater industries could benefit from the verification of in-line advanced oxidation processes (AOPs). As such, it helps the design of AOPs for wastewater treatment.
ACKNOWLEDGMENT
Amrutha Shah for Laboratory assistance is acknowledged. The National Science Foundation (NSF)-Water and Environmental Technology (WET) Center funded (Grant ID: NSF-WET_TU-09-3) this project. Opinions, findings and conclusion expressed in this article are those of the authors and do not necessarily reflect the views of NSF-WET center of Temple university.