Next Article in Journal
Trends, Cycles, and Spatial Distribution of the Precipitation, Potential Evapotranspiration and Aridity Index in Xinjiang, China
Next Article in Special Issue
Research on Micropollutants in Urban Water
Previous Article in Journal
Macroplastics in Lakes: An Underrepresented Ecological Problem?
Previous Article in Special Issue
Occurrence and Removal of Priority Substances and Contaminants of Emerging Concern at the WWTP of Benidorm (Spain)
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:

Experimental Design and Bioassays as Tools to Investigate the Impact of Anodic Oxidation on Progestins Degradation

Juliana Mendonça Silva de Jesus
Allan dos Santos Argolo
Flávio Kiyoshi Tominaga
Maria Elena Taqueda
Daniele Maia Bila
Sueli Ivone Borrely
3 and
Antonio Carlos Silva Costa Teixeira
Research Group in Advanced Oxidation Processes (AdOx), Department of Chemical Engineering, University of São Paulo, Av. Prof. Luciano Gualberto, 380, São Paulo CEP 05508-010, Brazil
Department of Sanitary and Environmental Engineering, Rio de Janeiro State University, Rio de Janeiro CEP 20550-900, Brazil
Nuclear and Energy Research Institute (IPEN-CNEN/SP), Av. Prof. Lineu Prestes, 2242, São Paulo CEP 05508-000, Brazil
Authors to whom correspondence should be addressed.
Water 2023, 15(1), 61;
Submission received: 6 November 2022 / Revised: 20 December 2022 / Accepted: 21 December 2022 / Published: 24 December 2022
(This article belongs to the Special Issue Research on Micropollutants in Urban Water)


The present study investigated the degradation of levonorgestrel (LNG) and gestodene (GES) through an anodic oxidation process mediated by active chlorine species. The independent variables [LNG]0 and [GES]0, current density (mA cm−2), and [NaCl]0 (mol L−1) were optimized through a response surface methodology (RSM) based on a four-level central composite design (CCD). Specific energy consumption allowed CCD-RSM analysis and optimization. The decay of progestins was followed to verify the kinetics of the anodic degradation process. Chlorine monitoring showed that excess Cl concentration did not mean high hormones removal, as well as the excess of current density. Central point conditions ([NaCl]0 = 0.07 mol L−1, j = 32.5 mA cm−2, [LNG]0, and [GES]0 1.0 mg L−1) proved to be the best operational option. The performance with real pharmaceutical wastewater confirmed model optimization (2.2 ± 0.2 kWh g−1, with removals of 83.1 ± 0.9% and 75.1 ± 2.8% for LNG and GES, respectively). The selected condition was used for estrogenic activity and acute toxicity assays. The first allowed the identification of the initial estrogenic activity for the mixture of LNG and GES (924 E2-EQ ng L−1). Additionally, the electrochemical process could decrease this environmental parameter by 74.6%. The progestin mixture was classified as acute toxicity to Daphnia similis, with a toxicity unit (TU) of 2.5 100/EC50%. After electrolysis, the hormone solutions reached a fourfold increase in TU value, classified as high acute toxicity.

1. Introduction

The fate and occurrence of endocrine disrupting compounds (EDCs) promoted an emerging concern worldwide. As a result of anthropogenic actions, EDCs have been found in distinct environmental matrices, such as water, soil, and air. Among them, synthetic hormones, such as progestins, have been used in ecotoxicological assays to highlight their hazard to Dreissena polymorpha [1], fathead minnow [2], and zebrafish [3].
Levonorgestrel (LNG) and gestodene (GES) are progestins used in hormonal maintenance via contraceptive pills and intrauterine devices, whose consumption has been increasing yearly. According to the United Nations [4], 151 million women worldwide chose the contraceptive pill as their primary contraceptive method; in Brazil, 2.6 million emergency contraceptive pills (ECPs) were consumed in 2017 [5]. ECPs have LNG as their only active principle (1.50 mg), besides using GES and LNG in a combined formulation with estrogens, in doses of 0.075 to 0.15 mg, respectively [6].
Pharmaceutical production and consumption levels are related to the number of substances discharged into environmental matrices (Table 1). The primary sources of micropollutants are household, pharmaceutical, and livestock farm runoff, which flow out into wastewater treatment plants (WWTP) [7,8,9,10]. Administered hormones are not fully metabolized by humans and animals. In human metabolism, about 77% and 10% of ingested LNG and GES are eliminated in their active form via urine, respectively [8,9].
In the last 26 years, studies have investigated electrochemical advanced oxidation processes (EAOPs) to remove hormones from water matrices, among which hormones have been used as the target pollutant. A brief search in the Scopus and Web of Science online platforms using the keywords “anodic oxidation” AND “hormones” showed 24 and 28 articles. In this sense, Table 2 presents the main studies evaluating the effectiveness of anodic oxidation for the degradation of hormones in synthetic and real matrices.
Based on that, the electrodes applied to the degradation of organic compounds are classified as either active or non-active. The first is characterized by the strong affinity of HO radicals (the primary oxidizing species in AOPs) with the surface of the electrode (M), reducing the number of radicals that will be available to directly react with the target contaminant (Equations (1) and (2)). This behavior is observed for RuO2, IrO2, and Pt, for example [24,25,26].
H2O + M → M(OH) + H+ + e
M(OH) → M + ½ O2 + H+ + e
In contrast, electrodes such as SnO2 and boron-doped diamond (BDD) can easily generate and release OH radicals to the solution, improving degradation and potentially leading to the total mineralization of the contaminants (R(aq)) [24,26,27,28]:
R(aq) + M(OH)n/2 → M + transformation products + n/2 H+ + n/2 e
In addition to hydroxyl radicals, EAOPs can mediate the generation of other strong oxidants, such as chlorine (Cl), persulfate (S2O82−), or superphosphate (P2O84−). This occurs in parallel to the electron generation of HO [24,25,26].
However, to obtain the maximum performance of contaminant removal by EAOP, it is necessary to identify the main factors that affect these processes, such as the initial concentration of the target contaminant(s), the concentration of oxidants, and the current density. All these parameters are modified depending on the process studied and the matrices [29]. Furthermore, the best operating condition must achieve not only high removal rates but also low energy consumption and little generation of toxic by-products. To this end, EAOP studies are usually performed together with ecotoxicological, cytotoxic, and estrogenic activity tests. Such experimental proposals contribute to motivating the use of EAOPs, and under some conditions, the combined oxidative process reaches the potential of pollutant mineralization and toxicity removal [30,31,32].
Bioassays have been used to confirm the effectiveness of removing hazardous chemicals from target matrices [33,34,35]. For example, the in vitro yeast estrogen screen (YES) assay has been applied to environmental samples. The advantages of YES include simple, easy handling; low costs; non-invasive, non-sample pre-treatment; and the possibility of dealing with trace concentration levels [36,37]. YES is based on the yeast Saccharomyces cerevisiae modified with a human estrogen receptor, which quantifies the total estrogenic potential of samples, considering the synergistic effects of multiple substances [37]. In complex environmental matrices, the matrix effect can inhibit the estrogen receptor, and, consequently, the response. However, cytotoxicity (yeast growth inhibition) and the presence of antagonistic substances are other pieces of evidence obtained from the assay. Estrogenic activity can still be determined [36,37,38].
In turn, ecotoxicity assays have been used to improve the risk assessment network on emerging contaminants, including hormones. Standard procedures and test organisms have been applied, e.g., Daphnia similis [39], Daphnia magna [40], and Vibrio fischeri [41]. Evaluation of the toxicity of organic compounds in aqueous matrices before and after treatment is recommended to motivate the inclusion of such compounds in current environmental legislation and highlight viable possibilities for ecological treatment and monitoring [41,42,43].
In this context, the present work aimed to optimize and determine the experimental conditions that allow low specific energy consumption (SEC) and removals ≥70% of both progestogens in synthetic solutions and real pharmaceutical wastewater (RPW). To this end, the experiments were conducted following a central composite experimental design coupled with response surface analysis. The study of hormones concentration as an independent variable, in the mixture, is a novelty among the various studies that address the optimization of advanced electrochemical processes [35,44,45], considering the future application of the process in an industrial environment where the concentration of such compounds is not fixed a priori. From this, it was possible to determine a work region in which lower energy consumption and high removals of both hormones can be obtained, close to the central conditions of the experimental domain. Additionally, the study brings the ecotoxicological verification of such conditions, which if applied, can provide a reduction in the initial estrogenic activity. In brief, the main objectives were: (i) to investigate LNG and GES degradation in a synthetic mixture and real pharmaceutical wastewater, through anodic oxidation; (ii) to evaluate the effects of initial progestin concentrations, current density, and NaCl concentration through a central composite experimental design; (iii) evaluate the anodic process through acute toxicity tests with Daphnia similis; and (iv) to assess the estrogenicity of progestin samples, before and after the anodic process, using the yeast estrogen screen (YES) assay.

2. Materials and Methods

2.1. Reagents

LNG (C21H28O2, ≥ 98.0%, CAS 797-63-7) and GES (C21H26O2, ≥ 98.0, CAS 60282-87-3) were purchased from Zhejiang Xianju Pharmaceutical Co. Ltd. and used as analytical standards and in the preparation of working solutions. In addition, 17 β-estradiol (E2, C18H24O2, ≥ 98%, CAS 50-28-2) was purchased from Sigma-Aldrich. Chlorophenol red-β-D-galactopyranoside (CPRG), sodium chloride (NaCl, ≥99.0%), and sodium thiosulfate pentahydrate (Na2S2O3·5 H2O, ≥ 99.0%) were purchased from Merck, Casa Americana Ltd., and Vetec Ltd., respectively. HPLC grade methanol (CH3OH), ethanol (C2H5OH), and acetic acid (CH3CO2H) were purchased from Sigma-Aldrich. All solutions were prepared using ultrapure water (18.2 MΩ cm) from a Milli-Q® system (Merck Millipore, Burlington, MA, USA).

2.2. Electrochemical Degradation Studies

An undivided electrochemical cell was used, containing DSA® (Ti/Ru0.3Ti0.7O2, De Nora do Brasil Ltd., Brazil) and stainless-steel electrodes as the anode and cathode, respectively. With 28 cm−2 of the active area, both electrodes were parallel at 13 mm. The power supply (30V/5A MPS 300-5B MINIPA) was used as a constant current source. NaCl was used as the electrolyte. The solutions (300 mL) were recirculated from a borossilicate glassreservoir to the electrochemical cell with the aid of a peristaltic pump (DM 500) in a continuous flow of 0.75 mL s−1 (Figure S1). The internal volume of the electrochemical cell was 36.4 mL, and all experiments were performed at a space-time of 16 s in each pass. The conditions of the experiments (NaCl and progestin concentrations; current density) are summarized in Table 3. Progestin solutions were sampled at 0, 2, 4, 6, 8, 10, 15, and 20 min for analysis. The sample volume was 900 L, to which 100 µL of Na2S2O3 solution (10 g L−1) was added to stop any further reaction [46].

2.3. Analytical Methods

Ultra-fast liquid chromatography (UFLC) was performed using a Shimadzu (SPD 10 AV) device equipped with a UV-visible detector (SPD 20A, Shimadzu, Kyoto, Japan) and a C18 column (ACE, 250 mm × 4.60 mm, 5 µm, ThermoFisher Scientific, Waltham, MA, USA), for the simultaneous monitoring of GES and LNG concentrations. An isocratic method was applied, with 70% of MeOH and 30% ultrapure water containing 1% v/v of acetic acid as the mobile phase.
LNG and GES were detected at 244 nm. The sample injection volume, oven temperature, and flow rate were 20 µL, 40 °C, and 0.2 mL min−1, respectively. Under these conditions, the retention times of LNG and GES were 10.0 and 8.0 min, respectively. Due to the low water solubility of LNG (Table S1), stock solutions of LNG and GES (10 mg L−1) were prepared in methanol and used to spike progestin standards in concentrations of 0.05–10.0 mg L−1.
A calibration curve was obtained, and validation parameters were determined, where LOD and LOQ refer to detection and quantification limits, respectively. For LNG: (R2 = 0.9991, LOD = 20 µg L−1 and LOQ = 70 µg L−1); for GES: (R2 = 0.9980, LOD = 60 µg L−1, LOQ = 200 µg L−1). The efficiency of progestin removal was calculated by Equation (4):
p r o g e s t i n % = C 0 C t C 0 100
where C 0 and C t refer to initial and final LNG and GES concentrations (mg L−1), respectively.

2.4. Central Composite Design (CCD)

Considering the standard order of the CCD approach, 31 experiments were conducted based on seven replications of the central point, eight-star points, and 18 factorial points in a random sequence. Minitab 20 was used for regression and analysis of the data obtained. ANOVA allowed the evaluation of polynomial response surface models (RSM) at a 95% confidence interval. In this investigation, parameters such as F-value, p-value, and R2 were applied for data confirmation and accuracy. The F-value was the ratio of the parameter’s variance to the error variance (residual variance) or the ratio of the mean square parameter to the mean square error. RSM was applied for identifying the optimum operating conditions for the degradation of LNG and GES, aiming to minimize energy consumption.
CCD was applied for evaluating the effect of four independent variables: initial NaCl concentration (mol L−1), current density j (mA cm2), and initial LNG and GES concentrations (mg L−1). Table 3 presents these parameters, converted to dimensionless values coded as X1, X2, X3, and X4 at five levels (−2, −1, 0, +1, and +2). The variable levels were selected based on the progestins’ solubility (Table S1) and previous studies [47].
A second-order polynomial multiple regression equation was applied to represent the relationship between the response (SEC, specific energy consumption) and the operational parameters. The corresponding quadratic equation, along with the interaction effects of independent variables, is depicted in Equation (5):
S E C   kWh   g 1 = β 0 + i = 1 n β i X i + i = 1 n β ii X i 2 + i = 1 n 1 j = i + 1 n β ij X i X j +   ε  
where SEC is the predicted response variable; β0 is the intercept; βi, βij, and βii are the coefficients of the linear effect and double interactions; and Xi and Xj refer to the independent variables or factors. The specific energy consumption (SEC) is defined as the amount of energy consumed in kWh to degrade 1 g of organic load from an effluent—in this case, progestins, according to Equation (6):
S E C   kWh   g 1 = UIt Δ progestin V
where U is the average potential of the electrochemical cell (V); t is the electrolysis time (h); I is the electric current (A); Δ[progestin] corresponds to the variation in LNG and GES concentrations (g L−1); and V is the volume of the working solution (L).

2.5. In Vitro Estrogenic Assay—YES

The estrogenicity of progestin samples was assessed before and after anodic oxidation with YES, an in vitro recombinant reporter gene assay [48]. Before the YES assay, progestin samples were concentrated through C18 cartridges (SPE Strata 200 mg/3 mL).
The SPE protocol consisted of the following steps: (i) cartridge conditioning with 10 mL of methanol and 10 mL of pure water; (ii) percolation of progestin samples at 4 mL min−1; (iii) rinse with 10 mL of methanol (2% v/v); and (iv) elution of LNG and GES analytes with 2 mL of ethanol. LNG and GES extraction recoveries were (91.60 ± 0.04) % and (84.90 ± 2.47) %, respectively.
The bioassay was performed in 96-well microplates with serial dilution of sample extracts in ethanol. E2 was used as the positive control and standard curve, 2724 to 1.33 ng L−1, and ethanol was used as the negative control. An amount of 10 μL of each sample dilution was transferred to a test plate and allowed to evaporate. Then, 200 μL of culture medium containing yeast and the chromogenic substrate chlorophenol red-β-D-galactopyranoside (CPRG) were added. After incubation for 72 h at 30 °C, absorbances at 575 nm and 620 nm were measured with a VersaMax microplate reader (Molecular Devices, San Jose, CA, USA).

Data Analysis

Measured absorbances were corrected with Equation (7) to discount the turbidity effect from the estrogenic response:
A b s c o r r   s a m p l e = A b s 575   s a m p l e A b s 620   s a m p l e A b s 620   b l a n k s
Absorption and corrected concentration data were plotted, and the resulting sigmoidal curves were fitted to a symmetric logistic function using the software Origin 2020 (OriginLab). The estradiol equivalent (E2-EQ) results were obtained in ng L−1 by interpolating the standard E2 dose-response curve (Figure S3) and the sample data with a log-logistic model:
y = A 1 A 2 1 + x 0 / x p + A 2
where A1 and A2 refer to the maximum and minimum β-galactosidase induction in corrected absorbance, x0 is the median effect concentration EC50% for E2 in ng L−1, p corresponds to the slope of the sigmoidal curve, and (x, y) is the ordered pair related to a sample concentration and its response in corrected absorbance. Finally, the E2-EQ was calculated as the lowest x that elucidated an agonist response divided by the final sample enrichment factor in the assay [38].

2.6. Ecotoxicity Assays

Acute toxic assays were performed with the microcrustacean D. similis according to the Brazilian standard method, ABNT NBR 12713/2016 [49]. Daphnids were cultivated at the Laboratory of Biological and Environmental Assays (Nuclear and Energy Research Institute-IPEN-CNEN/SP, São Paulo, Brazil). The assays evaluated the effects of progestins in the mixture on the test organisms, before and after anodic oxidation conducted under central point conditions (Table 3). The immobility of D. similis after 48 h was the endpoint measured for this assay. For this purpose, neonates between 6 h and 24 h of age were exposed to several dilutions of treated and non-treated progestin solutions for 48 h.
The following solutions were evaluated through acute toxic assays: (A) a mixture of progestogens (LNG + GES) without electrochemical processing and in the absence of NaCl; (B) a mixture of progestogens (LNG + GES) without electrochemical processing in the presence of NaCl; (C) a NaCl solution without electrochemical processing; (D) a NaCl solution after electrochemical processing; and (E) a mixture of progestogens (LNG + GES) after electrochemical processing in the presence of NaCl. All samples had pH values corrected to 7.0 before toxicity assays, as requested by ABNT NBR 12713/2016 [49]. Acute toxicity is expressed in toxicity units (TU = 100/EC50%) and corresponds to the average effect concentrations that promoted 50% immobility of exposed living organisms (EC50%-48 h, expressed in % v/v).

3. Results and Discussion

3.1. Hormones Degradation and Response Surface Modeling

Table 4 presents the 31 experiments carried out in the standard CCD order, together with the values of SEC, percentage removals, and specific removal rates obtained for LNG and GES after 6 min of electrochemical treatment, with their respective coefficients of determination. Among the experimental conditions considered to evaluate the effectiveness of the anodic oxidation in the joint removal of progestins, runs 19 and 17 brought the removal and determination of the specific energy consumption for the treatment of solutions containing only LNG or GES, respectively. As a result, the lowest SEC of the entire experimental design (0.4 kWh g−1) was obtained in experiment 19, when the electrochemical system was applied to treat the solution containing 1.0 mg L−1 of GES with [NaCl]0 = 0.07 mol L−1 and current density of 32.5 mA cm−2. In addition to the minimum SEC, 96.4% GES was removed, with a specific degradation rate equal to 0.493 min−1. On the other hand, in run 17, performed only with LNG under the same experimental conditions, an SEC value eleven times greater than that observed for the GES solution was obtained. Furthermore, the percent removal and kobs for LNG were 78.1% and 0.212 min−1, respectively; thus, these results indicate different degradation behaviors exhibited by the hormones in relation to the anodic electrochemical process.
A closer inspection of the results in Table 4 shows that for experiments 1 and 5, in which the progestogen concentrations were 0.5 mg L−1 and the NaCl concentration was 0.04 mol L−1, it is observed that the effect of variable X3 (j) on hormone degradation resulted in higher LNG and GES removals (84.2% and 100%, respectively) for the lowest current density, 20 mA cm−2. On the other hand, for experiments 4 and 8, which started with concentrations of progestins three times higher (1.5 mg L−1) and the same concentration of electrolyte, the opposite effect was observed; that is, for higher initial concentrations of hormone it was necessary to use high values of current density (j = 45 mA cm−2) to achieve greater removal of hormones. The same is observed when comparing experiments 9 and 13, conducted with the lowest hormone concentrations, however, with a concentration of electrolyte 2.5 times higher. These results suggest that: (i) For initial concentrations of progestins ≤ 1.0 mg L−1, the current density has a positive impact on the removal; however, when exceeding such concentration, this density is insufficient. That is, the process is being operated above the threshold current density. Periyasamy et al. [50] report that increasing the current density increases the voltage, favoring the evolution of oxygen and disfavoring the evolution of chlorine. In this case, the oxygen generated in the electrode results in a reduction in the efficiency of the removal of organic molecules, inhibiting the transfer of mass to the electrode surface. (ii) There is a strong interaction between the variable’s current density and electrolyte concentration, as well as between the current density and the concentration of progestogens (in particular gestodene) so that the effect of increasing the current density will markedly depend on the levels of [NaCl]0 and [GES]0. In fact, for intermediate concentrations of LNG and GES, electrolyte content, and j = 32.5 mA cm−2 (central point), satisfactory SEC values (1.9 ± 0.2 kWh g−1) and progestin removals greater than 70%, with good specific removal rates (0.270 ± 0.1 min−1 and 0.242 ± 0.1 min−1 for LNG and GES, respectively) were obtained, in line with the study objectives.
The CCD-RSM was applied as a tool to identify and optimize the effects of the four main operating variables on the anodic oxidation of LNG and GES in a mixture using the DSA-Cl2 system to determine the experimental region that provides the lowest SEC and progestin removals ≥70%. Thus, a quadratic equation (Equation (9)), describing the mutual relationships between the experimental parameters and the response (SEC) was obtained from the significant effects (coefficients with p < 0.001), also observed through the Pareto chart (Figure S2).
SEC   kWh   g 1 = 2.1916 0.4777 X 1 + 1.8254 X 2 0.5044 X 3 0.9801 X 4 + 0.1399 X 1 2 + 0.2701 X 2 2 + 0.3016 X 3 2   + 0.6106 X 4 2   0.2407 X 1 X 2 0.1635 X 1 X 4 0.3329 X 2 X 3 0.7170 X 2 X 4 + 0.5627 X 3 X 4
where X1, X2, X3, and X4 are the coded experimental parameters; (X1)(X2), (X1)(X4), (X2)(X3), (X2)(X4), (X3)(X4) correspond to the interactions of quadratic terms; and (X1)2, (X2)2, (X3)2 and (X4)2 are the second-order terms. The ranges and levels of the experimental parameters are shown in Table 3. The predicted specific energy consumption (SEC) observed in progestin removal versus the experimental data is shown in Table 4. Furthermore, the ANOVA was implemented to determine the significance and adequacy of the statistical method (Table 5).
The main observation on the model equation (Equation (9)) and the Pareto chart (Figure S2) is associated with the variable X2 ([LNG]0), given the significance of first-order effects, second-order effects, and interactions. In the first two cases, the effects were positive; that is, the increase in the concentration of LNG contributes to the increase in energy consumption. On the other hand, Figure S2 highlights the non-significant interaction effect of variables X1 ([GES]0) and X3 (j), with no impact on specific energy consumption, which was not included in Equation (9).
The desirable value for “adequate precision” is F-value > 4.0. As shown in Table 5, the value obtained for “adequate precision” was 188.5. In addition, the fitted regression model was generally significant with a p-value <0.001, which was lower than 0.004, and the lack of fit (p-value = 0.004) was insignificant, indicating the suitability of the statistical model.
The evaluation of the regression model was also performed using Figure 1. As can be seen, Figure 1a indicates a strong correlation between model predictions and experimental values. Regarding the raw residuals (Figure 1b), the symmetrical distribution of the data and the clustering tendency in the middle of the plot is further evidence to confirm the quality of the model.

3.1.1. Effect of Operating Parameters

The CCD-RSM approach allows for the possibility of examining and interpreting both the underlying effects and their interactions. This interpretation was made using nonlinear forms in a multidimensional space. It is possible to interpret the effects of parameters on steady levels of the target variable using the response surface procedure (Figure 2).
As shown in Figure 2a, at the central levels of j = 32.5 mA cm2 (X3) and [NaCl]0 = 0.07 mol L−1(X4), the average response (SEC) was estimated to increase for high levels of X3 and low levels of X4. This observation was also verified using the model (Equation (9)), in which the coefficient of the interaction between these variables had a positive effect on the response (+ 0.5627X3X4) with a p-value < 0.001, indicating significant interaction.
The anodic oxidation in the presence of chloride anions leads to the formation of chlorine, hypochlorous acid, and hypochlorite, depending on the pH (Equations (11)–(13)), thus oxidizing organics on the electrode interface and/or in the bulk of the solution (Equation (13)) [25,26].
2   Cl aq     Cl 2 + 2 e aq
Cl 2 +   H 2 O     HOCl aq +   Cl aq + H aq +
HOCl aq     H aq + +   OCl aq
Organics +   OCl aq   intermediates   CO 2 +   Cl aq +   H 2 O       ( alkaline   medium )
For that reason, the use of low [NaCl] promotes an increase in SEC, due to the limited number of ions to be converted into reactive species and to the low conductivity of the solution, which provides an increase in the electric resistance of the system. For example, experiment 5 was conducted with all variables at their negative level, except X3; in experiment 13, the coded values of X1 and X2 were negative and X3 and X4 were positive. Under these conditions, SEC values ranged from 10.06 to 6.44 kWh g−1; that is, the increase in [NaCl] provided a twofold decrease in specific energy consumption.
Knowing the evolution of the concentration of Cl ions is necessary to test the hypothesis: high values of current density improve the electro generation of active species. Table 6 presents the consumption of Cl ions after 20 min of electrolysis for different process conditions, as previously established by the CCD matrix (Table 4).
According to Table 6, the anodic oxidation of progestins provided a consumption of Cl as a function of the initial concentration of NaCl and the applied current density. For example, runs 1 and 19 were performed with low values of j = 20 and 7.5 mA cm−2, respectively. In both experimental conditions, the electrolyte was applied at the lowest level of the empirical domain (0.04–0.07 mol L−1). Consequently, runs 1 and 19 consumed 100% of the Cl present in the solution to electrogenerated active species providing average progestin removals of 100% and 96.4%, respectively.
In contrast, run 20 was performed with a high level of X2 and an intermediate level of X4. In this condition, the maximum consumption of chlorine ions was 6.5%. Furthermore, the progestin removals achieved were 45.4% and 87.7% for LNG and GES, respectively. These results confirm the significant effect interaction among all studied variables and highlight that the use of high current densities does not always promote high generation of reactive species and consequently high removals of the target pollutants.
As can be seen in Figure 2b, at the central and constant levels of [LNG]0 (1.0 mg L−1) and [NaCl] (0.07 mol L−1), the SEC tends to be a minimum. The interaction coefficient between the variables showed a significant effect due to the p-value < 0.001, which was also observed in Equation (9), where the interaction coefficient was 0.7170 X 2 X 4 , indicating that to obtain a low SEC, it is necessary to operate the electrochemical system at low concentrations of both variables.
The effect of progestin concentrations on SEC can be observed in Figure 2c which has their levels set to central and constant levels. The use of low concentrations of LNG and GES (lower than −1.5 at the coded level) contributes to the increase in SEC. Nonetheless, the proportional increase of both progestins favors the process, and thus the specific energy consumption can be controlled. The negative interaction coefficient confirms this assumption, being a significant effect (p-value < 0.001).

3.1.2. Definition of Optimum Condition

Based on this set of evidence, it was possible to determine the optimal point of the experimental design (Table 7). However, for the concentrations of GES and NaCl used, the conditions suggested as optimal are outside the experimental domain, i.e., despite the practical possibility of these conditions, the result to be obtained will be outside the 95% statistical confidence interval, considered throughout the analysis.
Given that, the central point was chosen as the practical optimum point, corresponding to an SEC value of 2.2 ± 0.2 kWh g−1 and removal efficiencies of 82.0 ± 0.7% for LNG and 75.0 ± 0.5% for GES (Table 4). The accuracy of the model was also verified through two experiments carried out with real pharmaceutical wastewater (RPW) under optimal practical conditions. As result, the anodic oxidation treatment of RPW samples under central point conditions ([NaCl]0 = 0.07 mol L−1, j = 32.5 mA cm−2, [LNG]0 = [GES]0 = 1.0 mg L−1) resulted in an SEC of 2.32 ± 0.03 kWh g−1 and removal efficiencies of 83.1 ± 0.9% for LNG and 75.1 ± 2.8% for GES. Furthermore, considering the average price of USD 0.13 per kWh in Brazil [20], the estimated cost for treating wastewater containing LNG and GES is equivalent to about USD 0.26 m−3, under the conditions of the central point (Table 7). The motivation is the future implementation of anodic oxidation in wastewater treatment plants. In addition, the application of electrochemical oxidation technologies in other point sources, such as in the treatment of hospital effluents [51], pharmaceutical effluents [52], and agro-industrial effluents [53], can be seen as an opportunity to reduce the impacts caused by organic contaminants of emerging concern.

3.2. Estrogenic Activity Removal

The investigation of estrogenic activity was conducted by the YES assay as described in Section 2.5. E2 was used as a standard for obtaining a dose-response curve (Figure S3). Consequently, the sample treated under central point conditions (run 28, [NaCl]0 = 0.07 mol L−1, j = 32.5 mA cm−2, [LNG]0 = [GES]0 = 1.0 mg L−1) was used in the assay, which was also represented by the dose-response curve (Figure S4).
Figure 3 presents the removal of estrogenic activity observed for the LNG + GES solution before the AO process which reached a value of 924 E2-EQ ng L−1. This value was considered high compared with E2 which had an EC50 of 49.5 ng L−1. In addition, Figure 3 shows the efficiency of the AO process in reducing the initial estrogenic activity of the sample by 74.6%, achieving an average of 234 E2-EQ ng L−1. This result suggests that the degradation of progestin molecules did not form organic by-products with functional groups associated with the estrogenic activity of the parent compound, nor phenolic by-products after the anodic oxidation process.
Saggioro et al. [54] investigated the use of UV/chlorine to degrade bisphenol-A (BPA), a known endocrine disruptor that also has estrogenic activity. First, untreated solutions were evaluated ([BPA]0 = 100 µg L−1 and [Cl]0 = 2 mg L−1) showing ~280 µg L−1 E2-EQ. After 2 and 5 min of UV irradiation, both reached estrogenic activities below 26.0 ±12.0 ng·L−1 EQ-E2. Furthermore, Saggioro et al. [54] concluded that the samples evaluated did not produce cytotoxic by-products for S. cerevisiae cells.
Cunha et al. [55] evaluated the feasibility of the electrochemical filter with a carbon nanotube to remove E2 ethinylestradiol (EE2) and estrogenic activity. The authors applied an electrochemical system consisting of a titanium cathode and a carbon nanotube anode filter. Sodium sulfate (10 mmol L−1) was used as the supporting electrolyte and a voltage rate (0–2.5 V) was applied during 300 min of electrolysis. Under these conditions, the YES assays allowed verification that the process efficiently reduced the estrogenic activity of the target hormones (2.467 ± 0.0012 ng·L−1 E2-EQ). However, even under ideal conditions (2.5 V), mineralization was not achieved.

3.3. Toxicity Effect on D. similis

Samples A–E (Section 2.6) were applied to the assay as controls to investigate the effect of the electrogenerated species before and after the anodic oxidation process and to verify whether the effect on acute ecotoxicity on D. similis is a result of hormones, chlorinated species, or a combination of both (Figure 4). From this, the three mentioned conditions showed similar UT values (~3.0 UT 100/CE50%). This result suggests that the presence of NaCl in the mixture of LNG and GES does not interfere with the toxic effect attributed to the hormones before electrolysis.
After the oxidation process using the DSA-Cl2 system, the saline solution (sample D) showed an increase in UT of about three times, when compared to that observed for the untreated solution (sample C). Finally, sample E indicated an even greater increase of about four times that observed for sample B. In the latter case, the relevance of the process can be verified, since after the treatment there was the formation of chlorinated species, which may have contributed to the intensification of the acute effect to the organisms evaluated (Figure 4).
Lin et al. [34] applied bioassays with D. magna to monitor the degradation of orange II by the electro-Fenton (EF) process. The authors observed that the sample obtained by the electrochemical process after 5 min resulted in 100% immobilization of the test organisms. However, after 60 min of the treatment process, no immobilization was observed, thus confirming the feasibility of EF in removing acute toxicity.

4. Conclusions

The anodic oxidation process proved to be effective in decomposing progestins in a synthetic mixture and real pharmaceutical wastewater (RPW), with removals ≥70% in less than 20 min of electrolysis. The statistical analysis of the data allowed for identifying the main effects of first-order, second-order, and interaction of the independent variables ([LNG]0, [GES]0, j, and [NaCl]0). The statistical model obtained allowed for determining the relative importance of these variables on the response specific energy consumption (LNG concentration > NaCl concentration > current density > GES concentration). The model presented a correlation coefficient of 89.5%, indicating a good fit to the data with significant F and p-values, allowing estimation of the optimal point which, for practical purposes, was assumed as the central point ([LNG]0 = 1.05 ± 0.03 mg L−1 and [GES]0 = 1.10 ± 0.17 mg L−1; current density, 32.5 mA cm−2; and [NaCl]0 = 0.07 mol L−1). These conditions correspond to an SEC value of 2.2 ± 0.2 kWh g−1 and removal efficiencies of 82.0 ± 0.7% for LNG and 75.0 ± 0.5% for GES. The model accuracy was also verified under central point conditions for RPW, resulting in an SEC of 2.32 ± 0.03 kWh g−1 and removal efficiencies of 83.1 ± 0.9% for LNG and 75.1 ± 2.8 % for GES. Regarding estrogenic activity, the treatment under central point conditions allowed the reduction of estrogenicity by 74.6% after 20 min of anodic oxidation. Finally, the evolution of acute toxicity, evaluated with D. similis, proved to be strongly dependent on the presence of chlorinated species; however, prolonged anodic oxidation treatment can reduce residual ecotoxicity, as reported in the literature.

Supplementary Materials

The following supporting information can be downloaded at:, Figure S1—(I) Electrochemical apparatus used for the degradation of progestogens via anodic oxidation: (1) reservoir; (2) peristaltic pump; (3) electrochemical cell; (4) power supply; (5) multimeter. (II) Schematic representation of the electrochemical apparatus. (III) Disassembled electrochemical cell; Figure S2—Pareto chart of standardized effects; Figure S3—Dose-response curve for the positive control E2 (acetonitrile: water of 55:45 v/v. LOD = 2.3 ng L−1 and LOQ = 7.0 ng L−1); Figure S4—Dose-response curve for experiment 28 before (a) and after (b) the anodic oxidation process ([NaCl]0 = 0.07 mol L−1, j = 32.5 mA cm−2, [LNG]0 = [GES]0 = 1.0 mg L−1); Table S1——Physical and chemical properties of progestins under study; Table S2—Validation parameters of the LNG and GES calibration curves obtained by UFLC analysis. LOD: limit of detection; LOQ: limit of quantification; RSD: relative standard deviation; CI: confidence interval.

Author Contributions

Conceptualization, J.M.S.d.J.; methodology, J.M.S.d.J., A.d.S.A. and F.K.T.; software, J.M.S.d.J. and M.E.T.; validation, J.M.S.d.J.; formal analysis, J.M.S.d.J. and M.E.T.; investigation, J.M.S.d.J.; resources, S.I.B., D.M.B. and A.C.S.C.T.; data curation, J.M.S.d.J., A.d.S.A., F.K.T., and A.C.S.C.T.; writing—original draft preparation, J.M.S.d.J.; writing— review and editing, J.M.S.d.J. and A.C.S.C.T.; visualization, J.M.S.d.J.; supervision, M.E.T. and A.C.S.C.T.; project administration, A.C.S.C.T.; funding acquisition, A.C.S.C.T. All authors have read and agreed to the published version of the manuscript.


This research was funded in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior, Brasil (CAPES), Finance Code 001. The authors thank the support of the São Paulo Research Foundation (FAPESP) [grant number 2019/24158–9] and the National Council for Scientific and Technological Development—Brazil (CNPq) [grant number 311230/2020–2].

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Not applicable.

Conflicts of Interest

The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.


  1. Contardo-Jara, V.; Lorenz, C.; Pflugmacher, S.; Nützmann, G.; Kloas, W.; Wiegand, C. Molecular Effects and Bioaccumulation of Levonorgestrel in the Non-Target Organism Dreissena Polymorpha. Environ. Pollut. 2011, 159, 38–44. [Google Scholar] [CrossRef] [PubMed]
  2. Runnalls, T.J.; Beresford, N.; Kugathas, S.; Margiotta-Casaluci, L.; Scholze, M.; Scott, A.P.; Sumpter, J.P. From Single Chemicals to Mixtures-Reproductive Effects of Levonorgestrel and Ethinylestradiol on the Fathead Minnow. Aquat. Toxicol. 2015, 169, 152–167. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  3. Shi, W.J.; Ma, D.D.; Fang, G.Z.; Zhang, J.G.; Huang, G.Y.; Xie, L.; Chen, H.X.; Hou, L.P.; Ying, G.G. Levonorgestrel and Dydrogesterone Affect Sex Determination via Different Pathways in Zebrafish. Aquat. Toxicol. 2021, 240, 105972. [Google Scholar] [CrossRef] [PubMed]
  4. Contraceptive Use by Method 2019. Contracept; United Nations: New York, NY, USA, 2019. [CrossRef]
  5. Bahamondes, L.; Ali, M.; Monteiro, I.; Fernandes, A. Contraceptive Sales in the Setting of the Zika Virus Epidemic. Hum. Reprod. 2017, 32, 88–93. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  6. King, O.C.; van de Merwe, J.P.; McDonald, J.A.; Leusch, F.D.L. Concentrations of Levonorgestrel and Ethinylestradiol in Wastewater Effluents: Is the Progestin Also Cause for Concern? Environ. Toxicol. Chem. 2016, 35, 1378–1385. [Google Scholar] [CrossRef]
  7. Orlando, E.F.; Ellestad, L.E. Sources, Concentrations, and Exposure Effects of Environmental Gestagens on Fish and Other Aquatic Wildlife, with an Emphasis on Reproduction. Gen. Comp. Endocrinol. 2014, 203, 241–249. [Google Scholar] [CrossRef]
  8. Fent, K. Progestins as Endocrine Disrupters in Aquatic Ecosystems: Concentrations, Effects and Risk Assessment. Environ. Int. 2015, 84, 115–130. [Google Scholar] [CrossRef]
  9. Besse, J.P.; Garric, J. Progestagens for Human Use, Exposure and Hazard Assessment for the Aquatic Environment. Environ. Pollut. 2009, 157, 3485–3494. [Google Scholar] [CrossRef] [Green Version]
  10. Rocha, M.; Rocha, E. Synthetic Progestins in Waste and Surface Waters: Concentrations, Impacts and Ecological Risk. Toxics 2022, 10, 163. [Google Scholar] [CrossRef]
  11. Pal, A.; Gin, K.Y.H.; Lin, A.Y.C.; Reinhard, M. Impacts of Emerging Organic Contaminants on Freshwater Resources: Review of Recent Occurrences, Sources, Fate and Effects. Sci. Total Environ. 2010, 408, 6062–6069. [Google Scholar] [CrossRef]
  12. Golovko, O.; Šauer, P.; Fedorova, G.; Kroupová, H.K.; Grabic, R. Determination of Progestogens in Surface and Waste Water Using SPE Extraction and LC-APCI/APPI-HRPS. Sci. Total Environ. 2018, 621, 1066–1073. [Google Scholar] [CrossRef] [PubMed]
  13. Yu, Q.; Geng, J.; Zong, X.; Zhang, Y.; Xu, K.; Hu, H.; Deng, Y.; Zhao, F.; Ren, H. Occurrence and Removal of Progestagens in Municipal Wastewater Treatment Plants from Different Regions in China. Sci. Total Environ. 2019, 668, 1191–1199. [Google Scholar] [CrossRef] [PubMed]
  14. Nasuhoglu, D.; Berk, D.; Yargeau, V. Photocatalytic Removal of 17-Alfaethinylestradiol (EE2) and Levonorgestrel (LNG) from Contraceptive Pill Manufacturing Plant Wastewater under UVC Radiation. Chem. Eng. J. 2012, 185–186, 52–60. [Google Scholar] [CrossRef]
  15. de Jesus, J.M.S.; Tominaga, F.K.; dos Santos Argolo, A.; Nascimento, A.C.G.; Borrely, S.I.; Vieira, D.P.; Bila, D.M.; Costa Teixeira, A.C.S. Radiolytic Degradation of Levonorgestrel and Gestodene: Performance and Bioassays. Process Saf. Environ. Prot. 2022, 162, 520–530. [Google Scholar] [CrossRef]
  16. Martín de Vidales, M.J.; Sáez, C.; Cañizares, P.; Rodrigo, M.A. Electrolysis of Progesterone with Conductive-Diamond Electrodes. J. Chem. Technol. Biotechnol. 2012, 87, 1173–1178. [Google Scholar] [CrossRef]
  17. Brocenschi, R.F.; Rocha-Filho, R.C.; Bocchi, N.; Biaggio, S.R. Electrochemical Degradation of Estrone Using a Boron-Doped Diamond Anode in a Filter-Press Reactor. Electrochim. Acta 2016, 197, 186–193. [Google Scholar] [CrossRef]
  18. Frontistis, Z.; Brebou, C.; Venieri, D.; Mantzavinos, D.; Katsaounis, A. BDD Anodic Oxidation as Tertiary Wastewater Treatment for the Removal of Emerging Micro-Pollutants, Pathogens and Organic Matter. J. Chem. Technol. Biotechnol. 2011, 86, 1233–1236. [Google Scholar] [CrossRef]
  19. Nájera-Aguilar, H.A.; Gutiérrez-Hernández, R.F.; de Los Santos, R.G.; García-Lara, C.; Méndez-Novelo, R.; Rojas-Valencia, M.N. Degradation of Gestodene (GES) and 17α-Ethinylestradiol (EE2) by Electrochemical Oxidation. J. Water Health 2016, 14, 980–988. [Google Scholar] [CrossRef] [Green Version]
  20. Duarte, J.L.S.; Meili, L.; Gomes, L.M.; Melo, J.M.O.; Ferro, A.B.; Tavares, M.G.; Tonholo, J.; Zanta, C.L.P.S. Electrochemical Degradation of 17-A-Methyltestosterone over DSA® Electrodes. Chem. Eng. Process.-Process Intensif. 2019, 142, 107548. [Google Scholar] [CrossRef]
  21. Vieira, K.M.; Nascentes, C.C.; Motheo, A.J.; Augusti, R. Electrochemical Oxidation of Ethinylestradiol on a Commercial Ti/Ru 0.3 Ti 0.7 O 2 DSA Electrode. ISRN Environ. Chem. 2013, 2013, 354848. [Google Scholar] [CrossRef] [Green Version]
  22. Murugananthan, M.; Yoshihara, S.; Rakuma, T.; Uehara, N.; Shirakashi, T. Electrochemical Degradation of 17β-Estradiol (E2) at Boron-Doped Diamond (Si/BDD) Thin Film Electrode. Electrochim. Acta 2007, 52, 3242–3249. [Google Scholar] [CrossRef]
  23. Al-Qaim, F.F.; Mussa, Z.H.; Yuzir, A.; Latip, J.; Othman, M.R. The Fate of Prazosin and Levonorgestrel after Electrochemical Degradation Process: Monitoring by-Products Using LC-TOF/MS. J. Environamental Sci. 2018, 21, 134–146. [Google Scholar] [CrossRef] [PubMed]
  24. Cięciwa, A.; Fόti, G.; Comninellis, C. Basic Principles of the Electrochemical Mineralization of Organic Pollutants for Wastewater Treatment; Springer: Berlin/Heidelberg, Germany, 2010; ISBN 9780387369228. [Google Scholar]
  25. Garcia-Segura, S.; Ocon, J.D.; Chong, M.N. Electrochemical Oxidation Remediation of Real Wastewater Effluents—A Review. Process. Saf. Environ. Prot. 2018, 113, 48–67. [Google Scholar] [CrossRef]
  26. Panizza, M.; Cerisola, G. Direct and Mediated Anodic Oxidation of Organic Pollutants. Chem. Rev. 2009, 109, 6541–6569. [Google Scholar] [CrossRef] [PubMed]
  27. Torres, N.H.; de Oliveira Santiago Santos, G.; Ferreira, L.F.R.; Américo-Pinheiro, J.H.P.; Eguiluz, K.I.B.; Salazar-Banda, G.R. Environmental Aspects of Hormones Estriol, 17β-Estradiol and 17α-Ethinylestradiol: Electrochemical Processes as next-Generation Technologies for Their Removal in Water Matrices. Chemosphere 2021, 267, 128888. [Google Scholar] [CrossRef] [PubMed]
  28. Jiang, Y.; Zhao, H.; Liang, J.; Yue, L.; Li, T.; Luo, Y.; Liu, Q.; Lu, S.; Asiri, A.M.; Gong, Z.; et al. Anodic Oxidation for the Degradation of Organic Pollutants: Anode Materials, Operating Conditions and Mechanisms. A Mini Review. Electrochem. commun. 2021, 123, 106912. [Google Scholar] [CrossRef]
  29. Khan, S.; Sayed, M.; Sohail, M.; Shah, L.A.; Raja, M.A. Advanced Oxidation and Reduction Processes; Elsevier Inc.: Amsterdam, The Netherlands, 2019; ISBN 9780128147900. [Google Scholar]
  30. Lozano, I.; Pérez-Guzmán, C.J.; Mora, A.; Mahlknecht, J.; Aguilar, C.L.; Cervantes-Avilés, P. Pharmaceuticals and Personal Care Products in Water Streams: Occurrence, Detection, and Removal by Electrochemical Advanced Oxidation Processes. Sci. Total Environ. 2022, 827, 154348. [Google Scholar] [CrossRef]
  31. Nabgan, W.; Saeed, M.; Jalil, A.A.; Nabgan, B.; Gambo, Y.; Ali, M.W.; Ikram, M.; Fauzi, A.A.; Owgi, A.H.K.; Hussain, I.; et al. A State of the Art Review on Electrochemical Technique for the Remediation of Pharmaceuticals Containing Wastewater. Environ. Res. 2022, 210, 112975. [Google Scholar] [CrossRef]
  32. Santos, M.C.; Antonin, V.S.; Souza, F.M.; Aveiro, L.R.; Pinheiro, V.S.; Gentil, T.C.; Lima, T.S.; Moura, J.P.C.; Silva, C.R.; Lucchetti, L.E.B.; et al. Decontamination of Wastewater Containing Contaminants of Emerging Concern by Electrooxidation and Fenton-Based Processes–A Review on the Relevance of Materials and Methods. Chemosphere 2022, 307, 603. [Google Scholar] [CrossRef]
  33. Pillai, I.M.S.; Gupta, A.K. Batch and Continuous Flow Anodic Oxidation of 2,4-Dinitrophenol: Modeling, Degradation Pathway and Toxicity. J. Electroanal. Chem. 2015, 756, 108–117. [Google Scholar] [CrossRef]
  34. Lin, H.; Zhang, H.; Wang, X.; Wang, L.; Wu, J. Electro-Fenton Removal of Orange II in a Divided Cell: Reaction Mechanism, Degradation Pathway and Toxicity Evolution. Sep. Purif. Technol. 2014, 122, 533–540. [Google Scholar] [CrossRef]
  35. Wu, J.; Zhang, H.; Oturan, N.; Wang, Y.; Chen, L.; Oturan, M.A. Application of Response Surface Methodology to the Removal of the Antibiotic Tetracycline by Electrochemical Process Using Carbon-Felt Cathode and DSA (Ti/RuO2-IrO2) Anode. Chemosphere 2012, 87, 614–620. [Google Scholar] [CrossRef] [PubMed]
  36. Bistan, M.; Podgorelec, M.; Logar, R.M.; Tišler, T. Yeast Estrogen Screen Assay as a Tool for Detecting Estrogenic Activity in Water Bodies. Food Technol. Biotechnol. 2012, 50, 427–433. [Google Scholar]
  37. Arnold, S.F.; Robinson, M.K.; Notides, A.C.; Guillette, L.J.; McLachan, J.A. A Yeast Estrogen Screen for Examining the Relative Exposure of Cells to Natural and Xenoestrogens. Environ. Health Perspect. 1996, 104, 544–548. [Google Scholar] [CrossRef] [PubMed]
  38. Argolo, A.D.S.; Gomes, G.; Bila, D.M. Insights into Total Estrogenic Activity in a Sewage-Impacted Urban Stream Assessed via ER Transcriptional Activation Assay: Distribution between Particulate and Dissolved Phases. Ecotoxicol. Environ. Saf. 2021, 208, 111574. [Google Scholar] [CrossRef]
  39. Tominaga, F.K.; Silva, T.T.; Boiani, N.F.; de Jesus, J.M.S.; Teixeira, A.C.S.C.; Borrely, S.I. Is Ionizing Radiation Effective in Removing Pharmaceuticals from Wastewater? Environ. Sci. Pollut. Res. 2021, 28, 23975–23983. [Google Scholar] [CrossRef]
  40. Fernandes, A.; Pastorinho, M.R.; Sousa, A.C.; Silva, W.; Silva, R.; Nunes, M.J.; Rodrigues, A.S.; Pacheco, M.J.; Ciríaco, L.; Lopes, A. Ecotoxicological Evaluation of Electrochemical Oxidation for the Treatment of Sanitary Landfill Leachates. Environ. Sci. Pollut. Res. 2019, 26, 24–33. [Google Scholar] [CrossRef] [Green Version]
  41. Abbas, M.; Adil, M.; Ehtisham-ul-Haque, S.; Munir, B.; Yameen, M.; Ghaffar, A.; Shar, G.A.; Asif Tahir, M.; Iqbal, M. Vibrio Fischeri Bioluminescence Inhibition Assay for Ecotoxicity Assessment: A Review. Sci. Total Environ. 2018, 626, 1295–1309. [Google Scholar] [CrossRef]
  42. Christen, V.; Hickmann, S.; Rechenberg, B.; Fent, K. Highly Active Human Pharmaceuticals in Aquatic Systems: A Concept for Their Identification Based on Their Mode of Action. Aquat. Toxicol. 2010, 96, 167–181. [Google Scholar] [CrossRef]
  43. Persoone, G.; Marsalek, B.; Blinova, I.; Törökne, A.; Zarina, D.; Manusadzianas, L.; Nalecz-Jawecki, G.; Tofan, L.; Stepanova, N.; Tothova, L.; et al. A Practical and User-Friendly Toxicity Classification System with Microbiotests for Natural Waters and Wastewaters. Environ. Toxicol. 2003, 18, 395–402. [Google Scholar] [CrossRef]
  44. Le, G.T.; Ta, N.T.; Pham, T.Q.; Dao, Y.H. Response Surface Analysis of Fenobucarb Removal by Electrochemically Generated Chlorine. Water 2019, 11, 899. [Google Scholar] [CrossRef] [Green Version]
  45. Leili, M.; Shirmohammadi Khorram, N.; Godini, K.; Azarian, G.; Moussavi, R.; Peykhoshian, A. Application of Central Composite Design (CCD) for Optimization of Cephalexin Antibiotic Removal Using Electro-Oxidation Process. J. Mol. Liq. 2020, 313, 113556. [Google Scholar] [CrossRef]
  46. Li, M.; Xu, B.; Liungai, Z.; Hu, H.Y.; Chen, C.; Qiao, J.; Lu, Y. The Removal of Estrogenic Activity with UV/Chlorine Technology and Identification of Novel Estrogenic Disinfection by-Products. J. Hazard. Mater. 2016, 307, 119–126. [Google Scholar] [CrossRef] [PubMed]
  47. Aquino, J.M.; Rocha-Filho, R.C.; Bocchi, N.; Biaggio, S.R. Electrochemical Degradation of the Reactive Red 141 Dye on a β-PbO2 Anode Assessed by the Response Surface Methodology. J. Braz. Chem. Soc. 2010, 21, 324–330. [Google Scholar] [CrossRef] [Green Version]
  48. Routledge, E.J.; Sumpter, J.P. Estrogenic Activity of Surfactants and Some of Their Degradation Products Assessed Using a Recombinant Yeast Screen. Environ. Toxicol. Chem. 1996, 15, 241–248. [Google Scholar] [CrossRef]
  49. ABNT-NBR 12713; Ecotoxicologia aquática—toxicidade aguda—método de ensaio com Daphnia spp (crustacea, Cladocera). Associação Brasileira de Normas Técnicas: Rio de Janeiro, Brasil, 2012.
  50. Periyasamy, S.; Muthuchamy, M. Electrochemical Oxidation of Paracetamol in Water by Graphite Anode: Effect of PH, Electrolyte Concentration and Current Density. J. Environ. Chem. Eng. 2018, 6, 7358–7367. [Google Scholar] [CrossRef]
  51. Khan, A.H.; Khan, N.A.; Ahmed, S.; Dhingra, A.; Singh, C.P.; Khan, S.U.; Mohammadi, A.A.; Changani, F.; Yousefi, M.; Alam, S.; et al. Application of Advanced Oxidation Processes Followed by Different Treatment Technologies for Hospital Wastewater Treatment. J. Clean. Prod. 2020, 269, 122411. [Google Scholar] [CrossRef]
  52. Patel, S.; Majumder, S.K.; Mondal, S.; Das, P.; Ghosh, P. Treatment of a Pharmaceutical Industrial Effluent by a Hybrid Process of Advanced Oxidation and Adsorption. ACS Omega 2020, 5, 32305–32317. [Google Scholar] [CrossRef]
  53. Amor, C.; Fernandes, J.R.; Lucas, M.S.; Peres, J.A. Hydroxyl and Sulfate Radical Advanced Oxidation Processes: Application to an Agro-Industrial Wastewater. Environ. Technol. Innov. 2021, 21, 101183. [Google Scholar] [CrossRef]
  54. Saggioro, E.M.; Chaves, F.P.; Felix, L.C.; Gomes, G.; Bila, D.M. Endocrine Disruptor Degradation by UV/Chlorine and the Impact of Their Removal on Estrogenic Activity and Toxicity. Int. J. Photoenergy 2019, 2019, 7408763. [Google Scholar] [CrossRef]
  55. Cunha, G.D.S.; de Souza-Chaves, B.M.; Bila, D.M.; Bassin, J.P.; Vecitis, C.D.; Dezotti, M. Insights into Estrogenic Activity Removal Using Carbon Nanotube Electrochemical Filter. Sci. Total Environ. 2019, 678, 448–456. [Google Scholar] [CrossRef] [PubMed]
Figure 1. Predicted vs. observed values (a) and (b) raw residuals vs. predicted values for the specific energy consumption (SEC kWh g−1).
Figure 1. Predicted vs. observed values (a) and (b) raw residuals vs. predicted values for the specific energy consumption (SEC kWh g−1).
Water 15 00061 g001
Figure 2. Three-dimensional surface plot of the specific energy consumption (SEC) as a function of the independent variables X1 ([GES]0), X2 [LNG]0, X3 (j) (), and X4([NaCl]0) ). Each plot highlights the effects of two variables on the SEC with the other two variables fixed at their central point conditions: (a) X3 (j) and X4 ([NaCl]0); (b) X2 [LNG]0 and X4 ([NaCl]0);c) X1 ([GES]0) and X2 [LNG]0.
Figure 2. Three-dimensional surface plot of the specific energy consumption (SEC) as a function of the independent variables X1 ([GES]0), X2 [LNG]0, X3 (j) (), and X4([NaCl]0) ). Each plot highlights the effects of two variables on the SEC with the other two variables fixed at their central point conditions: (a) X3 (j) and X4 ([NaCl]0); (b) X2 [LNG]0 and X4 ([NaCl]0);c) X1 ([GES]0) and X2 [LNG]0.
Water 15 00061 g002
Figure 3. Results of estrogenic activity (E2-EQ) of LNG-GES solutions before and after 20 min of anodic oxidation mediated by chlorine active species ([NaCl]0 = 0.07 mol L−1, j = 32.5 mA cm−2, [LNG]0 = [GES]0 = 1.0 mg L−1). Measurements performed in duplicate.
Figure 3. Results of estrogenic activity (E2-EQ) of LNG-GES solutions before and after 20 min of anodic oxidation mediated by chlorine active species ([NaCl]0 = 0.07 mol L−1, j = 32.5 mA cm−2, [LNG]0 = [GES]0 = 1.0 mg L−1). Measurements performed in duplicate.
Water 15 00061 g003
Figure 4. Results of acute toxicity tests (in toxicity units, TU= 100/EC50%) to D. similis before and after anodic oxidation for samples (A) mixture of progestogens (LNG + GES) without electrochemical processing and in the absence of NaCl; (B) mixture of progestogens (LNG + GES) without electrochemical processing in the presence of NaCl; (C) NaCl solution without electrochemical processing; (D) NaCl solution after electrochemical processing (20 min); and (E) mixture of progestogens (LNG + GES) after electrochemical processing (20 min) in the presence of NaCl ([LNG]0 = [GES]0 = 1.0 mg L−1; j = 32.5 mA cm−2; [NaCl]0 = 0.07 mol L−1). Values correspond to an average of eight replicates.
Figure 4. Results of acute toxicity tests (in toxicity units, TU= 100/EC50%) to D. similis before and after anodic oxidation for samples (A) mixture of progestogens (LNG + GES) without electrochemical processing and in the absence of NaCl; (B) mixture of progestogens (LNG + GES) without electrochemical processing in the presence of NaCl; (C) NaCl solution without electrochemical processing; (D) NaCl solution after electrochemical processing (20 min); and (E) mixture of progestogens (LNG + GES) after electrochemical processing (20 min) in the presence of NaCl ([LNG]0 = [GES]0 = 1.0 mg L−1; j = 32.5 mA cm−2; [NaCl]0 = 0.07 mol L−1). Values correspond to an average of eight replicates.
Water 15 00061 g004
Table 1. Concentration of progestins in water matrices (ng L−1).
Table 1. Concentration of progestins in water matrices (ng L−1).
Levonorgestrel (LNG)Gestodene
Surface water5.3–7.05.03[8,9,11]
Wastewater treatment plants (WWTPs)<0.2–16.1 (influent)
<0.5–4.0 (effluent)
0.44 (influent)
0.19 (effluent)
Pharmaceutical wastewater4–5 × 106 (total concentration)
0.8–0.16 × 105 (aqueous concentration)
0.66 × 105[14,15]
Table 2. Anodic oxidation (AO) studies applied for the removal of hormones.
Table 2. Anodic oxidation (AO) studies applied for the removal of hormones.
HormoneConcentrationMatricesElectrodesSupport ElectrolyteCurrent DensityReference
Progesterone0.1 to 100 mg L−1Milli-Q waterBDD (anode)/SS (cathode)0.035 mol L−1 Na2SO4 or NaCl15–100 mA[16]
Estrone230, 410, and, 570 µg L−1Milli-Q waterBDD (anode)/SS (cathode)0.36 mol L−1 NaCl5, 10, and, 25 mA cm−2[17]
Estradiol (E2) and Ethyniletradiol (EE2)1 mg L−1Milli-Q water and WWTPBDD (anode)/zirconium (cathode)0.1 mol L−1 Na2SO4 or NaCl0.9–2.6 mA cm−2[18]
Gestodene (GES) and Ethyniletradiol (EE2)625 µg L−1 of GES and 250 µg L−1 of EE2
Milli-Q waterBDD as anode and cathode0.02, 0.05 and 0.1 mol L−1 Na2SO4
12, 32 and, 48 mA cm−2[19]
17-α-Methyltestosterone5.0 mg L−1Milli-Q waterDSA as anode and cathode0.5 mol L−1 Na2SO4 or NaCl
15, 30, and 45 mA cm−2
Ethyniletradiol (EE2)100 mg L−1Water/methanol 7: 3 v/vDSA (anode)/ SS (cathode)0.004 mol L−1 NaCl40 mA cm−2
Estradiol (E2)250–750 µg dm−3WaterBDD-WE
Hg/Hg2Cl2·KCl (sat.) (SCE)-RE
0.1 mol L−1 Na2SO4, NaNO3, and NaCl25 mA cm−2[22]
Levonorgestrel (LNG)2.5 mg L−1Milli-Q water and pharmaceutical wastewaterPVC-graphite (anode)/platinium (cathode)0.05, 0.1, and 0.2 g NaCl3, 6, and 8 V
SS = stainless steel, WE = working electrode, RE = reference electrode, CE = counter electrode.
Table 3. Independent variables and their coded and uncoded values.
Table 3. Independent variables and their coded and uncoded values.
Independent VariablesSymbolCoded Levels
Uncoded Values
[GES]0 (mg L−1)X10.
[LNG]0 (mg L−1)X20.
j (mA cm−2)X37.520.032.545.057.5
[NaCl]0 (mol L−1)X40.
Table 4. CCD matrix and variable levels used to study the influence of operating parameters on the anodic oxidation of levonorgestrel (LNG) and gestodene (GES), response variable (SEC) applied in the statistical analysis, and additional observed results.
Table 4. CCD matrix and variable levels used to study the influence of operating parameters on the anodic oxidation of levonorgestrel (LNG) and gestodene (GES), response variable (SEC) applied in the statistical analysis, and additional observed results.
RunCoded LevelsExperimental Variable LevelsResponseRemoval Efficiency (%)kobs (min−1)
(mg L−1)
(mg L−1)
j (mA cm−2)[NaCl]0
(mol L−1)
(kWh g−1)
Table 5. Analysis of variance (ANOVA) for the quadratic model and F-test.
Table 5. Analysis of variance (ANOVA) for the quadratic model and F-test.
SourceSum of SquaresDegree of FreedomMean SquareF-Valuep-Value
Regression146.991410.50188.4930.004 a
First-order effects114.60428.652057.370.0001 a
Second-order effects15.9043.97285.600.0207 a
[GES]0 × [LNG]00.9210.9216.630.006 a
[GES]0 × [NaCl]00.4210.427.6740.032 a
[LNG]0 × j1.7711.7731.820.0013 a
[LNG]0 × [NaCl]08.2218.22147.650.00001 a
Lack of fit16.59101.6629.780.0002 a
Pure error0.3360.05
R2 = 0.8951; a Significance at 95% confidence level.
Table 6. Consumption of chloride anions after 20 min of anodic oxidation for selected experiments.
Table 6. Consumption of chloride anions after 20 min of anodic oxidation for selected experiments.
Runj (mA cm−2)[NaCl]0 (mol L−1)Cl Consumption (%)
Table 7. Values obtained as the optimum point, coded and actual, and central point.
Table 7. Values obtained as the optimum point, coded and actual, and central point.
VariableOptimum Point (coded)Optimum Point (actual)Central Point
[GES]0 (mg L−1)3.944 a2.971.0
[LNG]0 (mg L−1)0.6231.311.0
j (mA cm−2)−1.43214.632.5
[NaCl]0 (mol L−1)2.356 a0.140.07
a value outside the experimental domain.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

de Jesus, J.M.S.; Argolo, A.d.S.; Tominaga, F.K.; Taqueda, M.E.; Bila, D.M.; Borrely, S.I.; Teixeira, A.C.S.C. Experimental Design and Bioassays as Tools to Investigate the Impact of Anodic Oxidation on Progestins Degradation. Water 2023, 15, 61.

AMA Style

de Jesus JMS, Argolo AdS, Tominaga FK, Taqueda ME, Bila DM, Borrely SI, Teixeira ACSC. Experimental Design and Bioassays as Tools to Investigate the Impact of Anodic Oxidation on Progestins Degradation. Water. 2023; 15(1):61.

Chicago/Turabian Style

de Jesus, Juliana Mendonça Silva, Allan dos Santos Argolo, Flávio Kiyoshi Tominaga, Maria Elena Taqueda, Daniele Maia Bila, Sueli Ivone Borrely, and Antonio Carlos Silva Costa Teixeira. 2023. "Experimental Design and Bioassays as Tools to Investigate the Impact of Anodic Oxidation on Progestins Degradation" Water 15, no. 1: 61.

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop