Listen
graphic
View largeDownload slide

In this work, the electrochemical degradation of ciprofloxacin (CIP) was studied in a filter-press-type reactor without division in a batch recirculation manner. For this purpose, two boron-doped diamond (BDD) electrodes (as cathode and anode) were employed. Also, the optimal operating conditions were found by response surface methodology (RSM) following a central composite face-centered design with three factors, namely current intensity (i), initial pH (pH0), and initial concentration ([C]0) with two responses, namely remotion efficiency (η) and operating cost. Optimal operating conditions were i = 3 A, pH0 = 8.49, and [C]0 = 33.26 mg L−1 within an electrolysis time of 5 h, leading to a maximum removal efficiency of 93.49% with a minimum operating cost of $0.013 USD L−1. Also, a TOC analysis shows an 80% of mineralization extent with an energy consumption of 5.11 kWh g−1 TOC. Furthermore, the CIP degradation progress was followed by mass spectrometry (LC/MS) and a degradation pathway is proposed.

Listen

  • Ciprofloxacin has been removed efficiently in a flow-by reactor equipped with two BDD electrodes.

  • Optimal operating conditions were pH0 = 8.49, i = 3 A, and [C]0 = 33.26 mg L−1.

  • Mineralization efficiency and extent of electrochemical combustion at optimal operating conditions were 80% and 0.85.

  • Three pathway reactions for the electrochemical degradation of CIP were described.

BDD electrodes, central composite face-centered design, ciprofloxacin, degradation pathway, DoE-driven optimization
Listen

In recent years, a special concern has been raised with the presence of emerging pollutant compounds (EC) in water bodies (e.g., rivers, lakes, and oceans) (Ojo et al. 2022) and soils (e.g., farming, cattle raising, and highway) because these can cause human health and environmental issues. EC, also known as micropollutants, can be either synthetic or from a natural origin (Khan et al. 2020; Arman et al. 2021; Firdaus et al. 2021). Common EC are personal cleaning products and drugs (e.g., hormones, antibiotics, analgesics, antidepressants, pesticides, among others) (Goswami et al. 2022). Although drugs are essential to human and animal welfare, they are harmful to the environment since they are introduced into water bodies (e.g., wastewater) (Rivera-Utrilla et al. 2013) by excretion and urine through sewage, slurry, and rainfall. Among the EC are fluoroquinolones such as ciprofloxacin (CIP), which is the most used antibiotic for many diseases (e.g., respiratory and bacterial) (Ashfaq et al. 2016). Also, CIP was employed during the COVID-19 pandemic as a complementary treatment (Cappelli et al. 2022). However, intensive use is dangerous to the environment since CIP is not completely metabolized by the human body and animals causing accumulation in water and promoting resistance to antibiotics resulting in risks to human health (Ahmadzadeh et al. 2017; Kim et al. 2020). In this context, the presence of EC (such as CIP) in wastewater has motivated the development of technologies to eliminate these since conventional wastewater treatments do not exhibit a high degradation efficiency. The advanced oxidation processes (AOPs, e.g., anodic oxidation (AO), electro-Fenton, photo-electro-Fenton, solar-photo-electro-Fenton, among others (Moreira et al. 2017)), emerge as an efficient technique to remove EC. AOPs are based on the generation of hydroxyl radicals (⁠⁠). In this sense, the AO becomes a green, efficient energetic, versatile, profitable, easily automatized, and promising technology (Martín de Vidales et al. 2020; Firdaus et al. 2021) for wastewater treatment.

AO has the advantage of producing hydroxyl radicals without the addition of chemicals (e.g., hydrogen peroxide and FeSO4, like in the Fenton process) since the water is oxidized in the anode surface according to the following equation (Chaplin 2014),
(1)

is capable of oxidizing and mineralizing organic compounds such as EC until carbon dioxide (CO2), ions, and water. Also, the amount of is dependent on the anode material because the overpotential of the oxygen evolution reaction is different for each anode material. In this sense, the boron-doped diamond (BDD) electrode exhibits a high overpotential of the oxygen evolution reaction. Therefore, the BDD anode is widely employed for the generation of (Auguste & Ouattara 2021).

Table 1 presents a summary of the literature related to the electrochemical degradation of CIP by using different anode materials, different reactor configurations, and different reaction environments. Although the literature review reveals high removal efficiencies of CIP, the treatment volume employed in many cases is very small, and the operating cost for all studies is not reported. In addition, it is observed in Table 1 that the optimization of the electrochemical degradation of CIP has not been yet intensively performed.

Table 1

Degradation of CIP at different environment reaction conditions

Operating conditions
Main results
OptimizedNot optimizedElectrodesV (L)ED (%)TOC (%)DQO (%)Ref.
Batch reactor, pH0 = 9, j = 17 mA cm−2, [Na2SO4] = 1.5 mg L−1, and t = 5 h  Ti/SnO2-Sb2O5, Ti/RuO2-Ti 0.2   89.5 Firdaus et al. (2021)  
 Flow reactor, Q = 2.5 L min−1, pH0 = 10, j = 30 mA cm−2, [C]0 = 50 mg L−1, [Na2SO4] = 0.1 mol L−1, and t = 10 h BDD-stainless steel 304 0.5 100 100  Wachter et al. (2019a)  
 Flow reactor, Q = 6.5 L min−1, pH0 = 10, j = 30 mA cm−2, [C]0 = 50 mg L−1, [Na2SO4] = 0.1 mol L−1, and t = 5 h Ni-TiPt/β-PBO2 – Ni BDD-Stainless steel 304 0.5 100 75  Wachter et al. (2019b)  
 Flow reactor, Q = 7 L min−1, j = 10 mA cm−2 [C]0 = 100 mg L−1, [NaCl] = 0.1 mol L−1, and t = 8 h BDD-Stainless steel 304 1 100 80  Carneiro et al. (2020)  
 Flow reactor, Q = 0.4 L min−1, j = 30 mA cm−2, pH = 7, [C]0 = 10 mg L−1, and t = 0.33 h BDD-Ti 0.3 100   Li et al. (2019)  
 Batch reactor, j = 50 mA cm−2, pH0 = 2.5, [C]0 = 30 mg L−1, [K2SO4] = 0.1 mol L−1, and t = 1.5 h Ti–Pt-EDG 0.25 99.3 25.3  Lima et al. (2020)  
 Batch reactor, j = 30 mA cm−2, pH0 = 5.4, [C]0 = 50 mg L−1, [Na2SO4] = 0.05 mol L−1, and t = 2 h SnO2-Sb/Ti-Ti 0.25 99.5 70  Wang et al. (2016)  
 Batch reactor, j = 20 mA cm−2, pH0 = 3, [C]0 = 30 mg L−1, [Na2SO4] = 25 g L−1, and t = 1.5 h Doped Sb SnO2-Stainless steel 0.05 100 93  Mu et al. (2019)  
 Batch reactor, j = 40 mA cm−2, pH0 = 7, [C]0 = 15 mg L−1, Chloride medium, and t = 0.33 h BDD (NCD)-Stainless steel 0.1 100 37.4  dos Santos et al. (2022)  
 Real wastewater, j = 40 mA cm−2, pH0 = 7, [C]0 = 15 mg L−1, and t = 1 h BDD (NCD)-Stainless steel 0.1 100 28.7 
 Synthetic urine, j = 40 mA cm−2, pH0 = 7, [C]0 = 15 mg L−1, and t = 1 h BDD (NCD)-Stainless steel 0.1 90.4 32.2 
 Batch reactor, pure water, j = 45 mA cm−2, pH0 = 7, [C]0 = 10 mg L−1, [Na2SO4] = 0.05 M, and t = 5 h BDD-Stainless steel 2.0 99.9   Montenegro-Ayo et al. (2023)  
 Batch reactor, tap water, [Na2SO4] = 0.05 M, [C]0 = 10 mg L−1, j = 45 mA cm−2, pH0 = 6.5, and, t = 5 h BDD-Stainless steal 2.0 95.8   
 Batch reactor, synthetic urine, [C]0 = 10 mg L−1, [Na2SO4] = 0.05 M, j = 45 mA cm−2, pH0 = 7, and t = 5 h BDD-Stainless steel 2.0 77.2   
 Batch reactor, synthetic urine, [C]0 = 30 mg L−1, [Na2SO4] = 0.05 M, j = 45 mA cm−2, pH0 = 7, and t = 5 h BDD-Stainless steel 2.0   94 
Batch reactor, j = 3.5 mA cm−2, pH0 = 3, [C]0 = 10 mg L−1, [NaCl] = 10 mg L−1, and t = 1.5 h  Ti/nanoSnO2MWCN-Stainless steel 0.25 89.61   Esmaelian et al. (2019)  
Operating conditions
Main results
OptimizedNot optimizedElectrodesV (L)ED (%)TOC (%)DQO (%)Ref.
Batch reactor, pH0 = 9, j = 17 mA cm−2, [Na2SO4] = 1.5 mg L−1, and t = 5 h  Ti/SnO2-Sb2O5, Ti/RuO2-Ti 0.2   89.5 Firdaus et al. (2021)  
 Flow reactor, Q = 2.5 L min−1, pH0 = 10, j = 30 mA cm−2, [C]0 = 50 mg L−1, [Na2SO4] = 0.1 mol L−1, and t = 10 h BDD-stainless steel 304 0.5 100 100  Wachter et al. (2019a)  
 Flow reactor, Q = 6.5 L min−1, pH0 = 10, j = 30 mA cm−2, [C]0 = 50 mg L−1, [Na2SO4] = 0.1 mol L−1, and t = 5 h Ni-TiPt/β-PBO2 – Ni BDD-Stainless steel 304 0.5 100 75  Wachter et al. (2019b)  
 Flow reactor, Q = 7 L min−1, j = 10 mA cm−2 [C]0 = 100 mg L−1, [NaCl] = 0.1 mol L−1, and t = 8 h BDD-Stainless steel 304 1 100 80  Carneiro et al. (2020)  
 Flow reactor, Q = 0.4 L min−1, j = 30 mA cm−2, pH = 7, [C]0 = 10 mg L−1, and t = 0.33 h BDD-Ti 0.3 100   Li et al. (2019)  
 Batch reactor, j = 50 mA cm−2, pH0 = 2.5, [C]0 = 30 mg L−1, [K2SO4] = 0.1 mol L−1, and t = 1.5 h Ti–Pt-EDG 0.25 99.3 25.3  Lima et al. (2020)  
 Batch reactor, j = 30 mA cm−2, pH0 = 5.4, [C]0 = 50 mg L−1, [Na2SO4] = 0.05 mol L−1, and t = 2 h SnO2-Sb/Ti-Ti 0.25 99.5 70  Wang et al. (2016)  
 Batch reactor, j = 20 mA cm−2, pH0 = 3, [C]0 = 30 mg L−1, [Na2SO4] = 25 g L−1, and t = 1.5 h Doped Sb SnO2-Stainless steel 0.05 100 93  Mu et al. (2019)  
 Batch reactor, j = 40 mA cm−2, pH0 = 7, [C]0 = 15 mg L−1, Chloride medium, and t = 0.33 h BDD (NCD)-Stainless steel 0.1 100 37.4  dos Santos et al. (2022)  
 Real wastewater, j = 40 mA cm−2, pH0 = 7, [C]0 = 15 mg L−1, and t = 1 h BDD (NCD)-Stainless steel 0.1 100 28.7 
 Synthetic urine, j = 40 mA cm−2, pH0 = 7, [C]0 = 15 mg L−1, and t = 1 h BDD (NCD)-Stainless steel 0.1 90.4 32.2 
 Batch reactor, pure water, j = 45 mA cm−2, pH0 = 7, [C]0 = 10 mg L−1, [Na2SO4] = 0.05 M, and t = 5 h BDD-Stainless steel 2.0 99.9   Montenegro-Ayo et al. (2023)  
 Batch reactor, tap water, [Na2SO4] = 0.05 M, [C]0 = 10 mg L−1, j = 45 mA cm−2, pH0 = 6.5, and, t = 5 h BDD-Stainless steal 2.0 95.8   
 Batch reactor, synthetic urine, [C]0 = 10 mg L−1, [Na2SO4] = 0.05 M, j = 45 mA cm−2, pH0 = 7, and t = 5 h BDD-Stainless steel 2.0 77.2   
 Batch reactor, synthetic urine, [C]0 = 30 mg L−1, [Na2SO4] = 0.05 M, j = 45 mA cm−2, pH0 = 7, and t = 5 h BDD-Stainless steel 2.0   94 
Batch reactor, j = 3.5 mA cm−2, pH0 = 3, [C]0 = 10 mg L−1, [NaCl] = 10 mg L−1, and t = 1.5 h  Ti/nanoSnO2MWCN-Stainless steel 0.25 89.61   Esmaelian et al. (2019)  
View Large

Based on the above-mentioned, the objective of this work was the electrochemical degradation of CIP through a DoE-driven optimization in a filter-press-type reactor under batch recirculation mode equipped with two BDD electrodes (both as anode and cathode). For this case, a central composite face-centered experimental design was utilized with three factors namely the initial concentration of CIP ([C]0), initial hydrogen potential (pH0), and current intensity (i) with two response variables namely removal efficiency (η) and operating cost (OCost).

Listen

Reagents and synthetic solution of CIP

Listen

CIP (CAS No: 85721-33-1, MW: 331.34 g mol−1, grade high-performance liquid chromatography (HPLC)), Na2SO4, NaOH, H2SO4 with a purity of 98, 99, 97, and 95–98%, respectively. All chemicals were purchased from Sigma Aldrich Company.

Synthetic solutions of CIP at different initial concentrations were prepared before running each experiment with 0.15 M of Na2SO4 as a supporting electrolyte. Also, 2 M NaOH and H2SO4 solutions, respectively, were prepared to adjust the pH0. It is worth mentioning that all aqueous solutions were prepared with distilled water.

Equipment

Listen
The pH was measured by using a potentiometer Hanna HI2210, and a power supply Gw instek GPR-351OHD was used to provide energy at the filter-press electrochemical reactor. The employed filter-press-type reactor was equipped with BDD electrodes that are separated 1.1 cm from each other. Also, has a reactor volume of 3.52 × 10−5 m3, a height of 20 cm, a width of 4.8 cm, and an electro active area of 32 cm2. In Figure 1, a schematic diagram of the filter-press-type electrochemical system can be observed. Additionally, at the optimal operating conditions, the total organic carbon was determined by employing a TOC Analyzer Shimadzu TOC-6001. For the optimization process, the CIP was quantified by UV-Vis spectrophotometry in a UV-Vis PerkinElmer Lambda 365 spectrophotometer.
Figure 1
Schematic diagram of the experimental set-up with a filter-press-type electrochemical reactor.
View largeDownload slide

Schematic diagram of the experimental set-up with a filter-press-type electrochemical reactor.

Figure 1
Schematic diagram of the experimental set-up with a filter-press-type electrochemical reactor.
View largeDownload slide

Schematic diagram of the experimental set-up with a filter-press-type electrochemical reactor.

Close modal

Analytical procedures

Listen

Removal efficiency

Listen
The removal efficiency (η) of CIP was determined based on absorbance lectures at 272 nm according to Equation (2) (Lupa et al. 2020). For this purpose, the absorbance of samples taken at the beginning (A0) and the end (At) of each run were measured in a spectrophotometer UV-Vis Perkin Elmer Lambda 365. While the η under optimal operating conditions was computed by using Equation (3),
(2)
(3)
where C0 is the initial concentration of CIP and Ct is the concentration of CIP at electrolysis time t (h).

Operating cost

Listen
The operating cost (OCost) was computed by the set of Equations (4)–(6) according to reference (Barrera-Díaz et al. 2018),
(4)
(5)
(6)
where E is the energy (kWh), 3.07 $MXN per kWh is the average price per kWh for commercial use in Mexico (Energy was supplied by the Mexican Company of Federal Commission of Electricity (CFE)), P is the power (kW), i is current intensity (A), ψ is the electrical potential (V).

Mineralization current efficiency

Listen
The mineralization current efficiency (MCE) provides a measurement of the effectiveness of the detoxification of wastewater treatment. The MCE is based on the difference between the initial and final TOC according to the following equation (Guinea et al. 2008),
(7)
where F is the Faraday constant (96,487 C mol−1), Vs is the volume of treatment (L), Δ(TOC)Exp is the change of experimental TOC at electrolysis time (mg L−1), 4.32 × 107 mg s mol−1 h−1 is a conversion factor, m is the number of carbon atoms, and n is the number of electrons required (in this case are 94) to reach the total mineralization of CIP into CO2 and inorganic ions as shown in the following equation
(8)
Since the wastewater treatment employed here uses electricity, it is necessary to compute the energy consumption (EC) because it is a very important economic parameter, which was estimated by using the following equation (Brillas et al. 2009),
(9)
where Ecell is the average cell voltage (V).

HPLC and mass spectrometry

Listen

The CIP degradation progress was conducted by using reverse-phase high-performance liquid chromatography (RP-HPLC) equipped with a photodiode detector (Agilent 6410). The column used was a Hypersil GOLD, 5 μm, 150 × 4.6 mm, the mobile phase was methanol grade HPLC and an aqueous solution of 0.1% formic acid. The isocratic separation was performed at 60/40 (v/v) with a flow rate of 1 mL min−1. The sample was concentrated for Sep-Pak® C18 Cartridge (water-methanol), the injection volume on HPLC was of 15 μL and a temperature of 30 °C was used. To identify the CIP degradation compounds, a triple Quad LC/MS with an electrospray ionization (ESI) source was used. The mass spectrometry (MS) was carried out in positive ion mode, the capillary voltage at 3,000 V, nitrogen gas flow rate at 10 L min−1 and the gas temperature was 350 °C.

Experimental design of CIP electrolysis

Listen

2.5 L of synthetic wastewater of CIP (prepared with different CIP initial concentrations according to experimental design) was prepared and recirculated through the experimental set-up by means of a pump at a volumetric flow rate of 1 L min−1 to homogenize the solution in each run.

The factors studied in this research were pH0, i, and [C]0. All runs were performed by duplicate and the presented results correspond to the mean. The low, medium, and high levels are encoded as −1, 0, and +1, respectively, by using the following equation as shown in Table 2.
(10)
Table 2

Levels and values of operational factors at a volumetric flow rate of 1 L min−1

FactorLevel
− 10+ 1
X1: pH0 4.5 6.5 8.5 
X2: i (A) 3.0 3.5 4.0 
X3: [C]0 (mg L−1) 10.0 30.0 50.0 
FactorLevel
− 10+ 1
X1: pH0 4.5 6.5 8.5 
X2: i (A) 3.0 3.5 4.0 
X3: [C]0 (mg L−1) 10.0 30.0 50.0 
View Large

For this optimization section, the removal efficiency of CIP was measured by UV-Vis spectrophotometric technique. For this purpose, the initial and final (within 5 h treatment) concentration ([C]) were measured for operating conditions according to Table 2. The experimental design was designed by a response surface methodology (RSM) employing a central composite face-centered design (CCFCD). The total experimental runs were 13 which are shown in Table 3. They were designed as follows: 2k−1 factorial points, 2k axial points, and k central points, with k = 3 and two responses (removal efficiency (η) and operating cost (OCost)).

Table 3

CCFCD for removal efficiency of CIP at a volumetric flow rate of 1 L min−1

RunSpace typeFactor
Response
pH0i (A)[C]0 (mg L−1)η (%)OCost ($MXN)
1 Factorial 8.5 4.0 10 95.261 1.048 
2 Center 6.5 3.5 30 96.651 0.839 
3 Axial 6.5 3.5 50 93.459 0.852 
4 Factorial 4.5 3.0 10 96.435 0.663 
5 Center 6.5 3.5 30 96.387 0.841 
6 Axial 6.5 3.5 10 95.776 0.858 
7 Axial 8.5 3.5 30 97.363 0.846 
8 Axial 6.5 4.0 30 96.573 1.053 
9 Axial 6.5 3.0 30 96.565 0.656 
10 Factorial 4.5 4.0 50 96.730 1.044 
11 Factorial 8.5 3.0 50 93.612 0.652 
12 Axial 4.5 3.5 30 97.678 0.860 
13 Center 6.5 3.5 30 96.517 0.863 
RunSpace typeFactor
Response
pH0i (A)[C]0 (mg L−1)η (%)OCost ($MXN)
1 Factorial 8.5 4.0 10 95.261 1.048 
2 Center 6.5 3.5 30 96.651 0.839 
3 Axial 6.5 3.5 50 93.459 0.852 
4 Factorial 4.5 3.0 10 96.435 0.663 
5 Center 6.5 3.5 30 96.387 0.841 
6 Axial 6.5 3.5 10 95.776 0.858 
7 Axial 8.5 3.5 30 97.363 0.846 
8 Axial 6.5 4.0 30 96.573 1.053 
9 Axial 6.5 3.0 30 96.565 0.656 
10 Factorial 4.5 4.0 50 96.730 1.044 
11 Factorial 8.5 3.0 50 93.612 0.652 
12 Axial 4.5 3.5 30 97.678 0.860 
13 Center 6.5 3.5 30 96.517 0.863 
View Large

Three additional experiments were performed at optimal operating conditions to validate the optimal operating conditions found and obtain the kinetic degradation of CIP, pathway reaction rate, EC, the MCE, and the mineralization grade (φ).

Optimization process

Listen
To accomplish this stage, the RSM designed above was employed to perform an analysis of the influence of the factors pH0, i, and [C]0 and their interactions, pH0 × i, pH0 × [C]0, and i × [C]0, on responses (η (%) and OCost ($MXN)) and to establish the optimal operating conditions. Upon performing the 13 runs, a quadratic (Equation (11)) and linear model (Equation (12)) for η (%) and OCost were fitted, respectively, in order to describe the removal efficiency of CIP and compute the optimal operating conditions of the electrochemical degradation treatment of CIP.
(11)
(12)
where β values are the regression coefficients, ε is the random error, and Xi and Xj are the encoded independent variables (given by Equation (10)). To evaluate the reliability and significance of the fitted polynomial functions analysis of variance (ANOVA) must be performed (Regalado-Méndez et al. 2020). It is worth mentioning that all plots and data analysis were carried out in Design Expert V 10.0 software package.
Finally, the accuracy precision of the fitted models was determined by the root mean square error index (RMSE) according to the following equation (Viana et al. 2018),
(13)
where ηi,pred and ηi,exp are the predicted and experimental values for the responses, respectively, and is the total number of experiments.
Listen

Model fitting

Listen
Based on all experimental data given by the set of runs in Table 3, two mathematical models for the removal efficiency of CIP and total operating cost were fitted, which are represented by the following equations, respectively,
(14)
(15)

Influence between the factors

Listen

The negative signs of β values in Equations (14) and (15) in terms of coded variables indicate a negative effect on responses in contrast to positive signs of β values (Zhou et al. 2020). From Equation (14), η (%) increases when increases i (X2) but decreases with pH0 (X1) and [C]0 (X3). Also, the interaction between i and [C]0 (X2X3) has a favorable effect on η (%), while the interaction between pH0 and i (X1X2) and pH0 and [C]0 (X1X3) has a negative effect on η (%). Additionally, in Equation (15), the negative sign of β values indicates a negative effect on OCost and in a similar way, the positive sign of β values indicates a positive effect on OCost.

Analysis of variance

Listen

ANOVA analysis for both selected responses (η and OCost) is shown in Table 4. ANOVA results show that both models fitted (quadratic (Equation (14)) and linear (Equation (15))) are significant since the F-values (224.319 and 1,364.81 for η and OCost, respectively, are greater than P-values (<0.0001 and <0.0001) for both responses (η and OCost, respectively) and that P-values are lower than 0.005 for both responses (η and OCost, respectively) according to reference (Dixit & Yadav 2019). Also, the lack of fit is not significant for both responses (η and OCost) because the F-values (0.2423 and 0.1213 for η and OCost, respectively) are lower than P-values (0.801 and 0.9855) and that P-values are greater than 0.005 for both responses (η and OCost, respectively), which implies that there is a significant correlation between the chosen responses (η and OCost) and the chosen variables (X1, X2 y X3). Also, determination coefficients (R2) were 0.9977 and 0.9978 for η and OCost, respectively, implying that the fitted models have high reproducibility of experimental data for both responses (η and OCost) (Korde et al. 2021). Moreover, the difference between and were 0.0296 and 0.0002 for η and OCost, respectively. The RMSE index performances were 0.2788 and 0.0062 for η and OCost, respectively, indicating that the fitted models (Equations (14) and (15)) have high concordance between predicted and experimental data. Furthermore, the adequate precession ratios for both responses were 48.62 and 95.84 for η and OCost, respectively, indicating an adequate signal since are greater than 4, according to reference (Peralta-Reyes et al. 2022) and the coefficient of variance was 0.1085 and 0.8794% for η and OCost, respectively, which indicates that the fitted models have high reproducibility since the C.V. was less than 10% for both responses (η and OCost). Hence, the models can be used to navigate the design space and are suitable for finding the optimal operating conditions of the electrochemical process employed.

Table 4

ANOVA for responses η (%) and OCost ($MXN)

SourceSum of squareDegree of freedomMean squareF-valueP-valueRemark
Removal efficiency, η (%) 
Model 19.5180 8 2.4397 224.319 <0.0001 Significant 
X1 0.0496 1 0.0496 4.5615 0.0995  
X2 0.000032 1 0.000032 0.0029 0.9593  
X3 2.6842 1 2.6842 246.7987 <0.0001  
X1X2 0.8965 1 0.8965 82.4303 0.0008  
X1X3 0.3097 1 0.3097 28.4808 0.0059  
X2X3 1.1175 1 1.1175 102.7486 0.0005  
 2.4326 1 2.4326 223.6650 0.0001  
 10.6589 1 10.6589 980.0169 <0.0001  
Residual 0.0435 4 0.0108    
Lack of fit 0.0086 2 0.0043 0.2483 0.8010 Not significant 
Pure error 0.0349 2 0.0174    
Cor error 19.5615 12     
R2 = 0.9977; = 0.9933; = 0.9637; Adequate precision = 48.6206; C.V. = 0.11% 
Operating cost, OCost ($MXN) 
Model 0.2298 3 0.0766 1364.81 <0.0001 Significant 
X1 0.0000735 1 0.0000735 1.3092 0.2820  
X2 0.2297 1 0.2297 4,091.8115 <0.0001  
X3 0.0000735 1 0.0000735 1.3092 0.2820  
Residual 0.0005 9 0.0000561    
Lack of fit 0.0001 7 0.0000215 0.1213 0.9855 Not significant 
Pure error 0.0003 2 0.0001    
Cor error 0.2303 12     
R2= 0.9978; = 0.9970; = 0.9968; Adequate precision = 95.8414; C. V. = 0.88% 
SourceSum of squareDegree of freedomMean squareF-valueP-valueRemark
Removal efficiency, η (%) 
Model 19.5180 8 2.4397 224.319 <0.0001 Significant 
X1 0.0496 1 0.0496 4.5615 0.0995  
X2 0.000032 1 0.000032 0.0029 0.9593  
X3 2.6842 1 2.6842 246.7987 <0.0001  
X1X2 0.8965 1 0.8965 82.4303 0.0008  
X1X3 0.3097 1 0.3097 28.4808 0.0059  
X2X3 1.1175 1 1.1175 102.7486 0.0005  
 2.4326 1 2.4326 223.6650 0.0001  
 10.6589 1 10.6589 980.0169 <0.0001  
Residual 0.0435 4 0.0108    
Lack of fit 0.0086 2 0.0043 0.2483 0.8010 Not significant 
Pure error 0.0349 2 0.0174    
Cor error 19.5615 12     
R2 = 0.9977; = 0.9933; = 0.9637; Adequate precision = 48.6206; C.V. = 0.11% 
Operating cost, OCost ($MXN) 
Model 0.2298 3 0.0766 1364.81 <0.0001 Significant 
X1 0.0000735 1 0.0000735 1.3092 0.2820  
X2 0.2297 1 0.2297 4,091.8115 <0.0001  
X3 0.0000735 1 0.0000735 1.3092 0.2820  
Residual 0.0005 9 0.0000561    
Lack of fit 0.0001 7 0.0000215 0.1213 0.9855 Not significant 
Pure error 0.0003 2 0.0001    
Cor error 0.2303 12     
R2= 0.9978; = 0.9970; = 0.9968; Adequate precision = 95.8414; C. V. = 0.88% 
View Large

Parity plots are displayed in Figure 2, which show that there are exceptionally good correlations between experimental data and predicted data by fitted models (Equations (14) and (15)) for both chosen responses (η and OCost).
Figure 2
Experimental data versus predicted values. (a) Response η; (b) Response OCost.
View largeDownload slide

Experimental data versus predicted values. (a) Response η; (b) Response OCost.

Figure 2
Experimental data versus predicted values. (a) Response η; (b) Response OCost.
View largeDownload slide

Experimental data versus predicted values. (a) Response η; (b) Response OCost.

Close modal
Perturbation diagrams (Figure 3(a) and 3(b)) reveal that the most important factor for response η is the initial concentration of CIP ([C]0 (X3)) following the initial hydrogen potential (pH0 (X1)) since it presents a pronounced curvature. Also, the less important factor is the current intensity (i (X2)) because has a relatively small slope. While for response OCost, the most important factor is the current intensity (i (X2)) because of their high slope. Also, the less important factors were [C]0 (X3) and pH0 (X1) since their slopes are practically null.
Figure 3
Perturbation diagrams. (a) Response η and (b) Response OCost.
View largeDownload slide

Perturbation diagrams. (a) Response η and (b) Response OCost.

Figure 3
Perturbation diagrams. (a) Response η and (b) Response OCost.
View largeDownload slide

Perturbation diagrams. (a) Response η and (b) Response OCost.

Close modal

Optimization of the responses (η and OCost)

Listen

To optimize the chosen responses (η (Equation (14) and OCost (Equation (15))), a multi-objective optimization was performed by using the Design Expert V.10 software package. Additional information to be supplied in the software, such as the optimization criteria are shown in Table 5 in which all factors (X1, X2, and X3) and both responses (η and OCost) have the same importance (+++).

Table 5

Restrictions and optimization criteria for the electrochemical degradation of CIP

Limits
ResponseObjectiveMinMaxUnitImportance
pH0 Is in range 4.50 8.50 Dimensionless ++ + 
i Is in range 3.00 4.00 A ++ + 
[C]0 Is in range 10.00 50.0 mg L−1 ++ + 
η Maximize 93.46 97.68 % ++ + 
OCost Minimize 0.65 1.05 $MXN ++ + 
Limits
ResponseObjectiveMinMaxUnitImportance
pH0 Is in range 4.50 8.50 Dimensionless ++ + 
i Is in range 3.00 4.00 A ++ + 
[C]0 Is in range 10.00 50.0 mg L−1 ++ + 
η Maximize 93.46 97.68 % ++ + 
OCost Minimize 0.65 1.05 $MXN ++ + 
View Large

The surface response plots are depicted in Figure 4(a)–4(f), which can be obtained by plotting Equations (14) and (15), respectively. In these 3D plots, the influence of pH0, i, and [C]0 on both responses (η and OCost) can be observed, maintaining constant one independent variable and varying the other two independent variables between study intervals.
Figure 4
(a) 3D plot of OCost as a function of pH0 and i; (b) 3D plot of OCost as a function of i and [C]0; (c) 3D plot of OCost as a function of pH0 and [C]0; (d) 3D plot of η as a function of pH0 and i; (e) 3D plot of η as a function of i and [C]0; (f) 3D plot of η as a function of pH0 and [C]0; (g) Overlay plot as a function of pH0 and i; (h) Desirability bar chart. Please refer to the online version of this paper to see this figure in colour: https://dx.doi.org/10.2166/wst.2023.279.
View largeDownload slide

(a) 3D plot of OCost as a function of pH0 and i; (b) 3D plot of OCost as a function of i and [C]0; (c) 3D plot of OCost as a function of pH0 and [C]0; (d) 3D plot of η as a function of pH0 and i; (e) 3D plot of η as a function of i and [C]0; (f) 3D plot of η as a function of pH0 and [C]0; (g) Overlay plot as a function of pH0 and i; (h) Desirability bar chart. Please refer to the online version of this paper to see this figure in colour: https://dx.doi.org/10.2166/wst.2023.279.

Figure 4
(a) 3D plot of OCost as a function of pH0 and i; (b) 3D plot of OCost as a function of i and [C]0; (c) 3D plot of OCost as a function of pH0 and [C]0; (d) 3D plot of η as a function of pH0 and i; (e) 3D plot of η as a function of i and [C]0; (f) 3D plot of η as a function of pH0 and [C]0; (g) Overlay plot as a function of pH0 and i; (h) Desirability bar chart. Please refer to the online version of this paper to see this figure in colour: https://dx.doi.org/10.2166/wst.2023.279.
View largeDownload slide

(a) 3D plot of OCost as a function of pH0 and i; (b) 3D plot of OCost as a function of i and [C]0; (c) 3D plot of OCost as a function of pH0 and [C]0; (d) 3D plot of η as a function of pH0 and i; (e) 3D plot of η as a function of i and [C]0; (f) 3D plot of η as a function of pH0 and [C]0; (g) Overlay plot as a function of pH0 and i; (h) Desirability bar chart. Please refer to the online version of this paper to see this figure in colour: https://dx.doi.org/10.2166/wst.2023.279.

Close modal

Figure 4(a)–4(c) forms a 3D plane since contour plots are shaped by linear functions. Changes in pH0 and [C]0 do not have a significant effect on OCost but an increase in i has a negative effect on OCost because the OCost must be minimized.

Figure 4(d)–4(f) forms a saddle point since the contour plots are shaped by hyperboles. Also, an increase in pH0 has a positive effect on η (see Figure 4(d) and 4(f)). An increase in [C]0 has a negative effect on η (see Figure 4(e) and 4(f)). An increase in i has a positive effect on η in Figure 4(d) but a negative effect on η in Figure 4(e).

The maximum η (97.67%) and minimum OCost ($0.6521 MXN or 0.0354 US$) were achieved at pH0 = 8.49, i = 3 A, and [C]0 = 33.26 mg L−1 within 5 h of electrolysis time with a global desirability of 99.97% indicating that all objectives were reached (see Figure 4h). Figure 4g depicts with black lines the restrictions (see Table 5) and a yellow section which represents the optimal operation region, which is bounded by a pH0 = 2.5–9.5 and i = 3–4 A, and the gray section that represents the not feasible region.

Model verification

Listen

Three complementary runs were performed at optimal operating conditions to carry out the verification of the fitting models (Equations (14) and (15)). It is worth mentioning that the response η was followed by UV-Vis and HPLC. The average values for experimental response η were 94.17% by HPLC and 93.43% by UV-Vis, the error between these values was 0.78%, indicating that the use of the spectrophotometric UV-Vis technique is acceptable in this study for the optimization process. For complex effluents, however, the use of HPLC is advised. The computed percentage error for response η was 3.6 and 4.35% for HPLC and UV-Vis, respectively. Also, the OCost was $0.664 MXN (0.0325 US$, 1 US$/$18.4238 MXN), with an error of 1.8%. Hence, the low (less than 5%) deviation between the values of experimental and modeled data corroborates the effectiveness of the optimization process employed in this work.

An additional trial at optimal operating conditions was carried out to show that the CIP is degraded with time (see Figure 5). Also, in Figure 5, the abatement of concentration of CIP is depicted.
Figure 5
Abatement of concentration of CIP as a function of time by HPLC at optimal operating conditions (pH0 = 8.49, i = 3 A and [C]0 = 33.26 mg L−1, and Q = 1 L min−1).
View largeDownload slide

Abatement of concentration of CIP as a function of time by HPLC at optimal operating conditions (pH0 = 8.49, i = 3 A and [C]0 = 33.26 mg L−1, and Q = 1 L min−1).

Figure 5
Abatement of concentration of CIP as a function of time by HPLC at optimal operating conditions (pH0 = 8.49, i = 3 A and [C]0 = 33.26 mg L−1, and Q = 1 L min−1).
View largeDownload slide

Abatement of concentration of CIP as a function of time by HPLC at optimal operating conditions (pH0 = 8.49, i = 3 A and [C]0 = 33.26 mg L−1, and Q = 1 L min−1).

Close modal
Additionally, a TOC analysis at optimal operating conditions was carried out to analyze the fraction of TOC abatement as a function of time (see Figure 6(a)), MCE (%) as a function of time (see Figure 6(b)), and EC (kWh g−1 TOC) as a function of time (see Figure 6(c)).
Figure 6
TOC analysis of CIP at optimal operating conditions (pH0 = 8.49, i = 3 A and [C]0 = 33.26 mg L−1, and Q = 1 L min−1). (a) Fraction of TOC abatement; (b) MCE (%); (c) EC (kWh/g TOC).
View largeDownload slide

TOC analysis of CIP at optimal operating conditions (pH0 = 8.49, i = 3 A and [C]0 = 33.26 mg L−1, and Q = 1 L min−1). (a) Fraction of TOC abatement; (b) MCE (%); (c) EC (kWh/g TOC).

Figure 6
TOC analysis of CIP at optimal operating conditions (pH0 = 8.49, i = 3 A and [C]0 = 33.26 mg L−1, and Q = 1 L min−1). (a) Fraction of TOC abatement; (b) MCE (%); (c) EC (kWh/g TOC).
View largeDownload slide

TOC analysis of CIP at optimal operating conditions (pH0 = 8.49, i = 3 A and [C]0 = 33.26 mg L−1, and Q = 1 L min−1). (a) Fraction of TOC abatement; (b) MCE (%); (c) EC (kWh/g TOC).

Close modal
It can be observed in Figure 6(a) that the maximum mineralization efficiency was 80% within 5 h of treatment according to Equation (8). These results indicate that there are some recalcitrant by-products and to achieve the total mineralization (CO2, H2O, and ions) a longer reaction time is required (Câmara Cardozo et al. 2022). Figure 6(b) shows a maximum of MCE at 5 h of electrolysis, indicating that CIP is transformed into byproducts easily degradable to transform into CO2 due to hydroxyl radical produced following Equation (1) (Flox et al. 2007), which minimizes the parasitic reactions given by Equations (16)–(18) (Lanzarini-Lopes et al. 2017). Also, in Figure 6(c), a minimal value of EC (5.11 kWh g−1 TOC) is achieved at 5 h of treatment in agreement with the removed fraction of TOC (Brillas & Martínez-Huitle 2015). Furthermore, the maximum EC (23.25 kWh g−1 TOC) was at 1 h of electrolysis time.
(16)
(17)
(18)
To understand if the electrochemical process employed in this research is controlled by mass transport or current, the limit current density (jlim) must be computed by using the following equations (Chang et al. 2017),
(19)
(20)
where n is the number of electrons, F is the Faraday constant, km average mass transfer coefficient, [C]0 is the initial concentration of CIP, TOCt is the TOC at any time t, TOC0 is the initial TOC, Vs is the volume treatment, and A is the geometric area of the electrode.

For this case of study, the jlim = 3.09 × 10−3 A cm−2, japply = 9.37 × 10−2 A cm−2, and km = 7.04 × 10−5 m s−1 were found. Hence, since the japply > jlim the electrochemical process is controlled by mass transport according to reference (Kapałka et al. 2008).

Finally, the extent of electrochemical combustion (φ) of the removed CIP was estimated as the ratio of the percentage TOC removed and the percentage CIP removed, according to Equation (21) (Miwa et al. 2006) in order to find the level of total combustion/removal. If φ tends towards 1, then almost all the CIP removed is subsequently mineralized into CO2.
(21)
where %TOCremoved is the percentage of TOC removed and %CIPremoved is the percentage of CIP removed.
The maximum value of φ was 0.85, indicating that the major amount of CIP was mineralized within 5 h of treatment time. These results corroborate the fact that BDD electrodes promote efficiently the complete oxidation of organic compounds such as CIP into CO2. Also, the profile of φ is shown in Figure 7.
Figure 7
Extent of electrochemical combustion (φ) of the removed CIP at optimal operating conditions (pH0 = 8.49, i = 3 A and [C]0 = 33.26 mg L−1,Q = 1 L min−1, and t = 5 h).
View largeDownload slide

Extent of electrochemical combustion (φ) of the removed CIP at optimal operating conditions (pH0 = 8.49, i = 3 A and [C]0 = 33.26 mg L−1,Q = 1 L min−1, and t = 5 h).

Figure 7
Extent of electrochemical combustion (φ) of the removed CIP at optimal operating conditions (pH0 = 8.49, i = 3 A and [C]0 = 33.26 mg L−1,Q = 1 L min−1, and t = 5 h).
View largeDownload slide

Extent of electrochemical combustion (φ) of the removed CIP at optimal operating conditions (pH0 = 8.49, i = 3 A and [C]0 = 33.26 mg L−1,Q = 1 L min−1, and t = 5 h).

Close modal

Degradation kinetics

Listen
Figure 8 depicts the electrochemical degradation of CIP followed by HPLC. An asymptotic behavior is observed, which is due to the intensive contact between CIP and generated in the electrode surface. Therefore, the CIP is relatively easily diffused on the BDD electrode surface and reacts with the ⁠. Based on the behavior displayed in Figure 8, the electrochemical degradation of CIP follows a pseudo-first-order kinetic rate. From a linear regression analysis, the kapp value was 0.53 h−1 with a determination coefficient (R2) value of 0.9921. Moreover, a good linear correlation (R2 > 0.9) suggests a high amount of production on the BDD electrode surface (Sirés et al. 2008).
(22)
where kapp is the apparent kinetic reaction constant and t is the reaction time.
Figure 8
Degradation profile of CIP at optimal operating conditions (pH0 = 8.49, i = 3 A and [C]0 = 33.26 mg L−1, Q = 1 L min−1, and t = 5 h).
View largeDownload slide

Degradation profile of CIP at optimal operating conditions (pH0 = 8.49, i = 3 A and [C]0 = 33.26 mg L−1, Q = 1 L min−1, and t = 5 h).

Figure 8
Degradation profile of CIP at optimal operating conditions (pH0 = 8.49, i = 3 A and [C]0 = 33.26 mg L−1, Q = 1 L min−1, and t = 5 h).
View largeDownload slide

Degradation profile of CIP at optimal operating conditions (pH0 = 8.49, i = 3 A and [C]0 = 33.26 mg L−1, Q = 1 L min−1, and t = 5 h).

Close modal

Finally, the kinetic order of the electrochemical mineralization of CIP was found by analyzing the TOC decay behavior depicted in Figure 6(a). After a linear regression analysis, it was found that the TOC removal fits a pseudo-zero-order kinetics rate, −dTOC/dt = k0, where k0 is the pseudo-zero-order rate constant (k0 = 3.3 mg TOC L−1 min−1 with an R2 value of 0.9566).

By-products identification

Listen

The identification of the by-products remaining after the electrochemical treatment at optimal operating conditions was performed by LC–MS. By this technique, 18 chemical organic structures were identified, and these are presented in Table 6. These chemical compounds were formed mainly by the attack of the on the rings of quinolone and cyclic diamines, which are the precursor molecules of CIP.

Table 6

Identified compounds after the electrochemical degradation of CIP at optimal operating conditions

Molecular formulaChemical structureExact mass (m/z)RT (min)Degradation time (h)
C17H18FN3O3  331.64052 8.86 0.5 
C17H18FN3O5  363.89234 9.86 1 
C17H16FN3O5  361.85437 10.80 1, 2 
C15H16FN3O3  305.79872 7.86 1, 2, 3 
C14H11FN2O4  290.18696 10.39 2, 3, 4 
C13H11FN2O3  263.07752 11.13 2, 3, 4 
C17H9FN2O3  236.38752 12.80 3, 4 
C17H18FN3O4  347.66728 8.93 0.5, 1 
C17H20FN3O4  349.40359 13.14 1, 2 
C9H12FN3O  197.52390 5.62 2, 3 
C7H6FN2O  153.17656 2.46 3, 4 
C14H14FN3O4  307.59945 9.23 0.5, 1 
C14H12FN3O5  321.28149 9.72 1, 2 
C11H12FN3O4  269.99877 13.0 2, 3 
C6H6FN2  125.97859 13.86 3, 4 
C6H6FN  111.50865 6.56 3, 4 
C6H10FN2  134.67847 1,43 4 
C7H7FN20  138.96735 4.56 3, 4 
Molecular formulaChemical structureExact mass (m/z)RT (min)Degradation time (h)
C17H18FN3O3  331.64052 8.86 0.5 
C17H18FN3O5  363.89234 9.86 1 
C17H16FN3O5  361.85437 10.80 1, 2 
C15H16FN3O3  305.79872 7.86 1, 2, 3 
C14H11FN2O4  290.18696 10.39 2, 3, 4 
C13H11FN2O3  263.07752 11.13 2, 3, 4 
C17H9FN2O3  236.38752 12.80 3, 4 
C17H18FN3O4  347.66728 8.93 0.5, 1 
C17H20FN3O4  349.40359 13.14 1, 2 
C9H12FN3O  197.52390 5.62 2, 3 
C7H6FN2O  153.17656 2.46 3, 4 
C14H14FN3O4  307.59945 9.23 0.5, 1 
C14H12FN3O5  321.28149 9.72 1, 2 
C11H12FN3O4  269.99877 13.0 2, 3 
C6H6FN2  125.97859 13.86 3, 4 
C6H6FN  111.50865 6.56 3, 4 
C6H10FN2  134.67847 1,43 4 
C7H7FN20  138.96735 4.56 3, 4 
View Large

Reaction pathway

Listen
The fragmentation patterns obtained through ESI-MS analysis at different exposure times of ciprofloxacin to AOP revealed three possible degradation routes (see Figure 9 and Table 6). In the first one, a double hydroxylation was observed on the piperazine ring (m/z: 363.89), followed by the oxidation of an alcohol group and the formation of a carbonyl group on the same ring (m/z: 361.85). This led to the fragmentation of the piperazine ring, resulting in secondary and primary amines (m/z: 305.80). Subsequently, the elimination of a primary amine formed a compound with m/z: 290.18. The last CO2 molecule allowed the formation of the compound with m/z: 263.08. The compound with m/z: 263.08 underwent further degradation, involving the removal of a carboxyl (COOH) group and the hydropyridine ring. This resulted in the formation of an anhydrous group on the quinoline ring (m/z: 236.39). The carbonyl group facilitated the formation of inorganic CO2, leading to the formation of a trisubstituted benzene ring (m/z: 125.97). The last two compounds identified at exposure times exceeding 5 h were aromatic compounds and cyclohexane (sp3), with masses of m/z: 111.51 and m/z: 134.67, respectively.
Figure 9
Reaction pathway for electrochemical degradation of CIP with two BDD electrodes. Optimal reaction conditions: pH0 = 8.49, i = 3 A, [C]0 = 33.26 mg L−1, and Q = 1 L min−1.
View largeDownload slide

Reaction pathway for electrochemical degradation of CIP with two BDD electrodes. Optimal reaction conditions: pH0 = 8.49, i = 3 A, [C]0 = 33.26 mg L−1, and Q = 1 L min−1.

Figure 9
Reaction pathway for electrochemical degradation of CIP with two BDD electrodes. Optimal reaction conditions: pH0 = 8.49, i = 3 A, [C]0 = 33.26 mg L−1, and Q = 1 L min−1.
View largeDownload slide

Reaction pathway for electrochemical degradation of CIP with two BDD electrodes. Optimal reaction conditions: pH0 = 8.49, i = 3 A, [C]0 = 33.26 mg L−1, and Q = 1 L min−1.

Close modal

In the second pathway, the formation of a carbonyl group is the initial step for bond breakage, like the first pathway. The difference is the place where radicals attacked the ciprofloxacin ring. This pathway begins with the hydroxylation of the quinolone ring (m/z: 347.66), followed by the opening of the hydropyridine ring from the quinolone, identified as m/z: 349.40. Subsequently, the loss of the piperazine ring (m/z: 363.89) and the formation of an aldehyde group in the quinolone ring are necessary for the formation of m/z: 197.52. Completely, the molecule with m/z: 363.89 initially loses the piperazine ring, resulting in the formation of compounds with masses of m/z: 138.96 and m/z: 125.97, respectively. Both pathways converge in the formation of 4-fluoro-1,3-diaminobenzene (m/z: 125.97).

The third pathway begins with the loss of the cyclopropyl group and joins the hydroxylation of the piperazine ring m/z: 307.60, then the oxidation of the alcohol group forms the carbonyl group and there is another hydroxylation on the same piperazine ring, obtained the molecule m/z: 321.28. Subsequently, the degraded hydropyridine could be the origin of the molecule with m/z: 269.99. The analysis of the two pathways described above suggests that the degradation of the molecule with m/z: 269.99 could follow the loss of piperazine and quinoline, obtaining the formation of 4-fluoro-1,3-diaminobenzene (m/z: 125.97).

The description of the degradation path reveals that the action of the radical produces intermediates with carbonyl groups and smaller sizes, which facilitates the degradation of the molecules into CO2 and H2O consistent with previous results (Chen et al. 2021). In this research, three degradation routes are proposed in which the most susceptible groups are the piperazine and quinolone ring, as well as the cyclopropyl group, which coincides with what was reported in Van Doorslaer et al. (2014) and Chen et al. (2021). In relation to the determined molecules and other degradation routes, it is observed that the first pathway presents a similar behavior to those reported in anodic oxidations with graphite and silver electrodes at a concentration of 20 mg L−1 of CIP, a current density of 6.25 mA cm−2, and a concentration of 0.1 mol L−1 of Na2SO4 (Chen et al. 2021). On the other hand, pathway two shows compounds like those reported in flow reactors with BDD electrodes at a current density of j = 30 mA cm−2, pH = 10.0, a flow rate of 2.5 L min−1, volume treated of 0.5 L, [CCIP] = 50 mg L−1 in 0.1 M Na2SO4 (Wachter et al. 2019a). Finally, the third pathway is consistent with the results obtained in experiments with graphite and silver electrodes (Chen et al. 2021). It is important to mention that the by-products obtained are carbon compounds with fluorine atoms, which can be perfluoroalkyl pollutants (PFAS) that are frequently detected in water bodies. The stability of the carbon-fluorine bond hinders its degradation and favors its accumulation in plants and animals. The methods used (biotic and abiotic) for water purification do not degrade PFAS, which means that in industrialized societies a person can consume up to 70 ng of PFAS per liter of drinking water (Sunderland et al. 2019). Therefore, it is necessary to develop AOPs to achieve the complete degradation of the CIP.

It is also plausible that sulfate anions could affect the CIP oxidation by either scavenging hydroxyl radicals (Equation (23)) which may affect negatively the organic compounds oxidation (do Vale-Júnior et al. 2019)
(23)
Also, the sulfate radicals (⁠⁠) can be recombined with water to form hydroxyl radicals (Amado-Piña et al. 2017), as described by Equation (24). Therefore, this reaction is expected to favor the removal efficiency of the organic contaminant (e.g., CIP). It is worth mentioning, that this reaction only occurs when two BDD electrodes are employed.
(24)
In this case, the scavenging of hydroxyl radicals is not expected to readily proceed since has been usually related to high amounts of sulfate ions (Duan et al. 2018; Ganiyu et al. 2021). Also, a series of other reactions could take place (Ganiyu et al. 2021) (Equations (25)–(30)). The amount of sulfate radical (⁠⁠) and persulfate ion (⁠⁠) is also expected to increase when the electro-oxidation process is conducted at temperatures higher than 40 °C (Ganiyu & Martínez-Huitle 2019; Shin et al. 2019),
(25)
(26)
(27)
(28)
(29)
(30)

In this study, the electrochemical oxidation of the CIP was carried out at a controlled temperature of 25 °C by using a heat exchanger. Thus, even if the previous reactions proceed, the impact on the CIP is expected to be low in comparison with the well-known production of hydroxyl radicals on BDD electrodes (Amado-Piña et al. 2022).

This agrees with those findings by Farhat et al. (2015), who mention that there is not enough evidence that activated persulfate species have a large effect on the removal efficiency of the organic compound when BDD anode is employed. Also, there is not enough evidence that high supporting electrolyte concentrations (Na2SO4, greater than 0.05 M) form high amounts of sulfate radicals and persulfate, when two BDD electrodes are employed. However, an excess of could inhibit the removal efficiency of organic compounds through the self-quenching of or the reaction between and its maternal persulfate ion (⁠⁠) molecule (see Equation (31)) when the concentration of persulfate is higher than the optimal concentration (Cai et al. 2018; Zhi et al. 2020).
(31)

A comparison of the different electrochemical degradation processes of CIP carried out in a flow reactor is shown in Table 7. In this table is distinguished that only this work reports the operating cost. Although the employed process achieves less degradation efficiency (94.17%) than all references given in Table 7 (100%) (Wachter et al. 2019a; dos Santos et al. 2022) the used volume in this research is 2.5–8.33 times greater than that employed in the literature. Albeit the mineralization efficiency achieved in reference (Wachter et al. 2019b) (100%) is greater than this research (80%), the volume used in this study is five times greater than that used in the literature. Given the results displayed in this work, the electrochemical treatment employed here could be considered adequate technology and profitable to be applied to wastewater that contains emerging contaminants such as CIP.

Table 7

Comparison of the different electrochemical degradation processes of CIP in flow reactors

Reaction environmental conditions
Results
OptimizedNon optimizedElectrodes Cath-AnV (L)ED (%)TOC (%)OCost (US$ L−1)Ref.
Q = 1 L/min, pH0 = 8.49, i = 3 A, [C]0 = 33.26 mg L−1, and t = 5 h  BDD-BDD 2.5 94.17 80 0.014 This work 
 Q = 2.5 L min−1, pH0 = 10, j = 30 mA cm−2, [C]0 = 50 mg L−1, [Na2SO4] = 0.1 mol L−1, and t = 10 h  BDD-Stainless steel 304 0.5 100 100  Wachter et al. (2019a)  
 Q = 6.5 L min−1, pH0 = 10, j = 30 mA cm−2, [C]0 = 50 mg L−1, [Na2SO4] = 0.1 mol L−1, and t = 5 h Ni-TiPt/β-PBO2-Ni BDD-Stainless steel 304 0.5 100 75  Wachter et al. (2019b)  
 Q = 7 L min−1, j = 10 mA cm−2, [C]0 = 100 mg L−1, [NaCl] = 0.1 mol L−1, and t = 8 h BDD-Stainless steel 304 1.0 100 80  Carneiro et al. (2020)  
 Q = 0.4 L min−1, j = 30 mA cm−2, pH0 = 7, [C]0 = 10 mg L−1, and t = 0.33 h BDD-Ti 0.3 100   Li et al. (2019)  
Reaction environmental conditions
Results
OptimizedNon optimizedElectrodes Cath-AnV (L)ED (%)TOC (%)OCost (US$ L−1)Ref.
Q = 1 L/min, pH0 = 8.49, i = 3 A, [C]0 = 33.26 mg L−1, and t = 5 h  BDD-BDD 2.5 94.17 80 0.014 This work 
 Q = 2.5 L min−1, pH0 = 10, j = 30 mA cm−2, [C]0 = 50 mg L−1, [Na2SO4] = 0.1 mol L−1, and t = 10 h  BDD-Stainless steel 304 0.5 100 100  Wachter et al. (2019a)  
 Q = 6.5 L min−1, pH0 = 10, j = 30 mA cm−2, [C]0 = 50 mg L−1, [Na2SO4] = 0.1 mol L−1, and t = 5 h Ni-TiPt/β-PBO2-Ni BDD-Stainless steel 304 0.5 100 75  Wachter et al. (2019b)  
 Q = 7 L min−1, j = 10 mA cm−2, [C]0 = 100 mg L−1, [NaCl] = 0.1 mol L−1, and t = 8 h BDD-Stainless steel 304 1.0 100 80  Carneiro et al. (2020)  
 Q = 0.4 L min−1, j = 30 mA cm−2, pH0 = 7, [C]0 = 10 mg L−1, and t = 0.33 h BDD-Ti 0.3 100   Li et al. (2019)  
View Large

Listen

CIP was degraded by an electro-oxidation process with BDD electrodes in a filter-press-type reactor. The optimal operating conditions were: pH0 = 8.49, i = 3 A, [C]0 = 33.26 mg L−1 at Q = 1 L min−1 within 5 h of electrolysis time.

The employed wastewater treatment in this research is an efficient and profitable technology because reaches high degradation efficiency (94.17%) and mineralization efficiency (80%) with a minimum operational cost (0.014 US$ L−1 or $MXN 0.2661 L−1).

The optimization by RSM was successful in the electrochemical degradation of CIP because the differences between experimental data and predicted values from fitted models were 4.3 and 1.8% for η and Ocost, respectively.

The wastewater treatment proposed is environmentally friendly since there are not any solid residues produced and the EC is low (5.11 kWh g−1 TOC), which could be supplied easily by solar panels.

Listen

All relevant data are included in the paper or its Supplementary Information.

Listen

The authors declare there is no conflict.

Amado-Piña
D.
,
Roa-Morales
G.
,
Barrera-Díaz
C.
,
Balderas-Hernandez
P.
,
Romero
R.
,
Martín del Campo
E.
 &
Natividad
R.
2017
Synergic effect of ozonation and electrochemical methods on oxidation and toxicity reduction: phenol degradation
.
Fuel
198
,
82
–
90
.
Available from: https://www.sciencedirect.com/science/article/pii/S0016236116310808.
Google Scholar
Crossref
Search ADS
 
Amado-Piña
D.
,
Roa-Morales
G.
,
Molina-Mendieta
M.
,
Balderas-Hernández
P.
,
Romero
R.
,
Barrera Díaz
C. E.
 &
Natividad
R.
2022
E-peroxone process of a chlorinated compound: oxidant species, degradation pathway and phytotoxicity
.
Journal of Environmental Chemical Engineering
10
(
4
),
108148
.
Available from: https://www.sciencedirect.com/science/article/pii/S2213343722010211.
Google Scholar
Crossref
Search ADS
 
Arman
N. Z.
,
Salmiati
S.
,
Aris
A.
,
Salim
M. R.
,
Nazifa
T. H.
,
Muhamad
M. S.
 &
Marpongahtun
M.
2021
A review on emerging pollutants in the water environment: existences, health effects and treatment processes
.
Water (Switzerland)
13
(
22
), 3258. https://doi.org/10.3390/w13223258.
Google Scholar
 
Ashfaq
M.
,
Khan
K. N.
,
Rasool
S.
,
Mustafa
G.
,
Saif-Ur-Rehman
M.
,
Nazar
M. F.
,
Sun
Q.
 &
Yu
C. P.
2016
Occurrence and ecological risk assessment of fluoroquinolone antibiotics in hospital waste of Lahore, Pakistan
.
Environmental Toxicology and Pharmacology
42
,
16
–
22
.
Google Scholar
Crossref
Search ADS
PubMed
 
Brillas
E.
,
Sirés
I.
 &
Oturan
M. A.
2009
Electro-Fenton process and related electrochemical technologies based on Fenton's reaction chemistry
.
Chemical Reviews
109
(
12
),
6570
–
6631
.
https://doi.org/10.1021/cr900136g
.
Google Scholar
Crossref
Search ADS
PubMed
 
Câmara Cardozo
J.
,
da Silva
D. R.
,
Martínez-Huitle
C. A.
,
Quiroz
M. A.
 &
dos Santos
E. V.
2022
Photovoltaic electrochemically driven degradation of Calcon dye with simultaneous green hydrogen production
.
Materials
15
(
21
).
Available from: https://www.mdpi.com/1996-1944/15/21/7445.
Google Scholar
 
Cappelli
F.
,
Longoni
O.
,
Rigato
J.
,
Rusconi
M.
,
Sala
A.
,
Fochi
I.
,
Palumbo
M. T.
,
Polesello
S.
,
Roscioli
C.
,
Salerno
F.
,
Stefani
F.
,
Bettinetti
R.
 &
Valsecchi
S.
2022
Suspect screening of wastewaters to trace anti-COVID-19 drugs: potential adverse effects on aquatic environment
.
Science of The Total Environment
824
,
153756
.
Available from: https://www.sciencedirect.com/science/article/pii/S0048969722008488.
Google Scholar
Crossref
Search ADS
PubMed
 
Chaplin
B. P.
2014
Critical review of electrochemical advanced oxidation processes for water treatment applications
.
Environmental Sciences: Processes and Impacts
16
(
6
),
1182
–
1203
.
Google Scholar
 
Chen
Z.
,
Lai
W.
,
Xu
Y.
,
Xie
G.
,
Hou
W.
,
Zhanchang
P.
,
Kuang
C.
 &
Li
Y.
2021
Anodic oxidation of ciprofloxacin using different graphite felt anodes: kinetics and degradation pathways
.
Journal of Hazardous Materials
405
,
124262
.
Available from: https://www.sciencedirect.com/science/article/pii/S0304389420322524.
Google Scholar
Crossref
Search ADS
PubMed
 
dos Santos
A. J.
,
Fortunato
G. V.
,
Kronka
M. S.
,
Vernasqui
L. G.
,
Ferreira
N. G.
 &
Lanza
M. R. V.
2022
Electrochemical oxidation of ciprofloxacin in different aqueous matrices using synthesized boron-doped micro and nano-diamond anodes
.
Environmental Research
204
,
112027
.
Available from: https://www.sciencedirect.com/science/article/pii/S0013935121013220.
Google Scholar
Crossref
Search ADS
PubMed
 
do Vale-Júnior
E.
,
dos Santos
A. J.
,
da Silva
D. R.
,
Fajardo
A. S.
 &
Martínez-Huitle
C. A.
2019
Electrochemical technologies for detecting and degrading benzoquinone using diamond films
.
ChemElectroChem
6
(
17
),
4383
–
4390
.
https://doi.org/10.1002/celc.201900541
.
Google Scholar
Crossref
Search ADS
 
Farhat
A.
,
Keller
J.
,
Tait
S.
 &
Adjenovic
J.
2015
Removal of persistent organic contaminants by electrochemically activated sulfate
.
Environmental Science & Technology
49
(
24
),
14326
–
14333
.
https://doi.org/10.1021/acs.est.5b02705
.
Google Scholar
Crossref
Search ADS
 
Firdaus
I.
,
Yaqub
A.
,
Ajab
H.
,
Khan
I.
,
Ahmed
B.
,
Amin
Z.
,
Baig
A.
 &
Isa
M. H.
2021
Electrochemical oxidation of amoxicillin, ciprofloxacin and erythromycin in water: effect of experimental factors on COD removal
.
Pakistan Journal of Pharmaceutical Sciences
34
,
119
–
128
.
Google Scholar
PubMed
 
Flox
C.
,
Cabot
P.-L.
,
Centellas
F.
,
Garrido
J. A.
,
Rodríguez
R. M.
,
Arias
C.
 &
Brillas
E.
2007
Solar photoelectro-Fenton degradation of cresols using a flow reactor with a boron-doped diamond anode
.
Applied Catalysis B: Environmental
75
(
1
),
17
–
28
.
Available from: https://www.sciencedirect.com/science/article/pii/S0926337307000951.
Google Scholar
 
Kapałka
A.
,
Fóti
G.
 &
Comninellis
C.
2008
Kinetic modelling of the electrochemical mineralization of organic pollutants for wastewater treatment
.
Journal of Applied Electrochemistry
38
(
1
),
7
–
16
.
https://doi.org/10.1007/s10800-007-9365-6
.
Google Scholar
Crossref
Search ADS
 
Khan
N. A.
,
Khan
S. U.
,
Ahmed
S.
,
Farooqi
I. H.
,
Yousefi
M.
,
Mohammadi
A. A.
 &
Changani
F.
2020
Recent trends in disposal and treatment technologies of emerging-pollutants- A critical review
.
TrAC Trends in Analytical Chemistry
122
,
115744
.
Available from: https://www.sciencedirect.com/science/article/pii/S0165993619305722.
Google Scholar
Crossref
Search ADS
 
Kim
T. K.
,
Kim
T.
,
Park
H.
,
Lee
I.
,
Jo
A.
,
Choi
K.
 &
Zoh
K. D.
2020
Degradation of ciprofloxacin and inactivation of ciprofloxacin resistant E. faecium during UV-LED (275 nm)/chlorine process
.
Chemical Engineering Journal
394
,
124803
.
Google Scholar
Crossref
Search ADS
 
Li
G.
,
Zhou
S.
,
Shi
Z.
,
Meng
X.
,
Li
L.
 &
Liu
B.
2019
Electrochemical degradation of ciprofloxacin on BDD anode using a differential column batch reactor: mechanisms, kinetics and pathways
.
Environmental Science and Pollution Research
26
(
17
),
17740
–
17750
.
Google Scholar
Crossref
Search ADS
PubMed
 
Lima
V. B.
,
Goulart
L. A.
,
Rocha
R. S.
,
Steter
J. R.
 &
Lanza
M. R. V.
2020
Degradation of antibiotic ciprofloxacin by different AOP systems using electrochemically generated hydrogen peroxide
.
Chemosphere
247
.
Google Scholar
 
Martín de Vidales
M. J.
,
Rua
J.
,
de Juan
J. L. M.
,
Fernández-Martínez
F.
 &
dos Santos-García
A. J.
2020
Degradation of contaminants of emerging concern by electrochemical oxidation: coupling of ultraviolet and ultrasound radiations
.
Materials
13
(
23
),
1
–
12
.
Google Scholar
Crossref
Search ADS
 
Miwa
D. W.
,
Malpass
G. R. P.
,
Machado
S. A. S.
 &
Motheo
A. J.
2006
Electrochemical degradation of carbaryl on oxide electrodes
.
Water Research
40
(
17
),
3281
–
3289
.
Available from: https://www.sciencedirect.com/science/article/pii/S0043135406003988.
Google Scholar
Crossref
Search ADS
PubMed
 
Montenegro-Ayo
R.
,
Pérez
T.
,
Lanza
M. R. V.
,
Brillas
E.
,
Garcia-Segura
S.
 &
dos Santos
A. J.
2023
New electrochemical reactor design for emergent pollutants removal by electrochemical oxidation
.
Electrochimica Acta
458
,
142551
.
Available from: https://www.sciencedirect.com/science/article/pii/S0013468623007296.
Google Scholar
Crossref
Search ADS
 
Moreira
F. C.
,
Boaventura
R. A. R.
,
Brillas
E.
 &
Vilar
V. J. P.
2017
Electrochemical advanced oxidation processes: a review on their application to synthetic and real wastewaters
.
Applied Catalysis B: Environmental
202
,
217
–
261
.
Google Scholar
Crossref
Search ADS
 
Ojo
B. O.
,
Arotiba
O. A.
 &
Mabuba
N.
2022
Evaluation of FTO-BaTiO3/NiTiO3 electrode towards sonoelectrochemical degradation of emerging pharmaceutical contaminants in water
.
Colloids and Surfaces A: Physicochemical and Engineering Aspects
647
.
Google Scholar
 
Regalado-Méndez
A.
,
Ruiz
M.
,
Hernández-Servín
J. A.
,
Natividad
R.
,
Romero
R.
,
Cordero
M. E.
,
Estrada-Vázquez
C.
 &
Peralta-Reyes
E.
2020
Electrochemical mineralization of ibuprofen on BDD electrodes in an electrochemical flow reactor: numerical optimization approach
.
Processes
8
(
12
),
1666
.
Google Scholar
Crossref
Search ADS
 
Rivera-Utrilla
J.
,
Sánchez-Polo
M.
,
Ferro-García
M. Á.
,
Prados-Joya
G.
 &
Ocampo-Pérez
R.
2013
Pharmaceuticals as emerging contaminants and their removal from water. A review
.
Chemosphere
93
(
7
),
1268
–
1287
.
Google Scholar
Crossref
Search ADS
PubMed
 
Shin
Y.-U.
,
Yoo
H.-Y.
,
Ahn
Y.-Y.
,
Kim
M. S.
,
Lee
K.
,
Yu
S.
,
Lee
C.
,
Cho
K.
,
Kim
H.
 &
Lee
J.
2019
Electrochemical oxidation of organics in sulfate solutions on boron-doped diamond electrode: multiple pathways for sulfate radical generation
.
Applied Catalysis B: Environmental
254
,
156
–
165
.
Available from: https://www.sciencedirect.com/science/article/pii/S0926337319303820.
Google Scholar
Crossref
Search ADS
 
Sunderland
E. M.
,
Hu
X. C.
,
Dassuncao
C.
,
Tokranov
A. K.
,
Wagner
C. C.
 &
Allen
J. G.
2019
A review of the pathways of human exposure to poly- and perfluoroalkyl substances (PFASs) and present understanding of health effects
.
Journal of Exposure Science & Environmental Epidemiology
29
(
2
),
131
–
147
.
https://doi.org/10.1038/s41370-018-0094-1
.
Google Scholar
Crossref
Search ADS
 
Van Doorslaer
X.
,
Dewulf
J.
,
Van Langenhove
H.
 &
Demeestere
K.
2014
Fluoroquinolone antibiotics: an emerging class of environmental micropollutants
.
Science of the Total Environment
500–501
,
250
–
269
.
Available from: https://www.sciencedirect.com/science/article/pii/S0048969714012546.
Google Scholar
PubMed
 
Viana
D. F.
,
Salazar-Banda
G. R.
 &
Leite
M. S.
2018
Electrochemical degradation of Reactive Black 5 with surface response and artificial neural networks optimization models
.
Separation Science and Technology
53
(
16
),
2647
–
2661
.
https://doi.org/10.1080/01496395.2018.1463264
.
Google Scholar
Crossref
Search ADS
 
Wachter
N.
,
Aquino
J. M.
,
Denadai
M.
,
Barreiro
J. C.
,
Silva
A. J.
,
Cass
Q. B.
,
Rocha-Filho
R. C.
 &
Bocchi
N.
2019b
Optimization of the electrochemical degradation process of the antibiotic ciprofloxacin using a double-sided β-PbO2 anode in a flow reactor: kinetics, identification of oxidation intermediates and toxicity evaluation
.
Environmental Science and Pollution Research
26
(
5
),
4438
–
4449
.
https://doi.org/10.1007/s11356-018-2349-8
.
Google Scholar
Crossref
Search ADS
PubMed
 
© 2023 The Authors
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).