RuO2 anode

Science of the Total Environment 677 (2019) 84–97

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Electro-oxidation of amoxicillin trihydrate in continuous reactor by Ti/RuO2 anode Ravneet Kaur, Jai Prakash Kushwaha ⁎, Neetu Singh Chemical Engineering Department, Thapar Institute of Engineering and Technology, Patiala, Punjab, India

H I G H L I G H T S

G R A P H I C A L

A B S T R A C T

• Ti/RuO2 anode was used for amoxicillin (AMT) mineralization in continuous EO. • 37.82% TOC removal showed partial mineralization and AMT presence in degraded form. • Eight transformation products/reaction intermediates of AMT (ARIs) were determined. • Generated chlorinated ARIs underwent dechlorination via cathodic reduction. • A plausible AMT degradation mechanism was proposed.

a r t i c l e

i n f o

Article history: Received 28 March 2019 Received in revised form 19 April 2019 Accepted 23 April 2019 Available online 25 April 2019 Editor: Paola Verlicchi Keywords: Antibiotic degradation Ti/RuO2 Continuous electro-oxidation Transformation products Mineralization

a b s t r a c t Electro-oxidation (EO) of synthetic wastewater containing amoxicillin (AMT) antibiotic as a model pollutant was performed using dimensionally stable Ti/RuO2 electrodes in a continuous reactor set-up. Response surface methodology (RSM) was used for optimization of continuous EO process. Individual and interactive effects of initial pH of synthetic wastewater (2−10), applied current, I (0.25–1.25 A), elapsed time, t (20–180 min) and retention time, RT (15–195 min) on AMT removal, total organic carbon (TOC) removal and specific energy consumption (SEC, kWh (g TOC removed)−1) were investigated. At optimum conditions (pH = 7.53, I = 0.7 A, RT = 175.6 min, t = 128.89 min), 51.64% and 37.82% AMT and TOC removal was achieved, with SEC value of 0.408 kWh (g TOC removed)−1. AMT and TOC removal at optimum conditions was found to follow pseudo-first order kinetics. Mineralization current efficiency for optimum run of continuous EO came out to be 9.81%. Furthermore, 8 transformation products/reaction intermediates of AMT (ARIs) were determined by UPLC-Q-TOF-MS analysis, and subsequently, a plausible degradation scheme of AMT by anodic oxidation and cathodic reduction using Ti/RuO2 electrodes was proposed. © 2019 Elsevier B.V. All rights reserved.

1. Introduction

⁎ Corresponding author. E-mail addresses: [email protected] (J.P. Kushwaha), [email protected] (N. Singh).

https://doi.org/10.1016/j.scitotenv.2019.04.339 0048-9697/© 2019 Elsevier B.V. All rights reserved.

In past three decades, the production and usage of antibiotics have aggravated exponentially across the world, intended for treating bacterial infections in humans and animals (Boeckel et al., 2014). This has led to their presence in surface, ground and drinking water (Kummerer, 2001, 2009; Fick et al., 2009; Lin and Tsai, 2009), since majority of

R. Kaur et al. / Science of the Total Environment 677 (2019) 84–97

85

antibiotic content is left unmetabolized in human and animal body system and hence excreted in its actual form (Kemper, 2008; Rivera-Utrilla et al., 2013). Disposal of excess and expired drugs from hospitals, households, and untreated or partially treated effluent from manufacturing industries is another reason behind the serious problem of water contamination by antibiotics and other drugs (Kummerer, 2009; Homem and Santos, 2011). Moreover, conventional techniques such as biological treatment and physico-chemical methods used in wastewater treatment plants have been found to be inefficient in terms of antibiotics' removal because of their bio-recalcitrant nature. This has led to accumulation of antibiotics in different aquatic environments, with their concentration levels ranging from a few ng L−1 to 100 mg L−1 (Kummerer, 2001, 2009; Fick et al., 2009; Lin and Tsai, 2009; Mutiyar and Mittal, 2013; Petrie et al., 2015). Major harm associated with this issue is continuous development of antibiotic resistant bacteria, which is a serious threat to health of humans and animals (Khetan and Collins, 2007; Gadipelly et al., 2014). Moreover, humankind does not have much alternatives to available antibiotics yet. Amoxicillin trihydrate (AMT), a semi-synthetic β-lactam antibiotic, is widely prescribed for bacterial infections such as strep throat, pneumonia, urinary tract infections, skin infections and middle ear infections. Less than 30% of AMT is metabolised in liver and rest is defecated (Benito-Pena et al., 2006; Garcia-Reiriz et al., 2007). Electro-oxidation (EO) technique is a type of advanced oxidation process (AOP), which has gained much consideration of scientists, as it has shown promising results in terms of degradation and mineralization of a variety of organic pollutants (Anglada et al., 2009; Panizza and Cerisola, 2009; Rao and Venkatarangaiah, 2014; Yu et al., 2014; Nideesh et al., 2018). In EO, organics are destroyed through by direct and/or indirect oxidation. Hydroxyl radicals (•OH) and other reactive oxidant species (H2O2, O3 and HO2•generated at the anode surface (A) through water discharge oxidise the pollutants via direct oxidation at the anode surface. In case of non-active anodes (PbO2, BDD (boron doped diamond), Ti/SnO2), physisorption of hydroxyl radicals on to the anode surface is predominant (Eq. (1)), which leads to electrochemical combustion/ mineralization of organic pollutant (R) to CO2 and H2O (Eq. (2)) (Nideesh et al., 2019). However, in case of active anodes (Pt, Ti/IrO2, Ti/RuO2), chemisorbed hydroxyl radicals dominate on the anode surface (Eq. (3)), resulting in to electrochemical conversion of organics (Eq. (4)) (Comninellis, 1994; Chen, 2004; Martinez-Huitle and Ferro, 2006; Brillas and Martinez-Huitle, 2015).

Ganiyu et al., 2016; Frontistis et al., 2017). However, BDD anodes are comparatively much more expensive than other dimensionally stable anodes, thus their practical application at large scale is still dubious (Panizza and Cerisola, 2009). Other non-active anodes such as PbO2 and SnO2 have also proved their efficacy in antibiotic degradation (Wang et al., 2016; Yahiaoui et al., 2013). However, their poor performance in wastewaters comprising chlorides and inability to resist corrosion are major drawbacks (Cossu et al., 1998; Chen, 2004; Anglada et al., 2009). In such scenario, titanium coated with ruthenium dioxide (Ti/ RuO2) anodes have emerged as great alternative to other anode materials for electrochemical treatment of organic pollutants, as they are cost effective, dimensionally stable, highly chemical and mechanical resistant and possess high electro-catalytic activity for generation of strong oxidants (Cl−, •OH, H2O2, O3) (Kumar et al., 2015; Goyal et al., 2017; Kaur et al., 2018a, 20018b; Kaur et al., 2019). Ti/RuO2 electrode has shown promising results in terms of chemical oxygen demand (COD) removal and current efficiency for treatment of actual wastewaters from textile industry (Kaur et al., 2017; Kaur et al., 2018a). Goyal et al. (2017) reported successful treatment of highly acidic (pH b 2) and toxic wastewater with high COD value using Ti/RuO2 electrodes. Nonetheless, the studies on application of Ti/RuO2 anode for antibiotic degradation and mineralization are very scarce (Babu et al., 2009; Kaur et al., 2018b, 2019). Furthermore, there was no study in literature based on “continuous” reactor set-up for EO treatment of AMT. Batch electrolysis have been used for degradation studies which cannot be implemented on industrial scale where gigantic amounts of wastewater having high flow rate is generated on daily basis. Therefore, in view of research gap, in present study, treatment of wastewater comprising AMT antibiotic in a continuous EO reactor using Ti/RuO2 electrodes was explored. EO experiments were designed and optimized using central composite design (CCD). Effect of important operating parameters such as initial pH, applied current (I), retention time (RT), and elapsed time (t), on total organic carbon (TOC) removal, AMT removal and specific energy consumption (SEC) was studied. Decay kinetics of AMT and TOC removal were investigated at optimum conditions. In order to determine the AMT reaction intermediates (ARIs) formed during EO, UPLC-Q-TOF-MS analysis of samples extracted during optimum run was performed. Consequently, a tentative degradation pathway of AMT in a continuous EO reactor using Ti/RuO2 anodes was proposed.

A þ H2 O→Að˙OHÞ þ Hþ þ e− ðPhysisorptionÞ

ð1Þ

2.1. Chemicals

Að˙OHÞ þ R→nCO2 þ mHþ þ ze− þ Inorganic ions þ A ðElectrochemical combustionÞ

ð2Þ

A þ H2 O→AO þ 2Hþ þ 2e− ðChemisorptionÞ

ð3Þ

AO þ R→RO þ A→Reaction intermediates þ A ðElectrochemical conversionÞ

ð4Þ

Further, in case of active anodes, various pH dependent dominance active chlorine species (Cl2, HOCl, OCl−) are formed (Comninellis, 1994; Deborde and Von Gunten, 2008; Sires et al., 2014), and degrade the pollutants via indirect oxidation. There are quite a number of studies on abatement of different kinds of antibiotics by employing EO technique. Most researchers have reported successful degradation of antibiotics by EO using non-active BDD anodes (Li et al., 2008; Guinea et al., 2009; Gonzalez et al., 2011; Brinzilla et al., 2012; El-Ghenymy et al., 2013a, 2013b; Haidar et al., 2013; Fabianska et al., 2014; Korbahti and Tasyurek, 2015; Frontistis et al., 2017). Some scientists have also reported successful degradation and mineralization of AMT using BDD anode at current densities equal to or above 30 mA cm−2 (Panizza et al., 2014; Sopaj et al., 2015;

2. Experimental

The antibiotic, AMT (C16H19N3O5S.3H2O, with N95% purity), was procured from DSM Sinochem Pharmaceuticals (Punjab), India. Sodium chloride (NaCl, 99.5%), hydrochloric acid (HCl, 37%) and sodium hydroxide (NaOH, 98%) were of analytical grade and supplied by Loba Chemie Pvt. Ltd. (Mumbai), India. Both formic acid (CH2O2, N97.5%) and acetonitrile (C2H3N, 99.9%) were of LCMS grade, obtained from Fluka Sigma-Aldrich and Fluka Honeywell, respectively. 2.2. Continuous electro-oxidation experiments A cuboid shaped, open and undivided reactor (13 cm × 13 cm × 13 cm), made from acrylic sheet (Fig. 1), with inlet point at the bottom and outlet point at the top, was used to treat synthetic wastewater comprising AMT (50 mg L−1) and NaCl (2 g L−1) dissolved in double distilled water. Working volume of the reactor was 1.5 L. For each experiment, freshly prepared wastewater was continuously supplied from a reservoir to the reactor via peristaltic pump at a predetermined flow rate corresponding to required retention time (RT, min). Set of four rectangular Ti/RuO2 electrodes (10 cm × 8.5 cm; 1.5 cm thickness, Titanium Tantalum Products Ltd. Company (Chennai), India) connected in parallel and 1 cm apart from each other, with

86

R. Kaur et al. / Science of the Total Environment 677 (2019) 84–97

Fig. 1. Continuous EO reactor set-up.

alternate electrodes acting as anodes and the other two as cathodes, were immersed in to synthetic wastewater in the reactor. Electrodes were connected to a regulated direct current (DC) power supply (DIGITECH, Roorkee, India, Model: 4818A10; 0–5 A, 0–20 V), which performed current controlled electrolysis. Continuous mixing of wastewater solution was ensured by placing the reactor, having a magnetic bead at the bottom, on a magnetic stirrer. EO treatment was started by simply switching on DC power supply. The set of electrodes was thoroughly washed using distilled water before conducting each EO experiment. Experiments were designed based on five level central composite design (CCD), by response surface methodology (RSM) using DesignExpert software 6.0.8 (STAT-EASE Inc., Minneapolis, US). The range and levels assigned to independent variables is listed in Table 1. The design required 30 experiments with 6 replications at the center point to determine the effect of operating parameters or independent variables, i.e. pH, I (A), RT (min) and t (min), on TOC removal (X1), AMT removal (X2) and SEC (kWh (g TOC removed)−1) (X3) (Table 2). 2.3. Instruments and analytical procedures

X 2 ð%Þ ¼

C 0− C f  100 C0

ð6Þ

where, C0 is initial AMT concentration (mg L−1) and Cf is the residual AMT concentration (mg L−1) in wastewater sample after its continuous EO treatment. The economic viability of EO process was determined by computing SEC (X3) (Eq. (7)), which is defined as the amount of energy consumed in kWh for removal of 1 g of TOC from wastewater.
 −1 X 3 kWhðg TOC removedÞ ¼

UIΔt  TOC 0 −TOC f  V R

ð7Þ

Table 2 Design matrix and experimental responses.

The extent of mineralization of wastewater by EO process was analysed in terms of total organic carbon (TOC) removal. TOC was determined using TOC-VCPH, PC controlled TOC Analyzer (Shimadzu, Japan). TOC removal efficiency (X1) was computed using following equation (Eq. (5)): X 1 ð%Þ ¼

from calibration curve of AMT. AMT removal efficiency (X2) was then calculated using the following relationship (Eq. (6)):

TOC 0− TOC f  100 TOC 0

ð5Þ

where, TOC0 and TOCf denote the amount of TOC (mg L−1) in synthetic wastewater before and after continuous EO treatment, respectively. The concentration of residual AMT in samples of treated wastewater was determined using a double beam UV–visible spectrophotometer (Perkin Elmer, Lambda 35). Absorbance at λmax = 228 nm was analysed and concentration was calculated using a regression equation obtained

Table 1 Range and levels of operating parameters. Operating parameters

Levels (−2)

(−1)

(0)

(1)

(2)

pH: V1 Applied current, I (A): V2 Retention time, RT (min): V3 Elapsed time, t (min): V4

2 0.25 15 20

4 0.5 60 60

6 0.75 105 100

8 1.0 150 140

10 1.25 195 180

Std

Run

V1

V2

V3

V4

X1 (%)

X2 (%)

X3 (kWh/g TOC removed)

29 5 26 16 3 19 24 7 9 21 13 11 10 15 8 22 4 30 23 14 6 25 18 20 17 1 27 12 2 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

6 4 6 8 4 6 6 4 4 6 4 4 8 4 8 6 8 6 6 8 8 6 10 6 2 4 6 8 8 6

0.75 0.5 0.75 1 1 0.25 0.75 1 0.5 0.75 0.5 1 0.5 1 1 0.75 1 0.75 0.75 0.5 0.5 0.75 0.75 1.25 0.75 0.5 0.75 1 0.5 0.75

105 150 105 150 60 105 105 150 60 15 150 60 60 150 150 195 60 105 105 150 150 105 105 105 105 60 105 60 60 105

100 60 100 140 60 100 180 60 140 100 140 140 140 140 60 100 60 100 20 140 60 100 100 100 100 60 100 140 60 100

32.23 11.15 30.97 38.91 13.79 12.45 36.21 16.12 14.63 30.27 16.51 24.37 27.88 26.94 30.2 36.09 27.8 32.89 18.16 31.55 27.45 31.74 30.17 29.18 6.33 9.14 32.05 36.16 25.43 33.19

45.58 30.33 43.8 52.86 32.51 30.09 49.67 34.46 33.25 42.29 35.75 37.89 45.22 39.88 45.53 48.44 43.2 45.29 36.21 50.26 44.56 44.36 45.81 42.52 25.14 28.47 44.86 49.66 42.48 46.42

0.374 0.323 0.387 0.706 0.644 0.204 0.605 0.568 0.61 0.395 0.553 0.934 0.371 0.856 0.395 0.338 0.41 0.367 0.104 0.332 0.161 0.379 0.423 0.909 1.353 0.375 0.376 0.755 0.172 0.364

R. Kaur et al. / Science of the Total Environment 677 (2019) 84–97

where, X is the predicted response; βo is the constant coefficient; vi are the linear terms corresponding to independent variables' effects, v2i correspond to effect of squared terms; vivj are interaction terms and correspond to effect of each paired combination of variables; βi, βii and βij are the coefficients; and er represents the error value. Along with the abovementioned equation, experimental data were processed for ANOVA to analyze significance of model terms and the interaction between operating parameters and responses (Table 3). Model F values of 86.38, 53.12 and 500.19 for X1, X2 and X3, respectively, along with P value (Prob N F) b0.0001, signify that the models are significant for all three responses under consideration. Lack of fit F values of 4.12, 2.95 and 2.72 were insignificant for TOC removal, AMT removal and SEC, respectively, which is desired for the model. For a model term, if P value b 0.05, then it is significant. Therefore, based on this confidence level and data given in Table 3, the response values were calculated in terms of significant model terms (Eqs. (11)–(13)).

where, U is the average cell potential (V), Δt is the electrolysis time (h), and VR is the reactor volume (L). The mineralization current efficiency (MCE) of continuous EO process was calculated for optimum run using following equation:  MCEð%Þ ¼

 TOC 0 −TOC f zFV R 4:32  107 nIΔt

 100

ð8Þ

where, z is the number of electrons involved in AMT mineralization to CO2 (Eq. (9)) (z = 70), F is Faraday's constant (F = 96,485C mol−1), n denotes number of carbon atoms in AMT molecule (n = 16), VR is the volume of synthetic wastewater, Δt is the electrolysis time (h), and 4.32 × 107 is the conversion factor (3600 s h−1 × 12,000 mg mol−1). C16 H19 N3 O5 S  3H2 O þ 28H2 O→16CO2 þ 3NH4 þ þ SO4 2− þ 69Hþ þ 70e−

ð9Þ


 X1 ¼ −77:81 þ ð15:77  pHÞ þ ð80:43  IÞ þ 6:43  10−3  RT

 þ ð0:21  t Þ− 0:92  pH2 − 48:73  I2
 ð11Þ þ −9:08  10−4  t2 þ ð0:13  I  tÞ

The initial pH of synthetic wastewater was set using required amounts of 0.1 M HCl or 0.1 M NaOH solutions. Thermo Scientific Orion 5 star pH meter was used to check the pH of solution. AMT reaction intermediates (ARIs) generated during the continuous EO treatment of wastewater by Ti/RuO2 electrodes were identified using ultra performance liquid chromatography coupled with quadrupole time-of-flight mass-spectrometry (UPLC-XEVO-G2-XS/QTOF-MS, Waters), fitted with BEH (Ethylene bridged hybrid) C18 analytical column (100 mm × 2.1 mm i.d., 1.7 μm particle size). The method conditions for UPLC system, as well as the details and operating conditions of mass spectrometer are reported in our previous works (Kaur et al., 2019).


 X2 ¼ −30:46 þ ð10:32  pHÞ þ ð63:02  IÞ− 9:19  10−3  RT

 ð12Þ þ ð0:08  t Þ− 0:58  pH2 – 33:99  I2
 X3 ¼ 1:41−ð0:39  pHÞ−ð0:55  IÞ− 3:34  10−4  RT 


þ 2:56  10−3  t þ 0:03  pH2 þ 0:61  I2

 þ 1:18  10−4  pH  RT þ 2:69  10−3  I  t

3. Results and discussion

The correlation of responses and independent variables was estimated using quadratic model (which also includes linear model), given by second order polynomial equation (Eq. (10)) as mentioned below: k X

βi vi þ

k X

i¼1

βii v2i þ

i¼1

XX

βij vi v j þ er

ð13Þ

The adequate precision (signal to noise ratio) values were 36.05, 29.35 and 85.13 for X1, X2 and X3, respectively, which signify that the models have a very strong signal to be used for optimization. The quality of fit of the quadratic model was justified on the basis of R2 (coefficient of determination) values: 0.988 (X1), 0.98 (X2) and 0.998 (X3). Moreover, the values of predicted R2 (X1: 0.935, X2: 0.898, X3: 0.989) were in agreement with the values of adjusted R2 (X1: 0.976, X2: 0.962, X3: 0.996). The CV values for all the responses were in acceptable range (X1: 5.47, X2: 3.41, X3: 3.17), thus implying enhanced reproducibility. Furthermore, the models were validated using plots between

3.1. Model fitting and analysis of variance (ANOVA)

X ¼ β0 þ

87

ð10Þ

ib j

Table 3 ANOVA results of quadratic models of responses for EO treatment of AMT wastewater. Source

Model pH I RT t pH2 I2 R2T t2 pH × I pH × RT pH × t I × RT I×t RT × t Residual Lack of fit Pure error Cor total

X1 (%TOC removal)

X2 (%AMT removal)

X3 (SEC (kWh/g TOC removed))

Sum of squares

DF

Mean square

F– value

Prob N F

Sum of squares

DF

Mean square

F– value

Prob N F

Sum of squares

DF

Mean square

F– value

Prob N F

2384.95 1072.14 294.07 40.74 352.43 372.9 254.47 0.056 57.94 5.09 0.26 4.65 0.013 27.74 0.28 29.58 26.38 3.2 2414.53

14 1 1 1 1 1 1 1 1 1 1 1 1 1 1 15 10 5 29

170.35 1072.14 294.07 40.74 352.43 372.9 254.47 0.056 57.94 5.096 0.26 4.65 0.014 27.74 0.28 1.97 2.64 0.64

86.38 543.64 149.11 20.659 178.71 189.08 129.03 0.028 29.38 2.58 0.13 2.36 0.007 14.07 0.14

b0.0001 b0.0001 b0.0001 0.0004 b0.0001 b0.0001 b0.0001 0.8681 b0.0001 0.1288 0.7202 0.1453 0.9344 0.0019 0.7125

104.75 846.92 106.38 46.06 205.04 149.16 123.79 0.54 5.95 4.21 1.18 0.09 0.25 2.21 1.27 1.97 2.53 0.86

b0.0001 b0.0001 b0.0001 0.0002 b0.0001 b0.0001 b0.0001 0.61 0.1 0.16 0.45 0.83 0.72 0.31 0.43 0.12

14 1 1 1 1 1 1 1 1 1 1 1 1 1 1 15 10 5 29

0.12 0.27 0.59 0.009 0.42 0.31 0.04 0.0001 0.0009 0.0002 0.002 0.0002 0.0003 0.01 0.0005 0.0002 0.0003 0.0001

b0.0001 b0.0001 b0.0001 b0.0001 b0.0001 b0.0001 b0.0001 0.5102 0.0698 0.4287 0.0144 0.4287 0.2726 b0.0001 0.1638

2.95

1.65 0.27 0.59 0.009 0.42 0.31 0.04 0.0001 0.0009 0.0002 0.002 0.0002 0.0003 0.01 0.0005 0.003 0.003 0.0005 1.66

500.19 1129.57 2508.16 38.98 1773.33 1324.31 167.93 0.45 3.81 0.66 7.65 0.66 1.29 48.94 2.14

0.06

14 1 1 1 1 1 1 1 1 1 1 1 1 1 1 15 10 5 29

53.12 429.46 53.95 23.36 103.97 75.64 62.77 0.27 3.02 2.14 0.59 0.05 0.13 1.12 0.64

4.12

1466.5 846.92 106.38 46.06 205.04 149.16 123.79 0.54 5.95 4.21 1.18 0.09 0.25 2.21 1.27 29.58 25.29 4.29 1496.08

2.72

0.1405

88

R. Kaur et al. / Science of the Total Environment 677 (2019) 84–97

Fig. 2. Normal % probability vs. studentized residuals plots for (a) %TOC removal, (b) %AMT removal, and (c) SEC (kWh/g TOC removed).

(a) normal % probability vs studentized residuals (Fig. 2), and (b) predicted vs actual values (Fig. 3), for responses X1, X2 and X3. The adequacy and accuracy of models was thus verified.

3.2. Effect of operating parameters and optimization The three dimensional response surface plots (Figs. 4–6) were used to analyze the individual and interactive effects of independent variables (pH, I (A), RT (min), t (min)) on responses: TOC removal (X1), AMT removal (X2) and SEC (kWh (g TOC removed)−1) (X3).

As shown in Fig. 4a, the TOC removal efficiency was very poor in highly acidic medium, i.e. pH 2–4. With further increase in pH, X1 was found to increase up to pH value of ≈7. Thereafter, in alkaline environment (pH 8–10), the TOC removal efficiency slightly decreased. For all the values of I (A), maximum X1 was attained in the neutral pH range only. X1 was extremely poor at low applied current of 0.25 A (current density, i = 1.47 mA cm−2). The interactive effect of applied current and elapsed reaction time is shown in Fig. 4b. TOC removal efficiency increased with increase in EO reaction time. However, the enhancement in X1 value with time was more prominent for higher values of applied current (0.5–1.25 A), than low applied current value ≈0.25 A. Retention

R. Kaur et al. / Science of the Total Environment 677 (2019) 84–97

89

Fig. 3. Predicted vs. actual plot for (a) %TOC removal, (b) %AMT removal, and (c) SEC (kWh/g TOC removed).

time is an important parameter in continuous EO treatment process. As shown in Fig. 4c, X1 was found to increase with both t and RT. Nonetheless, it is clear that the effect if t on X1 was slightly more prominent than the effect of RT. The interactive effects of pH – RT and I – RT on TOC removal is shown in Fig. 4d and f, respectively. As discussed previously, X1 increased with increase in RT, over entire range of applied I and initial pH of wastewater. Fig. 4e depicts the interactive effect of elapsed time and initial pH on X1. Maximum TOC removal efficiency was achieved in neutral pH range for elapsed time ≈ 150 min. Fig. 5a depicts the interactive effect of I and pH on AMT removal efficiency (X2). Substantial improvement in X2 was seen when initial pH was increased from 2 to ≈7, for all values of I. However, further increase in pH up to 10 resulted in almost no change in X2. Similarly, over entire

range of elapsed time, t, maximum AMT removal was attained at pH ≈ 7 (Fig. 5e). The effect of I on X2 followed the same trend as X1. Hence, maximum AMT removal was achieved in neutral pH range at applied current ≈ 1.0 A (Fig. 5b). Increment in X2 value followed a linear trend when t value was increased from 20 to 180 min, for all values of I. X2 value improved when both t and RT were increased from 20 to 180 min, and 15 to 190 min, respectively (Fig. 5c). X2 always increased with increase in RT irrespective of solution pH and applied current (Fig. 5d, f). EO technique involves generation of a number of oxidants on anode surface for destruction of target pollutant either on the anode surface or in the bulk solution. The pH of wastewater plays a key role in generation of different types of oxidant species, particularly in chloride rich

90

R. Kaur et al. / Science of the Total Environment 677 (2019) 84–97

Fig. 4. 3D response surface graphs for %TOC removal as a function of independent variables pH, I, t and RT.

wastewaters. Moreover, Ti/RuO2 anode is known for its strong electrocatalytic activity for generation of chloro-oxidant species (Cl2, HOCl·OCl−), (Comninellis, 1994; Deborde and Von Gunten, 2008; Sires et al., 2014). In present study, best results for X1 and X2 were attained in the neutral pH range. Nevertheless, efficiency in terms of AMT and TOC removal was quite low in acidic pH range. Degradation of AMT antibiotic in acidic medium primarily occurred through chemisorbed hydroxyl radicals (Eq. (3)) and active chlorine in the form of Cl2. Oxidation potential of Cl2 is comparatively lower than other oxidants. Moreover, chemisorbed •OH results in partial conversion of AMT to its intermediates (ARIs) (Eq. (4)). Furthermore, chlorine reduction takes place in acidic conditions according to Eq. (14), and oxygen gets evolved (Trasatti, 1987). 2Cl2 þ 2H2 O→O2 þ 4HCl

ð14Þ

Thus, indirect oxidation of AMT by Cl2 and partial conversion by chemisorbed hydroxyl radicals resulted in to poor values of X1 and X2 in acidic conditions. Physisorbed hydroxyl radicals are predominant in neutral pH range, thus resulting into maximum AMT removal and TOC removal according to Eq. (2). Furthermore, HOCl is dominant in neutral pH, whose oxidation potential is higher than that of Cl2, thus resulting in improved values of X1 and X2. Alkaline pH (8–10) had a meagre undesirable impact on X1, but X2 remained unaffected. This could be attributed to lower oxidation potential of species, such as hydrogen peroxide and hydroperoxyl radical (Eqs. (15) and (16)), responsible for AMT abatement in alkaline pH conditions (Martinez-Huitle and

Ferro, 2006; Yavuz and Koparal, 2006; Brillas and Martinez-Huitle, 2015). 2˙OH→H2 O2

ð15Þ

˙OH þ H2 O2 →HO2 ˙ þ H2 O

ð16Þ

The amount of current applied during EO has a strong influence on different outcomes of this process. Both X1 and X2 were found to increase with increase in I value from 0.25 to ≈1.0 A. The results are justified, since the rate of generation of oxidant species on the anode surface is directly proportional to the amount of current supplied to the EO reactor system, thus giving enhanced values of %AMT and % TOC removal at higher values of applied current. Nonetheless, this was found to be true up to a certain level of applied current. When I was further raised up to 1.25 A, X1 and X2 tend to decrease. During EO process, there are two inevitable reactions of oxygen and hydrogen evolution taking place on anode and cathode, respectively (Eqs. (17) and (18)). H2 O→1=2O2 þ 2Hþ þ 2e−

ð17Þ

2H2 O þ 2e− →H2 þ 2OH−

ð18Þ

Furthermore, with increase in applied current, the generation rate of •OH increases, and excess hydroxyl radicals react to form hydrogen peroxide and hydroperoxyl radical according to reactions mentioned above (Eqs. (15) and (16)) (Brillas and Martinez-Huitle, 2015; Moreira et al.,

R. Kaur et al. / Science of the Total Environment 677 (2019) 84–97

91

Fig. 5. 3D response surface graphs for %AMT removal as function of independent variables pH, I, t and RT.

2017). Consequently, at higher applied current, its significant portion got exhausted in these undesirable but inevitable side reactions, thus resulting in diminutive performance of EO in terms of AMT and TOC removal. EO is an energy intensive process. In order to check its feasibility on economic grounds, the third response, specific energy consumption (X3) was evaluated. Fig. 6 shows the individual and interactive effects of independent variables on X3. SEC increased with the increase in value of I, for all values of pH and RT (Fig. 6a, f). The average cell potential increased with the increase in applied current, thus resulting in higher values of SEC. Minimum SEC was found in neutral pH range (6–8). This is because maximum mineralization took place in this pH range. Maximum energy was consumed in acidic conditions, again due to poor mineralization, because SEC is inversely proportional to the amount TOC removed from the wastewater (Eq. (7)). This signifies that most part of the energy supplied during EO was consumed in parasitic reactions. As shown in Fig. 6b, there is no visible effect on SEC with elapsed time (20–180 min) for lower values of applied current (0.25 to ≈0.75 A). However, with further increase in I value up to 1.25 A, X3 value always increased. The increment in X3 was however quite sharp towards higher values of elapsed time, t. SEC always increased with increase in elapsed time, but decreased with the increase in retention time (Fig. 6c, f). Fig. 6d depicts the interactive effect of pH and RT on X3. For acidic pH range from 2 to ≈5, SEC decreased with the increase in value of RT. However, when the pH approaches neutral zone, the effect of RT on SEC diminishes. In alkaline pH zone, SEC remained almost unaffected by RT. Nevertheless, SEC always increased with increase in elapsed time for all values of pH (Fig. 6e).

The operating conditions for continuous EO treatment were optimized in order to achieve maximum %AMT and %TOC removal, and minimum SEC (kWh (g TOC removed)−1). The optimum condition achieved by RSM for given target was: pH = 7.53, I = 0.7 A, RT = 175.6 min, and t = 128.89 min, and the responses predicted was: X 1 = 38.36%, X 2 = 52.86% and X 3 = 0.385 kWh (g TOC removed)−1). In order to verify the suitability of this optimization, a set of three experiments was performed at the given optimum conditions. The mean values of responses thus obtained were: X1 = 37.82%, X2 = 51.64% and X3 = 0.408 kWh (g TOC removed)−1) in good agreement with their predicted values. The computed value of MCE at optimum conditions came out to be 9.81% (Eq. (8)), implying that this much part of total current supplied to the EO cell was utilized in mineralization of AMT, and the rest was consumed in side reactions occurring during the treatment process. Other authors have also reported the EO of various other antibiotics (Table 4). It can be seen that BDD, Ti/SnO 2 and PbO 2 based electrodes performed well. However, the reactor used was batch type. As discussed in introduction section, BDD anodes are comparatively expensive and large scale application is dubious. PbO 2 and SnO 2 electrodes showed poor performance in wastewaters comprising chlorides and inability to resist corrosion. 3.3. Kinetics of treatment A continuous EO experiment was run at optimum conditions (pH = 7.53, I = 0.7 A, RT = 175.6 min, t = 128.89 min), and 10 mL samples of synthetic wastewater were collected from the reactor outlet after fixed intervals of time, for studying the AMT and TOC decay kinetics using

92

R. Kaur et al. / Science of the Total Environment 677 (2019) 84–97

Fig. 6. 3D response surface graphs for SEC (kWh/g TOC removed) as function of independent variables pH, I, t and RT.

Table 4 Reported studies on antibiotics with EO method. Antibiotic

Concentration/reactor −1

Electrode

Comments

Reference

-99% antibiotic removal -93% COD removal -87% mineralization -95.6% mineralization

Brinzilla et al. (2012)

Tetracycline

150 mg L batch reactor

BDD

Sulfamethoxazole

100 mg L−1 batch reactor 618 mg L−1 batch reactor 50 mg L−1

Ti/SnO2-Sb/Ce-PbO2

Ampicillin Ciprofloxacin

Nb/BDD Ti/SnO2-Sb

Ofloxacin

50 mg L−1 batch reactor

Ti/RuO2

Tetracycline

50 mg L−1 batch reactor 50 mg L−1 flow reactor 50 mg L−1 batch reactor

Nb/BDD Ti-Pt/β-PbO2

100 mg L−1 batch reactor

SnO2-Sb BDD

Ciprofloxacin Amoxicillin trihydrate

Norfloxacin

Ti/RuO2

-97.1% antibiotic removal -92.5% COD removal -99.5% antibiotic removal -86% COD removal -70% mineralization -≈80% antibiotic removal within 30 min -46.3% mineralization in 4 h -Mineralization current efficiency = 4.9% -80% mineralization in 2 h -Mineralization current efficiency = 7.87% -100% antibiotic removal within 2 h -≈75% mineralization after 5 h -≈60% antibiotic removal within 1 h -48 % mineralization in 4 h -Mineralization current efficiency = 7.67% Sb-SnO2: -63% mineralization -9.5% mineralization current efficiency BDD: -92% mineralization -3.9% mineralization current efficiency

Lin et al. (2013) Korbahti and Tasyurek (2015) Wang et al. (2016)

Kaur et al. (2018b)

Vasilie et al. (2018) Wachter et al. (2019) Kaur et al. (2019)

Mora-Gomez et al. (2019)

R. Kaur et al. / Science of the Total Environment 677 (2019) 84–97

93

where, CRt (mg L−1) and TOCRt (mg L−1) correspond to the amounts of AMT and TOC removed from the wastewater at any time t (min), respectively, CRe (mg L−1) and TOCRe (mg L−1) denote the amounts of AMT and TOC removed at equilibrium conditions, and kf (min−1) symbolizes pseudo-first order rate constant. Marquardt's percent standard deviation (MPSD) (Eq. (21)) was employed as an error function to fit the experimental data using non – linear regression (Marquardt, 1963). vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi u  d  u 1 X C Rt;i; exp −C Rt;i;cal 2 MPSD ¼ 100t ðd−pÞ i¼1 C Rt;i; exp

Fig. 7. Kinetics of AMT and TOC removal at optimum condition.

modified equations (Eqs. (19) and (20)) of pseudo-first order kinetic model (Lucas and Peres, 2009). h i C Rt ¼ C Re 1−ek f t

ð19Þ

h i TOC Rt ¼ TOC Re 1−ek f t

ð20Þ

where, the subscripts ‘exp’ and ‘cal’ denote the experimental and calculated data, d and p stand for number of data points and number of parameters in the model, respectively. CRt values in Eq. (21) were replaced by TOCRt values to compute the error function corresponding to TOC removal kinetics. The AMT and TOC removal data obtained at optimum conditions for continuous EO was fitted on to pseudo-first order kinetic model according to Eqs. (19) and (20), and is demonstrated in Fig. 7. R2 value was found to be 0.998 for both AMT and TOC removal, which substantiates the fit of pseudo-first order kinetic model. Moreover, MPSD values were as low as 6.87 and 9.84 for AMT and TOC removal, respectively, adding to the validation of the kinetic model. The values of kf were found to be 4.4 × 10−2 (AMT removal) and 9.1 × 10−3 (TOC removal). This signify that AMT removal from the synthetic wastewater was

Table 5 AMT reaction intermediates (ARIs) formed during continuous EO identified by UPLC-Q-TOF-MS. Rt (min)

m/z [M + H]+

Elemental composition

ARI 1

1.30

368

C15H17N3O6S

ARI 2

4.70

453

C14H18N3O12S

ARI 3

0.72

339

C15H15ClN2O3S

ARI 4

0.86

412

C15H15Cl2N3O6S

ARI 5

5.08

340

C14H17N3O5S

ARI 6

0.77

269

C8H10Cl2N2O2S

ARI 7

1.45

300

C7H13N3O8S

ARI 8

2.88

334

C12H16ClN3O4S

Compound

ð21Þ

Proposed Structure

94

R. Kaur et al. / Science of the Total Environment 677 (2019) 84–97

almost five times faster than TOC removal using Ti/RuO2 electrodes in continuous EO method. As discussed in Section 3.2, EO is a complex process involving a number of oxidants and their reactions with organic pollutants. Ti/RuO2 anodes majorly demonstrate active behavior resulting in to oxidation or electrochemical conversion of pollutant species. The reaction intermediates thus formed further undergo oxidation and break into organic entities with lower molecular weights. The process continues till their mineralization, i.e. conversion to CO2. Hence, TOC removal process was found to be slower as compared to degradation of target pollutant. 3.4. Identification of amoxicillin reaction intermediates, mechanism and pathway In order to degrade and mineralize the target pollutant, there are a number of reactions taking place at the anode and in the bulk solution during EO process, resulting in to formation of different reaction intermediates. Degradation of AMT by Ti/RuO2 electrodes also resulted in to formation of certain reaction intermediates (ARIs), which were identified by subjecting the samples collected during optimum run to UPLCQ-TOF-MS analysis (Table 5). Subsequently, a plausible AMT degradation pathway was proposed (Fig. 8). The retention time of AMT was 0.93 min (Fig. 9a). The corresponding mass spectra of initial sample yielded signals at mass/charge ratio (m/z) values of 366 and 388, corresponding to protonated AMT molecule [M + H]+ and its sodium adduct [M + Na]+, respectively. Signals at m/z values of 349, 208 and 160 corresponding to three AMT fragments [M − NH3 + H]+, [M − C6H10N2O3 + H]+ and [M − C10H10N2O3 + H]+, respectively, were also detected. These fragments were formed as a consequence of loss of amine group and opening of β-lactam ring of AMT molecule. The UPLC chromatograms of samples extracted from the reactor after 15 and 30 min (Fig. 9b, c) of electrolysis revealed the presence of 8 ARIs (ARI 1 – ARI 8). The details of retention time, m/z ratio,

elemental composition and proposed structure of these ARIs are presented in Table 5. In our previous work (Kaur et al., 2019), degradation study of AMT in batch electrolysis using Ti/RuO2 revealed the presence of three major reaction intermediates. However, in present study, out of 8 ARIs detected during continuous EO of AMT, two were found to be similar as reported earlier (corresponding to m/z ratio = 339 and 269), and 6 of them were detected only during continuous EO treatment of AMT. Based on the identified ARIs, a tentative degradation scheme of AMT antibiotic was derived and is shown in Fig. 8.The degradation of AMT on Ti/RuO2 electrodes followed different routes such as direct anodic oxidation via hydroxylation and decarboxylation, and opening of β-lactam ring via indirect oxidation. 3.4.1. Direct anodic oxidation In direct anodic degradation, electron transfer from the Ti/RuO2 anode to the AMT molecule adsorbed on to its surface took place, leading to its degradation/mineralization according to Eq. (22). Ti=RuO2 þ AMT→Ti=RuO2  ðAMTÞads →nCO2 þ other degradation products þ zHþ

ð22Þ

Apart from direct electron transfer to AMT molecule, its degradation occurred through highly reactive hydroxyl radicals. Direct oxidation via hydroxylation of AMT led to its de-alkylation and formation of ARI 1 (m/ z 368). Subsequent hydroxylation and decarboxylation of ARI 1 generated ARI 5 (m/z 340). ARI 2 (m/z 453) was formed by attack of multiple hydroxyl radicals on different bonds of AMT molecule resulting in to its demethylation. 3.4.2. Indirect oxidation via opening of β-lactam ring In another route of degradation pathway, attack of hydroxyl radicals and chloro-oxidant species (Cl2, HOCl·OCl−) generated via Eqs. (23)– (26) (Comninellis, 1994; Deborde and Von Gunten, 2008; Sires et al.,

Fig. 8. Proposed degradation pathway for AMT antibiotic by continuous EO treatment using Ti/RuO2 electrodes.

R. Kaur et al. / Science of the Total Environment 677 (2019) 84–97

2014), on AMT molecule resulted in its β-lactam ring opening and decarboxylation, respectively, and thus forming ARI 4 (m/z 412). Subsequent hydroxylation and replacement of entire aromatic moiety of ARI 4 led to ARI 7 (m/z 300). Furthermore, the instability of ARI 4 molecule and action of active oxygen yielded ARI 8 (m/z 334). Alternatively, combined attack of active oxygen and active chlorine species on ARI 5 also resulted in to formation of ARI 8. Loss of amine (-NH3) group and decarboxylation by active chlorine from AMT molecule consequently formed ARI 3 (m/z 339). Replacement of entire aromatic moiety by Cl− in ARI 3 resulted in generation of ARI 6 (m/z 269). Further attack of hydroxyl radicals, would lead to mineralization of the reaction intermediates + (ARI 1–8) to CO2 and inorganic ions such as SO2− 4 and NH4 .

95

3.4.3. Direct cathodic reduction The chlorinated ARIs (R–Cl) (ARI 3, 4, 6 and 8) would further undergo cathodic degradation, through hydrogen generated due to water dissociation (Eq. (22)) and chemisorbed on to cathode surface, leading to their dechlorination (Eq. (28)). The set of reactions taking place on Ti/RuO2 cathodes involving adsorbed chlorinated reaction intermediates (Eq. (27)) and adsorbed hydrogen atom are shown below: Ti=RuO2 þ H2 O þ e− →Ti=RuO2  ðHÞad þ OH–

ð27Þ

Ti=RuO2 þ R–Cl→Ti=RuO2  ðR–ClÞad

ð28Þ

A þ Cl →Að˙Clad Þ þ e−

ð23Þ

Ti=RuO2  ðR–ClÞad þ Ti=RuO2  ðHÞad →Ti=RuO2  ðR–HÞad þ HCl

ð29Þ

2Að˙Clad Þ→2A þ Cl2 ðpHb1Þ

ð24Þ

Ti=RuO2  ðR–HÞad →R–H þ Ti=RuO2

ð30Þ

Cl2 þ H2 O→HOCl þ Hþ þ Cl ð3bpHb8Þ

ð25Þ

−

ð26Þ

The reduced entities would further undergo either direct anodic oxidation or indirect oxidation in the bulk via different chloro-oxidant species, further lowering the amount of TOC in the wastewater.

−

−

HOCl þ OH− →ClO þ H2 O ðpHN8Þ

Fig. 9. UPLC-Q-TOF-MS chromatogram (a) AMT (Retention time = 0.93 min) (b) wastewater sample after 15 min of EO (c) wastewater sample after 30 min of EO.

96

R. Kaur et al. / Science of the Total Environment 677 (2019) 84–97

Possible degradation routes of AMT by EO using non-active anodes such as Ti4O7 and BDD were explored by Ganiyu et al., 2016 and Frontistis et al., 2017. In both the cases, hydroxylation via physisorbed •OH, was clearly the prime force behind degradation and mineralization of AMT antibiotic. However, in present work active chlorine and hydroxyl radicals, both play their parts in AMT abatement process. Degradation by EO thus follows different routes depending on the type of electrode employed for the purpose. 4. Conclusions In this work, synthetic wastewater containing amoxicillin antibiotic as a model pollutant was treated by continuous EO method using dimensionally stable Ti/RuO2 electrodes, in order to optimize the process parameters and identify major reaction intermediates and degradation pathway, not reported earlier. At optimum conditions (pH = 7.53, I = 0.7 A, RT = 175.6 min, t = 128.89 min), the response values came out to be: AMT removal = 51.64%, TOC removal = 37.82%, and SEC = 0.408 kWh (g TOC removed)−1. The experimental values of optimum conditions were in agreement with the values predicted by RSM. AMT and TOC removal at optimum conditions was found to follow pseudofirst order kinetics, with kf values of 4.4 × 10−2 (AMT removal) and 9.1 × 10−3 (TOC removal). This signifies that AMT removal from the synthetic wastewater was almost five times faster than TOC removal using Ti/RuO2 electrodes in continuous EO method. Mineralization current efficiency at optimum conditions came out to be 9.81%. Furthermore, 8 major transformation products/reaction intermediates of AMT (ARIs) were determined by UPLC-Q-TOF-MS analysis, and subsequently, a plausible degradation scheme of AMT by anodic oxidation and cathodic reduction using Ti/RuO2 electrodes was proposed. List of abbreviations

AMT amoxicillin trihydrate ANOVA analysis of variance AOP advanced oxidation process ARI amoxicillin reaction intermediate BDD boron doped diamond CCD central composite design COD chemical oxygen demand EO electro-oxidation RSM response surface methodology SEC specific energy consumption MCE mineralization current efficiency TOC total organic carbon UPLC-Q-TOF-MS ultra-performance-quadrupole-time-of-flight-massSpectrometry Acknowledgement Authors are much grateful to University Grants Commission (UGC), India for providing research fellowship (MANF-2014-15-SIK-PUN43596) to the first author. References Anglada, A., Urtiaga, A., Ortiz, I., 2009. Contributions of electrochemical oxidation to waste-water treatment: fundamentals and review of applications. J. Chem. Technol. Biotechnol. 84, 1747–1755. Babu, B.R., Venkatesan, P., Kanimozhi, R., Basha, C.A., 2009. Removal of pharmaceuticals from wastewater by electrochemical oxidation using cylindrical flow reactor and optimization of treatment conditions. Journal of Environmental Science and Health Part A 44, 985–994. Benito-Pena, E., Partal-Rodera, A.I., Leon-Gonzalez, M.E., Moreno-Bondi, M.C., 2006. Evaluation of mixed mode solid phase extraction cartridges for the pre-concentration of

beta-lactam antibiotics in wastewater using liquid chromatography with UV-DAD detection. Anal. Chim. Acta 556, 415–422. Boeckel, T.P.V., Gandra, S., Ashok, A., Caudron, Q., Grenfell, B.T., Levin, S.A., Laxminarayan, R., 2014. Global antibiotic consumption 2000 to 2010: an analysis of national pharmaceutical sales data. Lancet Infect. Dis. 14, 742–750. Brillas, E., Martinez-Huitle, C.A., 2015. Decontamination of wastewaters containing synthetic organic dyes by electrochemical methods. An update. Rev. Applied Catalysis B: Environmental 166–167, 603–643. Brinzilla, C.I., Pacheco, M.J., Ciriaco, L., Ciobanu, R.C., Lopes, A., 2012. Electrodegradation of tetracycline on BDD anode. Chem. Eng. J. 209, 54–61. Chen, G., 2004. Electrochemical technologies in wastewater treatment. Sep. Purif. Technol. 38, 11–41. Comninellis, C., 1994. Electrocatalysis in the electrochemical conversion of organic pollutants for wastewater treatment. Electrochim. Acta 39, 1857–1862. Cossu, R., Polcaro, A.M., Lavagnolo, M.C., Mascia, M., Palmas, S., Renoldi, F., 1998. Electrochemical treatment of landfill leachate: oxidation at Ti/PbO2 and Ti/SnO2 anodes. Environ. Sci. Technol. 32, 3570–3573. Deborde, M., Von Gunten, U., 2008. Reactions of chlorine with inorganic and organic compounds during water treatment-kinetics and mechanisms: a critical review. Water Res. 42, 13–51. El-Ghenymy, A., Cabot, P.L., Centellas, F., Garrido, J.A., Rodriguez, R.M., Arias, C., Brillas, E., 2013a. Electrochemical incineration of the antimicrobial sulfamethazine at a borondoped diamond anode. Electrochim. Acta 90, 254–264. El-Ghenymy, A., Garrido, J.A., Rodriguez, R.M., Cabot, P.L., Centellas, F., Arias, C., Brillas, E., 2013b. Degradation of sulfanilamide in acidic medium by anodic oxidation with a boron-doped diamond anode. J. Electroanal. Chem. 689, 149–157. Fabianska, A., Bialk-Bielinska, A., Stepnowski, P., Stolte, S., Siedlecka, E.M., 2014. Electrochemical degradation of sulfonamides at BDD electrode: kinetics, reaction pathway and eco-toxicity evaluation. J. Hazard. Mater. 280, 579–587. Fick, J., Soderstrom, H., Lindberg, R.H., Phan, C., Tysklind, M., Larsonn, D.G.J., 2009. Contamination of surface, ground, and drinking water from pharmaceutical production. Environ. Toxicol. Chem. 28, 2522–2527. Frontistis, Z., Antonopoulou, M., Venieri, D., Konstantinou, I., Mantzavinos, D., 2017. Boron-doped diamond oxidation of amoxicillin pharmaceutical formulation: statistical evaluation of operating parameters, reaction pathways and antibacterial activity. J. Environ. Manag. 195, 100–109. Gadipelly, C., Perez-Gonzalez, A., Yadav, G.D., Ortiz, I., Ibanez, R., Rathod, V.K., Marathe, K.V., 2014. Pharmaceutical industry wastewater: review of the technologies for water treatment and reuse. Ind. Eng. Chem. Res. 53, 11571–11592. Ganiyu, S.O., Oturan, N., Raffy, S., Cretin, M., Esmilaire, R., Hullebusch, E.V., Esposito, G., Oturan, M.A., 2016. Sub-stoichiometric titanium oxide (Ti 4 O 7 ) as a suitable ceramic anode for electrooxidation of organic pollutants: a case study of kinetics, mineralization and toxicity assessment of amoxicillin. Water Res. 106, 171–182. Garcia-Reiriz, A., Damiani, P.C., Oliveri, A.C., 2007. Different strategies for the direct determination of amoxicillin in human urine by second-order multivariate analysis of kinetic-spectrophotometric data. Talanta 71, 806–815. Gonzalez, T., Dominguez, J.R., Palo, P., Sanchez-Martin, J., Cuerda-Correa, E.M., 2011. Development and optimization of the BDD-electrochemical oxidation of the antibiotic trimethoprim in aqueous solution. Desalination 280, 197–202. Goyal, A., Srivastava, V.C., Kushwaha, J.P., 2017. Treatment of highly acidic wastewater containing high energetic compounds using dimensionally stable anode. Chem. Eng. J. 325, 289–299. Guinea, E., Brillas, E., Centellas, F., Canizares, P., Rodrigo, M.A., Saez, C., 2009. Oxidation of enrofloxacin with conductive-diamond electrochemical oxidation, ozonation and Fenton oxidation. A comparison. Water Res. 43, 2131–2138. Haidar, M., Dirany, A., Sires, I., Oturan, N., Oturan, M.A., 2013. Electrochemical degradation of the antibiotic sulfachloropyridazine by hydroxyl radicals generated at a BDD anode. Chemosphere 91, 1304–1309. Homem, V., Santos, L., 2011. Degradation and removal of antibiotics from aqueous matrices – a review. J. Environ. Manag. 92, 2304–2347. Kaur, P., Kushwaha, J.P., Sangal, V.K., 2017. Evaluation and disposability study of actual textile wastewater treatment by electro-oxidation method using Ti/RuO2 anode. Process Saf. Environ. Prot. 111, 13–22. Kaur, P., Kushwaha, J.P., Sangal, V.K., 2018a. Electrocatalytic oxidative treatment of real textile wastewater in continuous reactor: degradation pathway and disposability study. J. Hazard. Mater. 346, 242–252. Kaur, R., Kushwaha, J.P., Singh, N., 2018b. Electrooxidation of ofloxacin antibiotic by dimensionally stable Ti/RuO2 anode: evaluation and mechanistic approach. Chemosphere 193, 685–694. Kaur, R., Kushwaha, J.P., Singh, N., 2019. Amoxicillin electro-catalytic oxidation using Ti/ RuO2 anode: mechanism, oxidation products and degradation pathway. Electrochim. Acta 296, 856–866. Kemper, N., 2008. Veterinary antibiotics in the aquatic and terrestrial environment. Ecol. Indic. 8, 1–13. Khetan, S.K., Collins, T.J., 2007. Human pharmaceuticals in the aquatic environment: a challenge to green chemistry. Chem. Rev. 107, 2319–2364. Korbahti, B.K., Tasyurek, S., 2015. Electrochemical oxidation of ampicillin antibiotic at boron-doped diamond electrodes and process optimization using response surface methodology. Environ. Sci. Pollut. Res. 22, 3265–3278. Kumar, S., Singh, S., Srivastava, V.C., 2015. Electro-oxidation of nitrophenol by ruthenium oxide coated titanium electrode: parametric, kinetic and mechanistic study. Chem. Eng. J. 263, 135–143. Kummerer, K., 2001. Drugs in the environment: emission of drugs, diagnostic aids and disinfectants into wastewater by hospitals in relation to other sources – a review. Chemosphere 45, 957–969.

R. Kaur et al. / Science of the Total Environment 677 (2019) 84–97 Kummerer, K., 2009. Antibiotics in aquatic environment – a review – part I. Chemosphere 75, 417–434. Li, S., Bejan, D., McDowell, M.S., Bunce, N.J., 2008. Mixed first and zero order kinetics in the electrooxidation of sulfamethoxazole at a boron-doped diamond (BDD) anode. J. Appl. Electrochem. 38, 151–159. Lin, A.Y., Tsai, Y., 2009. Occurrence of pharmaceuticals in Taiwan's surface waters: impact of waste streams from hospitals and pharmaceutical production facilities. Sci. Total Environ. 407, 3793–3802. Lin, H., Niu, J., Xu, J., Li, Y., Pan, Y., 2013. Electrochemical mineralization of sulfamethoxazole by Ti/SnO2-Sb/Ce-PbO2 anode: Kinetics, reaction pathways, and energy cost evaluation. Electrochim. Acta 97, 167–174. Lucas, M.S., Peres, J.A., 2009. Removal of COD from olive mill wastewater by Fenton's reagent: kinetic study. J. Hazard. Mater. 168, 1253–1259. Marquardt, D.W., 1963. An algorithm for least-squares estimation of nonlinear parameters. J. Soc. Ind. Appl. Math. 11, 431–441. Martinez-Huitle, C., Ferro, S., 2006. Electrochemical oxidation of organic pollutants for the wastewater treatment: direct and indirect processes. Chem. Soc. Rev. 35, 1324–1340. Mora-Gomez, J., Ortega, E., Mestre, S., Perez-Herranz, V., Garcia-Gabaldon, M., 2019. Electrochemical degradation of norfloxacin using BDD and new Sb-doped SnO2 ceramic anodes in an electrochemical reactor in the presence and absence of a cationexchange membrane. Sep. Purif. Technol. 208, 68–75. 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. Appl. Catal. B Environ. 202, 217–261. Mutiyar, P.K., Mittal, A.K., 2013. Occurrences and fate of an antibiotic amoxicillin in extended aeration-based sewage treatment plant in Delhi, India: a case study of emerging pollutant. Desalin. Water Treat. 51, 6158–6164. Nideesh, P.V., Zhou, M., Oturan, M.A., 2018. An overview on the removal synthetic dyes from water by electrochemical advanced oxidation process. Chemosphere 197, 210–227. Nideesh, P.V., Divyapriya, G., Oturan, N., Trellu, C., Oturan, M.A., 2019. Environmental applications of boron-doped diamond electrodes: 1. Application in water and wastewater treatment. ChemElectroChem 6, 1–20. Panizza, M., Cerisola, G., 2009. Direct and mediated anodic oxidation of organic pollutants. Chem. Rev. 109, 6541–6569. Panizza, M., Dirany, A., Sires, I., Haidar, M., Oturan, N., Oturan, M.A., 2014. Complete mineralization of the antibiotic amoxicillin by electro-Fenton with the BDD anode. J. Appl. Electrochem. 44, 1327–1335.

97

Petrie, B., Barden, R., Kaspryzk-Hordern, B., 2015. A review on emerging contaminants in wastewaters and the environment: current knowledge, understudied areas and recommendations for future monitoring. Water Res. 72, 3–27. Rao, A.N.S., Venkatarangaiah, V.T., 2014. Metal-oxide-coated anodes in wastewater treatment. Environ. Sci. Pollut. Res. 21, 3197–3217. Rivera-Utrilla, J., Sanchez-Polo, M., Ferro-Garcia, M.A., Prados-Joya, G., Ocampo-Perez, R., 2013. Pharmaceuticals as emerging contaminants and their removal. Chemosphere 93, 1268–1287. Sires, I., Brillas, E., Oturan, M.A., Rodrigo, M.A., Panizza, M., 2014. Electrochemical advanced oxidation processes: today and tomorrow. A review. Environ. Sci. Pollut. Res. 21, 8336–8367. Sopaj, F., Rodrigo, M.A., Oturan, N., Podvorica, F.I., Pinson, J., 2015. Influence of the anode materials on the electrochemical oxidation efficiency. Application to oxidative degradation of the pharmaceutical amoxicillin. Chem. Eng. J. 262, 286–294. Trasatti, S., 1987. Progress of the understanding of the mechanism of chlorine evolution at oxide electrodes. Electrochim. Acta 32, 369–382. Vasilie, S., Manea, F., Baciu, A., Pop, A., 2018. Dual use of boron-doped diamond electrode in antibiotics-containing water treatment and process control. Process. Saf. Environ. Prot. 117, 446–453. Wachter, N., Aquino, J.M., Denadai, M., Barreiro, J.C., Silva, A.J., Cass, Q.B., Rocha-Filho, R.C., Bocchi, N., 2019. 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. Environ. Sci. Pollut. Res. 26 (5), 4438–4449. Wang, Y., Shen, C., Zhang, M., Zhang, B.T., Yu, Y.G., 2016. The electrochemical degradation of ciprofloxacin using a SnO2-Sb/Ti anode: influencing factors, reaction pathways and energy demand. Chem. Eng. J. 296, 79–89. Yahiaoui, I., Aissani-Benissad, F., Fourcade, F., Amrane, A., 2013. Removal of tetracycline hydrochloride from water based on direct anodic oxidation (Pb/PbO2 electrode) coupled to activated sludge culture. Chem. Eng. J. 221, 418–425. Yavuz, Y., Koparal, A.S., 2006. Electrochemical oxidation of phenol in a parallel plate reactor using ruthenium mixed metal oxide electrode. J. Hazard. Mater. 136, 296–302. Yu, X., Zhou, M., Hu, Y., Serrano, K.G., Yu, F., 2014. Recent updates on electrochemical degradation of bio-refractory organic pollutants using BDD anode: a mini review. Environ. Sci. Pollut. Res. 21, 8417–8431.