G Model

JECE 632 1–8

Journal of Environmental Chemical Engineering xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Journal of Environmental Chemical Engineering

journal homepage: [Link]/locate/jece

1 Oxidative degradation of levofloxacin in aqueous solution by

2 S2O82
/Fe2+, S2O82
/H2O2 and S2O82
/OH
processes:
3 A comparative study
4 Q1 Irina Epold, Niina Dulova *
5 Department of Chemical Engineering, Tallinn University of Technology, Ehitajate tee 5, Tallinn 19086, Estonia

A R T I C L E I N F O A B S T R A C T

Article history: The performance of levofloxacin (LFX) degradation in aqueous solutions in ferrous ion-activated
Received 27 February 2015 persulfate (S2O82
/Fe2+), peroxide-activated persulfate (S2O82
/H2O2), and base-activated persulfate
Accepted 25 April 2015 (S2O82
/OH
) systems was evaluated and compared. The LFX degradation by all studied activated
persulfate systems was divided into two oxidation periods: a fast degradation of target compound within
Keywords: the first minute and subsequent gradual oxidation within the remaining reaction time. Notably, without
Activated persulfate consideration of the first minute, the rest of the LFX degradation in the S2O82
/Fe2+ and S2O82
/H2O2
Advanced oxidation
system followed the pseudo-first-order kinetic model. Among the studied activation techniques, the Fe2
Pharmaceuticals +
Sulfate radical
-activated persulfate system demonstrated the highest efficacy in LFX degradation, mineralization and
Water/groundwater treatment persulfate utilization, followed by the peroxide-activated persulfate oxidation. Generally, all studied
sulfate radical-based advanced oxidation technologies proved to be promising tool for the treatment of
LFX contaminated water/wastewater and, especially, groundwater.
ã 2015 Published by Elsevier Ltd.

6 Introduction The application of advanced oxidation technologies (AOTs) 26

could be a viable solution for the prevention of antibiotics escape 27

7 Antibiotics are one of the most important pharmaceuticals used into the environment as well as for the in situ treatment of already 28
8 worldwide in human and veterinary medical practices [1]. The existing groundwater contamination. The radicals involved in AOTs 29
9 primary pathways for antibiotics release in to the environment are are mainly the hydroxyl radical (HO, E = 2.73 V [4]) and sulfate 30
10 wastewater treatment plants (WWTPs) effluents, animal waste radical (SO4
, E = 2.5–3.1 V [5,6]). The former can be generated by 31
11 and municipal landfill leachate [2]. Once administered, antibiotics combination of strong oxidants (hydrogen peroxide, ozone) with 32
12 are metabolized to varying degrees, and then excreted as activators (transition metals, semiconductors), UV/vis and ultra- 33
13 metabolites or unchanged parent compounds, which subsequently sound irradiation [7]. The HO reacts with organic compounds 34
14 can undergo further modifications due to biological and chemical mainly by abstracting hydrogen-atom from C

H, N

H, or O

H 35
15 processes in both WWTPs and receiving water bodies [3]. bonds, adding to double and triple bonds, or adding to aromatic 36
16 Nevertheless, antibiotics are known to be resistant to conventional rings [4,8]. The advantages of HO-AOTs as aqueous matrices 37
17 biological oxidation and usually escape intact from WWTPs, treatment techniques include high reaction rates and non- 38
18 causing long-term concerns in the environment such as bio- selective oxidation due to HO, which allow the simultaneous 39
19 accumulation and contributing to the spread of antibiotic degradation of multiple organic contaminants including pharma- 40
20 resistance in microorganisms. Particularly, the continuous intro- ceuticals [2,9–16]. The most common technique used for SO4
41
21 duction of antibiotics into the environment can affect natural generation is persulfate activation by heat, transition metal, UV or 42
22 waters quality and potentially impact drinking water supplies, ultrasound irradiation, base, or peroxide [17,18] in accordance with 43
23 ecosystem and human health [1,3]. Thus, developing effective Eq. (1): 44
24 treatment technology for degradation of antibiotics in aqueous
25 Q2 S2O82
+ activator ! SO4
+ SO4
+ SO42
) (1) 45
matrices is of a great scientific, environmental and public concern.
The activated persulfate oxidation has several advantages over 46

HO-AOTs. Accordingly, different from hydroxyl radicals, sulfate 47

radicals are more stable, more selective for oxidizing unsaturated 48

* Corresponding author: Tel.: +372 6202850; fax: +372 6202856.
E-mail address: [Link]@[Link] (N. Dulova). bond and aromatic constituents, and preferably undergo electron 49

[Link]
2213-3437/ ã 2015 Published by Elsevier Ltd.

Please cite this article in press as: I. Epold, N. Dulova, Oxidative degradation of levofloxacin in aqueous solution by S2O82
/Fe2+, S2O82
/H2O2
and S2O82
/OH
processes: A comparative study, J. Environ. Chem. Eng. (2015), [Link]
 G Model
JECE 632 1–8

2 I. Epold, N. Dulova / Journal of Environmental Chemical Engineering xxx (2015) xxx–xxx

50 transfer reactions with organics [17,18]. The SO4
-AOTs have been Sodium hydroxide (NaOH) and sulfuric acid (H2SO4) aqueous 97
51 recently used for the degradation of organic pollutants in aqueous solutions were used to adjust the pH. 98
52 matrices [19–24]. However, only few studies concerning anti-
53 biotics abatement using activated persulfate systems can be found Experimental procedure 99
54 in the literature [25,26]. Generally, high aqueous stability,
55 relatively low cost and benign end products make persulfate All the persulfate oxidation trials were performed in batch 100
56 oxidation a promising choice for in situ groundwater treatment mode and in non-buffered solutions at ambient room temperature 101
57 among the other AOTs. (21 1  C). LFX solutions (75 mM, 0.4 L) were treated in a 0.6 L 102
58 Owning to the advantages of cost effectiveness, high activity cylindrical glass reactor with a permanent agitation speed 103
59 and the environmentally friendly nature, ferrous iron (Fe2+) has (400 rpm) for a period of 3 h. The pH of the samples was adjusted 104
60 been commonly selected as the activator of persulfate in practical to 3 with H2SO4 (0.5 M), if not specified otherwise. The activator 105
61 applications [17]: (FeSO47H2O) was added, and after its complete dissolution, 106

oxidation was initiated by adding Na2S2O8. Samples were 107

62 S2O82
+ Fe2+ ! Fe3+ + SO4
+ SO42
(2) 108
withdrawn at pre-determined time intervals. The oxidation
63 2+
quenching was done by the addition of Na2SO3 at a [oxidant]0/ 109
Moreover, several studies have shown that Fe -activated
64 SO32
molar ratio (m/m) of 1/10. The experiments on LFX oxidation 110
persulfate system proved effective to degrade various persistent
65 with non-activated persulfate or hydrogen peroxide, base-activat- 111
organic pollutants in aqueous matrices [27–30]. Conversely, the
66 ed (pH 11) and peroxide-activated persulfate were conducted in 112
data on the use of other persulfate activation techniques promising
67 identical reactors and treatment conditions for the respective Fe2 113
for in situ applications such as peroxide and base have not been
68 +
fully evaluated. -activated oxidation trials. In the case of peroxide-activated 114

69 persulfate, both oxidants were added simultaneously. All experi- 115

In the present study, a broad-spectrum antibiotic of the
70 116
fluoroquinolone (FQ) drug class levofloxacin (LFX) was selected ments were duplicated; the results of the analysis are presented as
71 the mean with a standard deviation less than 5%. 117
as the target compound due to its and other FQs frequent detection
72 in aquatic environment [31]. LFX is principally used to treat severe
73 Analytical methods 118
or life-threatening bacterial infections, since it has substantial
74 activity against a broad array of Gram-positive and -negative
75 The pH was measured using a digital pH/Ion meter (Mettler 119
bacteria [32]. Similarly to other FQs, it is known to be extremely
76 Toledo S220). The initial hydrogen peroxide concentration in the 120
resistant to conventional biological oxidation and usually escapes
77 stock solutions was measured spectrophotometrically at 121
intact from wastewater treatment plants [31,33]. The main goal of
78 this work was to investigate and compare the performance of LFX l = 254 nm; the residual hydrogen peroxide concentration in the
79 122
degradation and mineralization in S2O82
/Fe2+, S2O82
/H2O2 and treated samples was measured as described in Dulov et al. [15] at
80 S2O82
/OH
systems. To the best of our knowledge, the comparison l = 410 nm by a Helios-b UV/vis spectrophotometer (Thermo
81 Electron Corporation). The residual persulfate concentration in the 123
of different persulfate activation techniques efficacies for FQs as
82 treated samples was measured spectrophotometrically at 124
well as other antibiotics degradation in aqueous matrices has not
83 been investigated yet. Generally, the data obtained within this l = 446 nm with o-dianisidine [36]. Non-purgeable organic carbon
84 study provides valuable knowledge for further implementation in (NPOC) was measured by a TOC analyzer multi N/C1 3100 125
85 (Analytik Jena). 126
water/wastewater treatment and in situ groundwater purification
86 The concentration of LFX was quantified by means of high 127
by means of activated persulfate oxidation.
performance liquid chromatography combined with a diode array 128
87 detector (HPLC-PDA, Prominence SPD-M20A, Shimadzu) equipped 129
Materials and methods
with a Phenomenex Gemini (150  2.0 mm, 1.7 mm) NX-C18 (110 Å, 130
88 Chemicals and materials 5 mm) column. The isocratic method with a solvent mixture of 10% 131

acetonitrile and 90% acetic acid aqueous solution (0.1%) was 132
89 TM
Hydrogen peroxide (Perdrogen , 30%), ferrous sulfate applied. The flow rate was maintained at 200 mL/min. Samples 133
90 heptahydrate (FeSO47H2O, 99%), levofloxacin (Fig. 1; were scanned at 190–800 nm and analyzed at l = 295 nm. The 134
91 135
C18H2OFN3O4, 98%, molecular weight 361.37 g/mol, pKa concentration of LFX was determined by using the standard
92 chemical to fit the retention time. 136
5.33 and 8.07 [34], log Kow
0.39 [35]), sodium persulfate
93 (Na2S2O8, 99%), and sodium sulfite (Na2SO3, 98%) were
94 Results and discussion 137
purchased from Sigma–Aldrich. All other chemicals of analytical
95 grade were used without further purification. Stock solutions were
96 prepared in ultrapure water (Millipore Simplicity1 UV System). In all studied persulfate systems, a fast decomposition of the 138

target compound during 1 min (the first measured time point after 139

the beginning of the reaction) with subsequent gradual degrada- 140

tion within the remaining reaction time was observed. Accordingly 141

it was suggested to divide the entire persulfate oxidation reaction 142

into two main phases: the first period of rapid LFX degradation and 143

the second period of steady target compound oxidation. Addition- 144

ally, the obtained data processing revealed that LFX degradation by 145

non-activated persulfate as well as Fe2+-and peroxide-activated 146

persulfate process followed a pseudo-first-order kinetic law during 147

the second reaction period and may be described with regard to 148

the LFX concentration through Eq. (3): 149

dC LFX
¼
k1  C LFX (3) 150
Fig. 1. Molecular structure of levofloxacin. dt

Please cite this article in press as: I. Epold, N. Dulova, Oxidative degradation of levofloxacin in aqueous solution by S2O82
/Fe2+, S2O82
/H2O2
and S2O82
/OH
processes: A comparative study, J. Environ. Chem. Eng. (2015), [Link]
 G Model
JECE 632 1–8

I. Epold, N. Dulova / Journal of Environmental Chemical Engineering xxx (2015) xxx–xxx 3

151 where k1 is the pseudo-first-order rate constant and CLFX is the LFX
152 concentration. The
k1 constants were calculated from the slopes
153 of the straight lines by plotting ln(C/C0) as a function of time t
154 through linear regression.

155 Non-activated persulfate oxidation

156 First, the performance of non-activated persulfate in LFX

157 degradation was studied. The persulfate ion (S2O82
) is known
158 to be a strong but relatively stable oxidant, which under acidic
159 conditions can hypothetically breakdown into SO4
in accordance
160 with Eqs. (4) and (5) [37]:
161 S2O82
+ H+ ! HS2O8
(4)

162 HS2O8
! H+ + SO4
+ SO42
(5)

163 Therefore, the efficacy of LFX degradation by non-activated
164 persulfate at an LFX/S2O82
m/m of 1/10 and pH 3 was evaluated.
165 The results indicated the negligible efficacy of non-activated Fig. 3. Reduction of NPOC in the S2O82
/Fe2+ system (LFX/S2O82
/Fe2+ is a molar
166 S2O82
oxidation in target compound degradation, and thus only ratio, [LFX]0 = 75 mM).
167 7% of LFX was removed in the first fast period with a k1 10
2 of
168 0.04 min
1 for the second stage, and more than 95% of the This observation can be explained by fast generation of SO4
189
169 190
unreacted S2O82
was observed after 3 h of oxidation. The observed through the decomposition of persulfate immediately after
170 activation and rapid accumulation of ferric ion (Fe3+) through 191
two-stage kinetic model of LFX degradation by non-activated
171 the complete oxidation of activator in accordance with Eqs. (2) and 192
persulfate process suggests, however, the radical-based target
172 (6). Subsequently, ineffectiveness of Fe3+ to activate persulfate 193
compound oxidation mechanism. The inhibition of oxidation in
194
173 this case was presumably caused by the preference of SO4
-to- decomposition even at acidic pH values [38] led to fast inhibition of
195
174 SO4
reactions over SO4
-to-LFX reactions. the oxidation of target compound.

175
SO4
+ Fe2+ ! Fe3+ + SO42
(6) 196
Ferrous ion-activated persulfate oxidation
Conversely, the existence of the second period of gradual LFX 197
176 198
The performance of the S2O82
/Fe2+ system in target compound degradation and steady persulfate decomposition suggests the
177 presence of other routes for reducing the activator and supporting 199
and oxidation by-products removal was studied at different LFX/
178 S2O82
and S2O82
/Fe2+ ratios. The effect of oxidant and activator SO4
generation. Recently, the simultaneous existence of both 200
179 dose on the efficacy of LFX degradation and mineralization in the SO4
and HO in the S2O82
/Fe2+ system was demonstrated by Wu 201
180 202
S2O82
/Fe2+ oxidation was evaluated, and the results are presented et al. [39]. Therefore, the second phase of gradual LFX degradation
181 in S2O82
/Fe2+ systems can be explained by the continuous 203
in Figs. 2 and 3.
182 reduction of Fe3+ in the reaction with organic radicals as presented 204
In the case of the S2O82
/Fe2+ system, the two-stage kinetic
183 in Eq. (7), by the formation of hydrogen peroxide in accordance 205
model was observed not only for LFX degradation but also for
184 with Eq. (8), and the subsequent direct reduction of Fe3+ through 206
persulfate decomposition (Figs. 2 and 4). Accordingly, more
185 Eq. (9), or by the combined action of both listed processes. 207
than 31 and 56% of LFX was removed as well as 7 and 11% of
186 persulfate was decomposed during the first minute and the rest of
187 R + Fe3+ ! Fe2+ + products (7) 208
the 3 h oxidation, respectively, in S2O82
/Fe2+ systems at an
188 LFX/S2O82
/Fe2+ m/m/m of 1/10/1.
1
1

0.9 1/10/0.5 1/10/1 1/20/1

0.9
0.8
1/20/2 1/20/4 1/30/3
0.7
[S2O82-]/[S2O82-]0

0.8
[LFX]/[LFX]0

0.6

0.5

0.4 0.7

0.3 1/10/0.5 1/10/1 1/20/1

0.2 0.6
1/20/2 1/20/4 1/30/3
0.1

0 0.5
0 20 40 60 80 100 120 140 160 180 0 20 40 60 80 100 120 140 160 180
Time, min Time, min

Fig. 2. Degradation of LFX in the S2O82
/Fe2+ system (LFX/S2O82
/Fe2+ is a molar Fig. 4. Decomposition of S2O82
in the S2O82
/Fe2+ system (LFX/S2O82
/Fe2+ is a
ratio, [LFX]0 = 75 mM). molar ratio, [LFX]0 = 75 mM).

Please cite this article in press as: I. Epold, N. Dulova, Oxidative degradation of levofloxacin in aqueous solution by S2O82
/Fe2+, S2O82
/H2O2
and S2O82
/OH
processes: A comparative study, J. Environ. Chem. Eng. (2015), [Link]
 G Model
JECE 632 1–8

4 I. Epold, N. Dulova / Journal of Environmental Chemical Engineering xxx (2015) xxx–xxx

Table 1
Kinetic parameters of LFX degradation in the S2O82
/Fe2+ system under different treatment conditions ([LFX]0 = 75 mM).

LFX/S2O82/Fe2+, m/m/m pHinitial [LFX]removed, % (t = 1 min) Pseudo-first-order constant,

k1 10
2 min
1 (r2 > 0.99)
1/10/0.5 3 25 0.65  0.02
1/10/1 3 31 0.93  0.02
1/20/1 3 36 1.57  0.05
1/20/2 3 43 2.42  0.04
1/20/2 5 39 2.41  0.05
1/20/2 7 36 2.33  0.03
1/20/2 9 31 1.44  0.04
1/20/4 3 64 3.61  0.08
1/30/3 3 56 5.74  0.15

degradation was found to decrease gradually with the increase in 250

the initial pH value. Accordingly, a 1.7-fold decrease in k1 value of 251

209 2HO ! H2O2 (8) 252
the second reaction period and 12% lower removal of LFX during
1 min was observed at initial pH 9 compared to pH 3. Conversely, 253

the obtained results revealed the comparable performance of LFX 254

210 Fe3+ + H2O2 ! Fe2+ + HO2 + H+ (9) 255
degradation in the S2O82
/Fe2+ system at initial pH values of 3,
211 The termination of oxidation in the S2O82
/Fe2+
systems is 256
5 and 7, indicating a high potential of this process for in situ
212 believed to occur through Eq. (6) and Eqs. (10)–(13) [4,40]: 257
applications at natural waters pH values. Notably, the solution pH
in all trials decreased to around 3 after 3 h of oxidation, mainly due 258
213  
HO2 + HO ! H2O + O2 (10) 259
to acidity of oxidant and its decomposition products, metal-related
acidity and acidic LFX decomposition by-products. 260


 In order to suppress the activator-consuming and SO4
- 261

214 SO4 + HO ! HSO5
(11) 262
scavenging reaction presented in Eq. (6), the adjustment of
S2O82
/Fe2+ system by correction of the activator dose was studied. 263

Accordingly, three S2O82
/Fe2+ m/m ratios of 20/1, 10/1, and 264

215 HO2 + SO4
! HSO4
+ O2 (12) 265
5/1 were evaluated at a fixed LFX/S2O82
m/m of 1/20. The results
indicated equally faster target compound degradation and 266

persulfate decomposition in both reaction periods as well as 267

216 SO4
+ SO4
! S2O82
(13) better mineralization at a S2O82
/Fe2+ m/m ratio of 5/1 (Figs. 2–4, 268

Table 1). However, considering the high stability of persulfate 269

217 As it was above-mentioned, without consideration of the first
along with its high acidity ensuring acidic pH values in reaction 270
218 minute, the rest of the performance of LFX degradation in the
media, prolonged oxidation at a S2O82
/Fe2+ m/m of 10/1 could be a 271
219 S2O82
/Fe2+ system followed the pseudo-first-order kinetic model.
more reasonable alternative to the elevated activator dose (5/1 and 272
220 The values of the kinetic constants calculated for the S2O82
/Fe2+
higher) both for water and wastewater treatment as well as for in 273
221 system under different treatment conditions are presented in
situ groundwater decontamination. 274
222 Table 1.
223 The obtained results indicated a stable improvement in the LFX
Peroxide-activated persulfate oxidation 275
224 degradation rate with the increase in applied oxidant dose at the
225 same S2O82
/Fe2+ ratio. Accordingly, the k1 10
2 of the second 276
To evaluate the efficacy of LFX degradation by peroxide-
226 reaction period was 0.93  0.02, 2.42  0.04, and 5.74  0.15 min
1 277
227 activated persulfate oxidation, the effect of persulfate and
for the S2O82
/Fe2+ oxidation at an LFX/S2O82
/Fe2+ m/m/m of 278
228 hydrogen peroxide dosages was studied (Fig. 5). Similarly to the
1/10/1, 1/20/2, and 1/30/3, respectively. Similarly, the decomposi-
229 tion of LFX during the first stage of reaction improved with the
230 increase in oxidant dose, indicating a more comprehensive target 1
231 compound oxidation until the added activator is completely
232 oxidized through Eqs. (2) and (6). In the case of S2O82

233 decomposition, a more obvious dependence on oxidant dose
234 was observed during the first minute, with almost similar 0.9
235 tendency of decomposition within the rest of 3 h oxidation.
[LFX]/[LFX]0

236 Unlike LFX degradation, the extent of mineralization in the

237 S2O82
/Fe2+ system was considerably lower, mainly indicating
238
0.8
the accumulation of transformation products in the reaction
239 mixture. Consequently, the NPOC concentration remained nearly
240 unchanged after 3 h of oxidation at an LFX/S2O82
/Fe2+ m/m/m
241 lower than 1/20/2 (less than 3% mineralization). The highest 0.7
242 obtained NPOC removal under the studied treatment conditions
243
1/10/2.5 1/10/5 1/10/10 1/15/5 1/20/5
was 11% at an LFX/S2O82
/Fe2+ m/m/m of 1/30/3.
244 The pH value of the solution is known to be an important
245 parameter affecting degradation of organic contaminants. There- 0.6
246 fore, the effect of initial pH on the degradation of LFX by Fe2 0 20 40 60 80 100 120 140 160 180
247 +
-activated persulfate was evaluated. The results of LFX oxidation Time, min
248 at an LFX/S2O82
/Fe2+ m/m/m of 1/20/2 and four different pH Fig. 5. Degradation of LFX in the S2O82
/H2O2 system (LFX/S2O82
/H2O2 is a molar
249 values (3, 5, 7 and 9) are presented in Table 1. The efficacy of LFX ratio, [LFX]0 = 75 mM).

Please cite this article in press as: I. Epold, N. Dulova, Oxidative degradation of levofloxacin in aqueous solution by S2O82
/Fe2+, S2O82
/H2O2
and S2O82
/OH
processes: A comparative study, J. Environ. Chem. Eng. (2015), [Link]
 G Model
JECE 632 1–8

I. Epold, N. Dulova / Journal of Environmental Chemical Engineering xxx (2015) xxx–xxx 5

Table 2
Kinetic parameters of LFX degradation in the S2O82
/H2O2 system under different treatment conditions ([LFX]0 = 75 mM).

LFX/S2O82
/H2O2, pHinitial [LFX]removed, % (t = 1 min) Pseudo-first-order constant, k1 10
2 min
1 (r2 > 0.99)
m/m/m
1/10/2.5 3 10 0.08  0.003
1/10/5 3 10.5 0.08  0.002
1/10/5 5 10 0.06  0.001
1/10/5 7 10.5 0.05  0.002
1/10/5 9 10 0.02  0.002
1/10/10 3 8 0.07  0.003
1/15/5 3 16 0.06  0.001
1/20/5 3 7.5 0.07  0.002

279 non-activated and Fe2+-activated persulfate system LFX degrada- LFX degradation during the first period. A further increase in the 310
280 tion in S2O82
/H2O2 process was divided to two oxidation periods S2O82
/H2O2 ratio to 4/1 demonstrated decrease in target com- 311
281 and the rate of LFX degradation followed the pseudo-first order in pound oxidation efficacy. Thus, the highest LFX degradation efficacy 312
282 the second steady oxidation stage (Table 2). Notably, in the case of was observed at an LFX/S2O82
/H2O2 m/m/m of 1/15/5 with 16 and 313
283 the S2O82
/H2O2 system, residual concentrations of persulfate 25% of LFX removed during the first minute and overall oxidation 314
284 were not measured due to interference caused by the presence of time, respectively. Irrespective of the applied S2O82
/H2O2 molar 315
285 hydrogen peroxide in samples. Conversely, the presence of residual ratio, the performance of peroxide-activated persulfate in LFX 316
286 persulfate still allowed the measurement of unused hydrogen degradation during the second oxidation period was comparable, 317
287 peroxide, and the results are presented in Fig. 6. and thus the kinetic constants (k1 10
2) ranged from 0.06 to 318
288 The obtained results demonstrated to some extent improvement 0.08 min
1. Similarly to LFX degradation, the most effective 319
289 in the target compound decomposition at an LFX/S2O82
/H2O2 of utilization of hydrogen peroxide was observed at an LFX/S2O82
/ 320
290 1/10/10 compared to non-activated persulfate oxidation at an H2O2 m/m/m of 1/15/5 with only 15% of utilized H2O2, indicating the 321
291 LFX/S2O82
m/m 1/10, and thus 8% of LFX removed in the first stage high potential of prolonged oxidation for subsequent increase in 322
292 with k1 10
2 of 0.07 min
1 for the second gradual period and more treatment efficacy. Additionally, the efficacy of LFX oxidation by- 323
293 than 99% of residual H2O2 after 3 h of oxidation was observed. products degradation in the S2O82
/H2O2 system was assessed, and 324
294 Conversely, the efficacy of LFX degradation in the S2O82
/H2O2 the results obtained indicated negligible NPOC removal with the 325
295 system was considerably lower compared with the S2O82
/Fe2+ highest achieved mineralization of 1–2%. 326
296 oxidation at the same LFX/S2O82
ratios (Figs. 2 and 5). To adjust the To evaluate the effect of initial pH on the performance of LFX 327
297 S2O82
/H2O2 ratio and as a result to improve the overall efficacy of degradation in the S2O82
/H2O2 system, the target compound 328
298 the peroxide-activated persulfate system, the effect of hydrogen degradation at an LFX/S2O82
/H2O2 m/m/m of 1/10/5 was 329
299 peroxide and persulfate dosages was studied at a fixed LFX/S2O82
additionally studied at pH 5, 7 and 9 (Table 2.). Similarly to the 330
300 and LFX/H2O2 ratio, respectively. The obtained results indicated a S2O82
/Fe2+ system, the efficacy of LFX removal in the S2O82
/H2O2 331
301 strong necessity for careful optimization of the S2O82
/H2O2 system process decreased with the increase in the initial pH value. 332
302 (Fig. 5, Table 2). Accordingly, the increase in S2O82
/H2O2 molar ratio Accordingly, a 4-fold decrease in k1 value of the second reaction 333
303 from 1/1 to 2/1 at a fixed LFX/S2O82
m/m of 1/10 resulted in 8 and period was observed at initial pH 9 compared to pH 3. This 334
304 10.5% LFX removal within the first minute of oxidation, respectively. observation is in agreement with previous studies in literature 335
305 A further increase in the S2O82
/H2O2 molar ratio to 4/1 resulted in concerning organic contaminants degradation by activated per- 336
306 similar 10% LFX removal within the first minute of oxidation, sulfate processes [19,37,38]. 337
307 indicating the irrationality of further decrease in hydrogen peroxide To clarify the predominant mechanism of LFX decomposition in 338
308 dose. In the case of trials with a fixed LFX/H2O2 ratio (1/5, m/m), an the S2O82
/H2O2 system, the oxidative potential of non-activated 339
309 increase in S2O82
/H2O2 molar ratio from 2/1 to 3/1 resulted in faster hydrogen peroxide was studied as well. The results of target 340

compound oxidation at an LFX/H2O2 m/m of 1/10 indicated no LFX 341

removal and more than 99% of unused H2O2 in solution after 3 h of 342
1
treatment. Generally, it is known that a dual oxidant system 343

utilizing S2O82
and H2O2 combines the reactivity of peroxide in 344

the reduction of target organic contaminants with the enhanced 345

stability of persulfate and involves the simultaneous generation of 346

0.9 HO and SO4
[17]. However, the mechanism of peroxide 347
348
[H2O2]/[H2O2]0

activation of persulfate is still uncertain; the suggested activation

mechanisms include the HO generated from peroxide or the heat 349

from the exothermic hydrogen peroxide reactions [17]. The later 350

can be described with the sequence of reaction proposed by Tsao 351

and Wilmarth [41] for the S2O82
/H2O2 system at T = 30  C, where 352

0.8 353
the chain initiation was suggested to be caused by the thermal
decomposition of S2O82
through Eq. (14). Additionally, it was 354
1/10/2.5 1/10/5 1/10/10 1/15/5 1/20/5 assumed that direct peroxide activation through Eq. (15) plays an 355

insignificant part in the initiation of persulfate decomposition to 356

SO4
. 357
0.7
0 20 40 60 80 100 120 140 160 180 S2O82
! 2SO4
(14) 358
Time, min

Fig. 6. Decomposition of H2O2 in the S2O82
/H2O2 system (LFX/S2O82
/H2O2 is a

molar ratio, [LFX]0 = 75 mM). S2O82
+ H2O2 ! SO4
+ HO2  + HSO4
(15) 359

Please cite this article in press as: I. Epold, N. Dulova, Oxidative degradation of levofloxacin in aqueous solution by S2O82
/Fe2+, S2O82
/H2O2
and S2O82
/OH
processes: A comparative study, J. Environ. Chem. Eng. (2015), [Link]
 G Model
JECE 632 1–8

6 I. Epold, N. Dulova / Journal of Environmental Chemical Engineering xxx (2015) xxx–xxx

360 Then, the chain-carrying step can be described through Eqs. Table 3
361 Kinetic parameters of LFX degradation in the S2O82
/OH
system under different
(16)–(19) and the termination of the oxidation in accordance with
362 treatment conditions ([LFX]0 = 75 mM).
Eqs. (10)–(13).
LFX/S2O82
, [LFX]removed, % [S2O82
]remained, %
363 SO4
+ H2O ! HO + SO42
+ H+ (16) m/m (t = 1 min) (t = 180 min)
1/5 6 94.5
1/10 13 94

364 HO + H2O2 ! H2O + HO2 (17) 1/15 14.5 96
1/20 17 98

365 HO2 + S2O82
! HSO4
+ SO42
+ O2 (18) was noticeably less effective than LFX degradation and in general 395

resulted in less than 1% of NPOC removal. Taking into account the 396

treatment performance, oxidant utilization and mineralization 397

366 HO2 + H2O2 ! HO + H2O + O2 (19) extent, the base-activated persulfate system proved the less 398

effective for LFX degradation in aqueous solution among other 399

367 According to the results obtained within this study, the
studied activated persulfate systems (Fig. 8). 400
368 application of the peroxide-activated system with carefully
A possible mechanism of base activation of persulfate was 401
369 adjusted activator dose and preferably prolonged treatment period
suggested by Furman et al. [42]. The initial step of the proposed 402
370 could be a viable solution for the degradation of LFX and other
mechanism is the base-catalyzed hydrolysis of S2O82
to perox- 403
371 micropollutants in water/wastewater and especially groundwater
omonosulfate ion (SO52
) and sulfate ion (SO42
) through Eq. (20) 404
372 were naturally occurring iron could play a significant role in
373 with subsequent formation of hydroperoxide (HO2
) in accordance 405
subsequent improvement of the overall treatment efficacy.
with Eq. (21). Then, the HO2
reduces another persulfate molecule, 406

374 generating SO4
and superoxide (O2
) through Eq. (22). Further- 407
Base-activated persulfate oxidation
more, in highly alkaline conditions, the SO4
reacts with OH
to 408

form HO in accordance with Eq. (23). 409

375 The results of LFX degradation by the system are S2O82
/OH

376 410
presented in Fig. 7 and Table 3. In the case of base-activated S2O82
+ H2O ! SO52
+ SO42
+ 2H+ (20)
377 persulfate system, the values of kinetic constants were not
378 calculated for the second oxidation period because the data did
379 not fit any kinetic model. SO52
+ H2O ! HO2
+ SO42
+ H+ (21) 411
380 The efficacy of LFX degradation in the base-activated persulfate
381 was strongly dependent on the LFX/S2O82
ratio applied.
382 Accordingly, LFX removal was 6/14 and 13/18% after 1 min/3 h HO2
+ S2O82
! SO4
+ SO42
+ H+ + O2
HO2
+ S2O82
! 412
383 oxidation at an LFX/S2O82
m/m of 1/5 and 1/10, respectively. A SO4
+ SO42
+ H+ + O2
(22) 413
384 further increase in the oxidant dose from 1/10 to 1/20 (LFX/S2O82
,
385 m/m) resulted in 4 and 7% of supplementary LFX removal during 1
386 min and 3 h oxidation, respectively, indicating the efficacy of SO4
+ OH
! SO42
+ HOSO4
+ OH
! SO42
+ HO (23) 414
387 elevated oxidant dose application to enhance target compound
388 removal. Conversely, the utilization of persulfate proved more The SO4
and HO generated in Eqs. (22) and (23) proceed 415

389 through propagation and scavenging reactions, Eqs. (10)–(13), in 416

effective at moderate LFX/S2O82
molar ratio of 1/10, but still
390 base-activated persulfate system, resulting mainly in the genera- 417
resulted in 94% of unused S2O82
in solution after 3 h of treatment.
391 Therefore, in the case of S2O82
/OH
system the prolonged tion of molecular oxygen. The O2
is likely scavenged by HO and 418

392 oxidation could be a reasonable solution to improve the perfor- SO4
through Eqs. (24) and (25), respectively, and thus expected 419

393 mance of target compound degradation. Similarly to the

394 S2O82
/H2O2 system, the mineralization in the S2O82
/OH
system
100
1
90

80 [LFX] [NPOC] [oxidant]

0.9 70
Removal, %

60
[LFX]/[LFX]0

50
0.8
40

30

0.7 20
1/5 1/10 1/15 1/20
10

0
0.6 Fe(II) activation Peroxide activation Base activation
0 20 40 60 80 100 120 140 160 180
Time, min Fig. 8. Removal of LFX, NPOC and oxidant after 3 h of oxidation in the S2O82
/Fe2+
system (LFX/S2O82
/Fe2+ m/m/m of 1/20/2, [LFX]0 = 75 mM), S2O82
/H2O2 system
Fig. 7. Degradation of LFX in the S2O82
/OH
system (LFX/S2O82
is a molar ratio, (LFX/S2O82
/H2O2 m/m/m of 1/15/5), and S2O82
/OH
system (LFX/S2O82
m/m of
[LFX]0 = 75 mM). 1/20).

Please cite this article in press as: I. Epold, N. Dulova, Oxidative degradation of levofloxacin in aqueous solution by S2O82
/Fe2+, S2O82
/H2O2
and S2O82
/OH
processes: A comparative study, J. Environ. Chem. Eng. (2015), [Link]
 G Model
JECE 632 1–8

I. Epold, N. Dulova / Journal of Environmental Chemical Engineering xxx (2015) xxx–xxx 7

420 to have a negligible impact on the overall oxidative potential of the 42 (3) (2012) 251–325, doi:[Link] 473
474
421 S2O82
/OH
system. 507698.
[8] W. Jiang, J.A. Joens, D.D. Dionysiou, K.E. O’Shea, Optimization of photocatalytic
475
422 HO + O2
! O2 + OH
(24) performance of TiO2 coated glass microspheres using response surface
476
methodology and the application for degradation of dimethyl phthalate, J.
Photochem. Photobiol. A Chem. 262 (2013) 7–13, doi:[Link] 477
10.1016/[Link].2013.04.008. 478
423 SO4
+ O2
! SO42
+ O2SO4
+ O2
! SO42
+ O2 (25) [9] M. Trapido, R. Munter, Y. Veressinina, N. Kulik, Oil shale semicoke leachate
479
treatment using ozonation and the Fenton oxidation, Environ. Technol. 27 (3)
424 (2006) 307–315, doi:[Link] 480
Generally, it is suggested that alkaline activation is one of the 481
425 16548211.
least efficient approaches due to the very fast decomposition of [10] N. Kulik, M. Trapido, Y. Veressinina, R. Munter, Treatment of surfactant sta-
426 persulfate under alkaline conditions [17]. However, the results of 482
bilized oil-in-water emulsions by means of chemical oxidation and coagu-
427 lation, Environ. Technol. 28 (12) (2007) 1345–1355, doi:[Link] 483
the present study indicated the fast decomposition of persulfate 484
428 10.1080/09593332808618896. 18341145.
only during the first fast period of target compound degradation [11] M. Klavarioti, D. Mantzavinos, D. Kassinos, Removal of residual pharma-
429 with the subsequent gradual decomposition during the rest of 485
ceuticals from aqueous systems by advanced oxidation processes, Environ. Int.
486
430 reaction in the S2O82
/OH
system, indicating a high potential of 35 (2) (2009) 402–417, doi:[Link]
487
431 18760478.
prolonged oxidation to enhance the overall treatment efficacy [12] A. Goi, Y. Veressinina, M. Trapido, Fenton process for landfill leachate treat-
432 mainly for in situ applications. 488
ment: evaluation of biodegradability and toxicity, J. Environ. Eng. 136 (1)
(2010) 46–53, doi:[Link] 489
433 [13] I. Oller, S. Malato, J.A. Sánchez-Pérez, Combination of advanced oxidation
Conclusions 490
processes and biological treatments for wastewater decontamination—A re-
view, Sci. Total Environ. 409 (20) (2011) 4141–4166, doi:[Link] 491
434 492
The ferrous ion-activated, peroxide-activated and base-activat- 10.1016/[Link].2010.08.061. 20956012.
435 ed persulfate systems proved to be promising techniques for the [14] E. Hapeshi, I. Fotiou, D. Fatta-Kassinos, Sonophotocatalytic treatment of
ofloxacin in secondary treated effluent and elucidation of its transformation 493
436 degradation of LFX in aqueous solution. The main conclusions of 494
products, Chem. Eng. J. 224 (2013) 96–105, doi:[Link]
437 this work are summarized in the following points: cej.2012.11.048. 495
[15] A. Dulov, N. Dulova, M. Trapido, Photochemical degradation of nonylphenol in
aqueous solution: the impact of pH and hydroxyl radical promoters, J. Environ. 496
438  The performance of the studied processes in target compound 497
Sci. (China) 25 (7) (2013) 1326–1330, doi:[Link]
439 oxidation was as follows: the S2O82
/Fe2+ system > the S2O82
/ 498
0742(12)60205-8. 24218843.
440 H2O2 system > the S2O82
/OH
system. [16] A. Babuponnusami, K. Muthukumar, A review on Fenton and improvements to
499
441  The highest mineralization and more complete persulfate the Fenton process for wastewater treatment, J. Environ. Chem. Eng. 2 (1)
(2014) 557–572, doi:[Link] 500
442 utilization were observed in the Fe2+-activated persulfate system [17] A. Tsitonaki, B. Petri, M. Crimi, H. Mosbaek, R.L. Siegrist, P.L. Bjerg, In situ
443 chemical oxidation of contaminated soil and groundwater using persulfate: a 501
followed by the peroxide-activated persulfate oxidation. 502
444  The results obtained within this study suggested that all studied review, Crit. Rev. Environ. Sci. Technol. 40 (1) (2010) 55–91, doi:[Link]
org/10.1080/10643380802039303. 503
445 activated persulfate systems with prolonged oxidation period [18] W. Chu, Y.R. Wang, H.F. Leung, Synergy of sulfate and hydroxyl radicals in UV/
446 S2O82
/H2O2 oxidation of iodinated X-ray contrast medium iopromide, Chem. 504
could be a reliable technology for wastewater and groundwater
Eng. J. 178 (2011) 154–160, doi:[Link] 505
447 remediation contaminated by the LFX.
[19] C. Liang, Z.S. Wang, C.J. Bruell, Influence of pH on persulfate oxidation of TCE at
ambient temperatures, Chemosphere 66 (1) (2007) 106–113, doi:[Link] 506
448 507
Generally, the findings in this study may provide important org/10.1016/[Link].2006.05.026. 16814844.
449 insight for further implementation in the treatment of FQ [20] S.-H. Yoon, S. Jeong, S. Lee, Oxidation of bisphenol A by UV/S2O82
: comparison
with UV/H2O2, Environ. Tech. 33 (1–3) (2012) 123–128, doi:[Link] 508
450 contamination in water/wastewater and, especially, groundwater. 509
10.1080/09593330.2011.579181. 22519095.
[21] J. Deng, Y. Shao, N. Gao, Y. Deng, S. Zhou, X. Hu, Thermally activated persulfate
451 (TAP) oxidation of antiepileptic drug carbamazepine in water, Chem. Eng. J. 510
Acknowledgments 511
228 (2013) 765–771, doi:[Link]
[22] B. Li, L. Li, K. Lin, W. Zhang, S. Lu, Q. Luo, Removal of 1,1,1-trichloroethane from
452 Q3 The financial support provided by Institutional Research 512
aqueous solution by a sono-activated persulfate process, Ultrason. Sonochem.
453 20 (3) (2013) 855–863, doi:[Link] 513
Funding IUT (1–7 of the Estonian Ministry of Education and 514
454 23266439.
Research is gratefully acknowledged. The authors thank Ms. Ave [23] S. Nachiappan, K. Muthukumar, Treatment of pharmaceutical effluent by
455 Kaskla for her assistance with the experiments. 515
ultrasound coupled with dual oxidant system, Environ. Technol. 34 (1–4)
(2013) 209–217, doi:[Link] 516
23530332. 517
456 References
[24] C. Tan, N. Gao, Y. Deng, W. Rong, S. Zhou, N. Lu, Degradation of antipyrine by
heat activated persulfate, Sep. Purif. Technol. 109 (2013) 122–128, doi:http:// 518
[1] K. Kümmerer, Antibiotics in the aquatic environment – a review – part I, [Link]/10.1016/[Link].2013.03.003. 519
457 Chemosphere 75 (4) (2009) 417–434, doi:[Link] [25] A. Ghauch, G. Ayoub, S. Naim, Degradation of sulfamethoxazole by persulfate
458 sphere.2008.11.086. 19185900. 520
assisted micrometric Fe0 in aqueous solution, Chem. Eng. J. 228 (2013) 1168–
[2] J. Rivera-Utrilla, M. Sánchez-Polo, M.Á. Ferro-García, G. Prados-Joya, R. 1181, doi:[Link] 521
459 Ocampo-Pérez, Pharmaceuticals as emerging contaminants and their removal [26] Y. Ji, C. Ferronato, A. Salvador, X. Yang, J.-M. Chovelon, Degradation of cip-
460 from water. A review, Chemosphere 93 (7) (2013) 1268–1287, doi:[Link] 522
rofloxacin and sulfamethoxazole by ferrous-activated persulfate: implications
461 [Link]/10.1016/[Link].2013.07.059. 24025536. 523
for remediation of groundwater contaminated by antibiotics, Sci. Total En-
[3] V. Homem, L. Santos, Degradation and removal methods of antibiotics from viron. 472 (2014) 800–808, doi:[Link] 524
462 aqueous matrices – a review, J. Environ. Manage. 92 (10) (2011) 2304–2347, 525
2013.11.008. 24342085.
463 doi:[Link] 21680081. [27] F. Vicente, A. Santos, A. Romero, S. Rodriguez, Kinetic study of diuron oxidation
[4] J.J. Pignatello, E. Oliveros, A. MacKay, Advanced oxidation processes for organic and mineralization by persulphate: effects of temperature, oxidant concen- 526
464 contaminant destruction based on the Fenton reaction and related chemistry, 527
tration and iron dosage method, Chem. Eng. J. 170 (1) (2011) 127–135, doi:
465 Crit. Rev. Environ. Sci. Technol. 36 (1) (2006) 1–84, doi:[Link] 528
[Link]
466 10.1080/10643380500326564. [28] C. Tan, N. Gao, W. Chu, C. Li, M.R. Templeton, Degradation of diuron by per-
[5] P. Neta, R.E. Huie, A.B. Ross, Rate constants for reactions of inorganic radicals in sulfate activated with ferrous ION, Sep. Purif. Technol. 95 (2012) 44–48, doi: 529
467 aqueous solution, J. Phys. Chem. Ref. Data 17 (3) (1988) 1027–1284, doi:http:// 530
[Link]
468 [Link]/10.1063/1.555808. [29] N. Sabri, K. Hanna, V. Yargeau, Chemical oxidation of ibuprofen in the presence
[6] F. Ghanbari, M. Moradi, M. Manshouri, Textile wastewater decolorization by of iron species at near neutral pH, Sci. Total Environ. 427–428 (2012) 382–389, 531
469 zero valent iron activated peroxymonosulfate: compared with zero valent 532
doi:[Link] 22578697.
470 copper, J. Environ. Chem. Eng. 2 (3) (2014) 1846–1851, doi:[Link] [30] S. Rodriguez, L. Vasquez, D. Costa, A. Romero, A. Santos, Oxidation of orange G
471 10.1016/[Link].2014.08.003. 533
by persulfate activated by Fe(II), Fe(III) and zero valent iron (ZVI), Chemo-
[7] J.L. Wang, L.J. Xu, Advanced oxidation processes for wastewater treatment: sphere 101 (2014) 86–92, doi:[Link] 534
472 formation of hydroxyl radical and application, Crit. Rev. Environ. Sci. Technol. 535
2013.12.037. 24439838.

Please cite this article in press as: I. Epold, N. Dulova, Oxidative degradation of levofloxacin in aqueous solution by S2O82
/Fe2+, S2O82
/H2O2
and S2O82
/OH
processes: A comparative study, J. Environ. Chem. Eng. (2015), [Link]
 G Model
JECE 632 1–8

8 I. Epold, N. Dulova / Journal of Environmental Chemical Engineering xxx (2015) xxx–xxx

[31] X. van Doorslaer, J. Dewulf, H. van Langenhove, K. Demeestere, Fluoroquin- [38] C. Liang, C.-P. Liang, C.-C. Chen, pH dependence of persulfate activation by
536 olone antibiotics: an emerging class of environmental micropollutants, EDTA/Fe(III) for degradation of trichloroethylene, J. Contam. Hydrol. 106 (3–4) 552
537 Sci. Total Environ. 500–501 (2014) 250–269, doi:[Link] (2009) 173–182, doi:[Link] 553
538 scitotenv.2014.08.075. 25226071. 19286273. 554
[32] N.H. El Najjar, A. Touffet, M. Deborde, R. Journel, N.K. Karpel Vel Leitner, [39] X. Wu, X. Gu, S. Lu, M. Xu, X. Zang, Z. Miao, Z. Qiu, Q. Sui, Degradation of
539 Levofloxacin oxidation by ozone and hydroxyl radicals: kinetic study, trans- trichloroethylene in aqueous solution by persulfate activated with citric acid 555
540 formation products and toxicity, Chemosphere 93 (4) (2013) 604–611, doi: chelated ferrous ion, Chem. Eng. J. 255 (2014) 585–592, doi:[Link] 556
541 [Link] 23850240. 10.1016/[Link].2014.06.085. 557
[33] M. Yasojima, N. Nakada, K. Komori, Y. Suzuki, H. Tanaka, Occurrence of lev- [40] H. Kusic, I. Peternel, S. Ukic, N. Koprivanac, T. Bolanca, S. Papic, A.L. Bozic,
542 ofloxacin, clarithromycin and azithromycin in wastewater treatment plant in Modeling of iron activated persulfate oxidation treating reactive azo dye in 558
543 Japan, Water Sci. Technol. 53 (11) (2006) 227–233. 16862794. water matrix, Chem. Eng. J. 172 (1) (2011) 109–121, doi:[Link] 559
[34] N.H. El Najjar, M. Deborde, R. Journel, N.K. Vel Leitner, Aqueous chlorination of 10.1016/[Link].2011.05.076. 560
544 levofloxacin: kinetic and mechanistic study, transformation product identi- [41] M.-S. Tsao, W.K. Wilmarth II, Oxidation-reduction reactions involving inor-
545 fication and toxicity, Water Res. 47 (1) (2013) 121–129, doi:[Link] ganic substrates. aqueous chemistry of inorganic free radicals. Part 3. The 561
546 10.1016/[Link].2012.09.035. 23084340. kinetics and mechanism of the reaction of peroxydisulphate ion and hydrogen 562
[35] C. Hansch, A. Leo, D. Hoekman, Exploring QSAR – Hydrophobic, Electronic, and peroxide, Discuss. Faraday Soc. 29 (1960) 137–145, doi:[Link] 563
547 Steric Constants, American Chemical Society, Washington, DC, 1995. 10.1039/df9602900137. 564
[36] N.A. Sof’ina, M.K. Beklemishev, A.L. Kapanadze, I.F. Dolmanova, Sorption- [42] O.S. Furman, A.L. Teel, R.J. Watts, Mechanism of base activation of persulfate,
548 catalytic method for the chromium determination, Mosc. Univ. Chem. Bull. 44 Environ. Sci. Technol. 44 (16) (2010) 6423–6428, doi:[Link] 565
549 (2003) 189–198 (in Russian). 10.1021/es1013714. 20704244. 566
[37] A. Romero, A. Santos, F. Vicente, C. González, Diuron abatement using activated
550 persulphate: effect of pH, Fe(II) and oxidant dosage, Chem. Eng. J. 162 (1)
551 (2010) 257–265, doi:[Link]

Please cite this article in press as: I. Epold, N. Dulova, Oxidative degradation of levofloxacin in aqueous solution by S2O82
/Fe2+, S2O82
/H2O2
and S2O82
/OH
processes: A comparative study, J. Environ. Chem. Eng. (2015), [Link]