Accepted Manuscript

Oxidation of antipyrine by chlorine dioxide: reaction kinetics and degradation

pathway

Xing-Hua Jia, Li Feng, Yong-Ze Liu, Li-Qiu Zhang

PII: S1385-8947(16)31471-1
DOI: http://dx.doi.org/10.1016/j.cej.2016.10.062
Reference: CEJ 15918

To appear in: Chemical Engineering Journal

Received Date: 21 July 2016

Revised Date: 3 October 2016
Accepted Date: 14 October 2016

Please cite this article as: X-H. Jia, L. Feng, Y-Z. Liu, L-Q. Zhang, Oxidation of antipyrine by chlorine dioxide:
reaction kinetics and degradation pathway, Chemical Engineering Journal (2016), doi: http://dx.doi.org/10.1016/
j.cej.2016.10.062

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1 Oxidation of antipyrine by chlorine dioxide: reaction

2 kinetics and degradation pathway

3 Xing-Hua Jia, Li Feng, Yong-Ze Liu, Li-Qiu Zhang
4 Beijing Key Lab for Source Control Technology of Water Pollution, College of Environmental

5 Science and Engineering, Beijing Forestry University, Beijing 100083, PR China

6 Abstract：
7 Antipyrine (ANT, phenazone), a widely used anti-inflammatory analgesic in medical

8 treatment, has been frequently detected in the aquatic environment. Chlorine disinfection process

9 is thought as an efficient way to remove ANT, however, the potential risks of chlorine disinfection

10 by-products (DBPs) such as trihalomethane (THMs) and haloacetic acids (HAAs) cannot be

11 ignored. Chlorine dioxide (ClO2) has been adopted as an effective alternative disinfectant of

12 chlorine to reduce THMs and HAAs formation. In this work, the reaction kinetics and degradation

13 pathway of ANT with ClO2 were studied to investigate the feasibility of using ClO2 as oxidant to

14 degrade ANT. Experimental results demonstrated that ANT oxidation by ClO2 followed

15 second-order kinetics, and the second-order rate constant (kapp) was determined to be 4.8×10-1

16 M-1s-1 at neutral pH. Higher pH could accelerate the reaction when pH<9, while strong alkaline

17 environment (pH>9) might significantly slow down the oxidation process. Structural change

18 during the reaction was proposed with the assistance of fourier transform infrared spectroscopy

19 (FT-IR), C=C and C-N bond of ANT were vulnerable under electrophilic attack of ClO2.

20 Degradation pathways of ANT with ClO2 were suggested based on the main intermediate products.

21 ANT was firstly transformed into ANT-Cl through single-electron-transfer (SET) and substitution

22 reaction. Further oxidation of this intermediate product involved ring-opening reaction and

23 de-carbonyl reaction.

24 Key words: antipyrine; chlorine dioxide; reaction kinetics; degradation pathway

25 1. Introduction
26 Pharmaceuticals are receiving increasing attention as potential bioactive chemicals in aquatic


Corresponding author. Address: Beijing Forestry University, No. 35 Tsinghua East Road, Beijing 100083, PR
China. Tel: +86 010 62336528; Fax: +86 010 62336900. E-mail address: zhangliqiu@163.com
27 environment since 1970s [1-6]. Under the assistance of new analytical techniques such as liquid

28 chromatography-mass spectra-mass spectra (LC-MS/MS), pharmaceuticals have been frequently

29 detected in treated sewage, surface and ground water in the range of ng/L to μg/L [7-11]. Inability

30 of traditional drinking water treatment process to completely remove pharmaceutical ingredients

31 makes it a widely concerned threaten to public health and ecological environment [12-15].

32 Antipyrine (ANT, phenazone, CAS: 60-80-0) is known as an anti-inflammatory analgesic and

33 is widely used to relieve headache, fever and general pain in medical treatment [16]. Due to the

34 relatively low absorption in human body, unignorable amounts of ANT are discharged into the

35 aquatic environment through municipal drainage system. It has been investigated that about

36 0.05-0.25 μg/L of ANT and its metabolites were detected in Germany municipal sewage effluents

37 and surface water. Chinese researchers have also found that the concentration of ANT in drinking

38 water plants was 1.34-2.22 ng/L [17]. Lungs as well as other target organs have been confirmed to

39 be vulnerable under ANT exposure, as a result, the potential risk of long-term ANT exposure

40 towards human body shall not be ignored though its concentration in aquatic environment is low

41 [18].

42 Due to the poor removal efficiency of pharmaceuticals by coagulation, sediment and

43 filtration (known as traditional drinking water treatment process), disinfection procedure seems to

44 be a feasible way to degrade ANT. Previous study has demonstrated that ANT can be removed by

45 most of frequently used disinfectants, such as chlorine, chlorine dioxide (ClO2), ozone (O3) and

46 UV [17-21].

47 ClO2 is an efficient alternative disinfectant of chlorine during water disinfection process for

48 its high biocidal efficacy at wide pH range and less potential of chlorine disinfection by-products

49 (DBPs) formation [22, 23]. Instead of chlorine substitution reaction, oxidation reaction is the

50 major approach to degrade pharmaceuticals, which can avoid the formation of two major kinds of

51 chlorine disinfection by-products, trihalomethane (THMs) and haloacetic acids (HAAs) [24-28].

52 Benefit from its high oxidation-reduction potential (ORP), ClO2 solution has stronger oxidation

53 capacity than chlorine solution. However, compared with other disinfectants, ANT oxidation

54 process by ClO2 was much slower, which is contradict to the fact that the ORP of ClO2 solution
55 (1.511V) is usually higher than chlorine solution(1.360V)[29]. The reason of the above-mentioned

56 phenomenon remain ambiguous and further study was needed.

57 ClO2 exhibits highly selective oxidation ability during the degradation reaction with

58 pharmaceuticals and personal care products (PPCPs). Studies have acknowledged that ClO2 has

59 relatively high oxidation efficiency with contaminants such as diclofenac [30], phenylurea [31],

60 fluoroquinolone [32], carbamazepine [33], tetracyclines [34], sulfamethoxazole [33], and some

61 selected amino acids [35], while other pharmaceuticals such as caffeine, clofibric acid, fenoprofen

62 and cyclophosphamide can rarely be degraded [22]. Degradation pathway of PPCPs by ClO2 can

63 generally be described as single electron transfer (SET) reaction, one electron of PPCPs is

64 transferred to ClO2 and results in ClO2-. Electron-rich moieties, such as amino group, conjugated

65 double bounds and piperazine ring are vulnerable under ClO2 attack. Under certain situation,

66 ClO2- can serve as oxidant in further degradation process of pharmaceuticals. However,

67 intermediate products and degradation pathways of ANT by ClO2 were still unknown.

68 The primary aims of our study were (1) to investigate the reaction kinetics of ANT oxidation

69 by ClO2, (2) to explain the different reactivity of ANT with ClO2 and free chlorine, and (3) to

70 explore the degradation pathway of ANT with ClO2.

71 2. Materials and Methods

72 2.1. Chemicals and reagents
73 ANT (purity>99%) was purchased from WAKO (Japan), stock solution of ANT (2.5 mM)

74 was prepared with ultrapure water and was then kept in dark and stored in 4℃. All other reagents

75 were purchased and used without further purification, such as NaClO2 (95%, AR), Na2S2O3 (98%

76 AR), NaOH (99% AR), H2SO4 (99% AR), Na2SO4 (99% AR), phosphate (95% AR), etc. Methanol,

77 acetonitrile, acetic acid and dichloromethane were HPLC grade (Fisher Scientific). All solutions

78 were prepared with ultrapure water (18 MΩ·cm) from a Water Purification System (ELGA

79 Purelab Classic, Veolia).

80 ClO2 was generated in a gas washing bottle by slowly dripping diluted H2SO4 into 500 mL of

81 400 mM NaClO2 solution. N2 stream (0.1 bar pressure) was used to blow off and carry the

82 generated gaseous ClO2 out of the washing bottle. In order to diminish the interference of chlorine
 83 in the experiment, N2 stream was pumped through a scrubber with saturated NaClO2 solution to

84 convert elemental chlorine in the gas stream to ClO2 and absorb other impurities [32, 36]. Finally,

85 the purified ClO2 was dissolved into ultrapure water and stored in brown bottle at 4℃ to prevent

86 self-decomposition.

87 2.2. ClO2 oxidation experiments

88 Batch experiments were conducted in 250 ml amber borosilicate bottles with glass stoppers

89 under magnetic stirring at room temperature (25±1℃). ANT stock solution (2.5 mM) was diluted

90 by ultrapure water to reach the initial reaction concentration (100 μM). Different dosage of ClO2

91 (from 5 mM to 15 mM) was added into the bottles to initiate the reaction. Phosphate buffer

92 solution was used to control pH at 7.00, the change between initial and final pH value was less

93 than ±0.10. ANT hydrolysis is extremely weak and can be considered to be negligible (less than 1%

94 consumption) within 6 hours under a wide range of pH (2-10).

95 To prevent ClO2 from volatilization, the volume of each reaction was 250 ml to keep a

96 smaller headspace of amber bottles. Reaction time was determined by preliminary experiments,

97 which ranged from 7 minutes to 30 minutes. At each sampling time, 2 ml sample was obtained by

98 a syringe and quenched with 100 μL sodium thiosulfate (0.3 g/L). All samples were filtrated by a

99 0.22 μM membrane and were further analyzed to obtain the residual concentration of ANT.

100 2.3. Analytical methods

101 ANT was analyzed by an ultra-performance liquid chromatograms system (UPLC 1260,

102 Agilient, USA) which includes a quatpump, liquid sampler, thermoregulation column

103 compartment and variable wavelength UV detector. A sample volume of 5 μL was injected onto a

104 Poroshell 120 EC-C18 column (4.6×50 mm, 2.7 Micron, Agilent, USA). The column was

105 maintained at 30℃ with a flow rate of 1.0 mL/min. The composition of the mobile phase was 10%

106 acetonitrile, 5% methanol and 85% acetic acid (0.02 vol.%, pH=4). ANT was detected at 242 nm

107 with an isocratic flow for 10 minutes. The limit of quantitation for ANT was approximately 5

108 ng/L.

109 High concentration ClO2 was analyzed by continuous iodometric method, sodium thiosulfate

110 and phosphate buffer were conducted to titrate ClO2, Cl2, ClO2- and ClO3- under different pH

111 conditions, respectively [36]. Freshly prepared ClO2 stock solution was analyzed, confirming that
112 the total concentration of Cl2 together with other impurities was less than 1% of the ClO2

113 concentration, and the concentration of ClO2 stock solution was 337 mM. The decay rate of ClO2

114 under storage condition was less than 10% within 4 months. Low concentration ClO2 was

115 analyzed by UV-spectrophotometry method, diluted ClO2 stock solution was prepared to plot the

116 standard curve, the relationship between ClO2 concentration (mg/L) and absorbance at 430 nm can

117 be described as A=1.5×10-5×[ClO2]-3.0×10-4, R2=0.9997.

118 2.4. Products identification

119 Oxidation experiment with higher initial ANT concentration (1 mM) was conducted to

120 explore the degradation products and pathway of ANT. The molar ratios of [ANT]0:[ClO2]0 ranged

121 from 10:1 to 1:10, the reaction time was set to be 72 hours, which was more than 20 times of the

122 reaction half-life, so the oxidation reaction should be regarded to be completed. The pH value and

123 temperature were maintained at 7.00 and 25±1℃, respectively. At each sampling time, 100 ml

124 sample was divided into two parts for fourier transform infrared spectroscopy (FT-IR) and gas

125 chromatography tandem mass spectrometry (GC-MS) analysis. In order to avoid interference of

126 sodium thiosulfate, samples were pre-treated and analyzed without quench [37]. Blank

127 experiments without ANT or ClO2 were also conducted simultaneously.

128 Pretreatment process of samples included freeze drying and re-dissolve, samples were

129 transferred into a sterilized plastic dishes and frozen within 8 hours in a ultra-low temperature

130 refrigerator at -80℃. The frozen samples were dehydrated by a vacuum freeze dryer (LGJ-12,

131 China) at -80℃ and 20 Pa. Dehydrated-solutes were obtained after 48 hours, one part of sample

132 was directly analyzed by FT-IR while the other part of sample was re-dissolved by 2 ml methanol

133 for GC-MS measurement.

134 An Agilent 7890 gas chromatograph with an Agilent 5975C MSD mass spectrometer

135 (GC-MSD) was used to identify ANT oxidation products. The capillary column was a DB-5ms

136 (30m×0.25mm×0.25 μm film thickness, 5% phenyl methylpolysiloxane, Agilent). 1 μL sample

137 was injected into GC-MS in splitless mode at inlet temperature 280℃, 99.99% helium was used as

138 carrier gas and maintained at flow rate of 0.8 mL/min. The column temperature was programmed

139 as follows: rise from 60℃ (1 min) to 160℃ (1 min) at 8℃/min, then to 290℃ (1 min) at 5℃
140 /min. The temperature of MS interface, MS ion source and quadrupole were maintained at 280℃,

141 230℃ and 150℃, respectively. Qualitative analysis was conducted at SCAN mode.

142 3. Results and Discussion

143 3.1. Reaction kinetics of ANT degradation by ClO2
144 Batch experiments were conducted under pseudo-first-order condition to investigate the

145 reaction rate and reaction order. The rate expression for the reaction between ClO2 and ANT can

146 be described as Eq. (1), where [ANT] and [ClO2] stand for the concentrations of ANT and ClO2

147 respectively, and kapp is the second-order reaction rate constant [22]. Previous studies have pointed

148 out that the degradation reaction between pharmaceuticals and ClO2 can be described as

149 second-order kinetics, each order for pharmaceuticals and ClO2, i.e. m=n=1 [36].

150 -d[ANT]/dt=kapp[ANT]m[ClO2]n (1)

151 ClO2 has relatively high reactivity with phenazone derivatives, such as ANT and aminopyrine

152 (AMP). Preliminary experiments showed that, under experimental conditions of 25.0±1℃,

153 pH=7.00, initial ANT concentration 100 μM, initial ClO2 dosage 10 mM, it required less than 20

154 minutes for the residual ANT to reach LOQ. Meanwhile, negligible consumption of ClO2 was

155 observed during the experiment time by UV-Vis spectrophotometer, which confirmed that the

156 change of ClO2 concentration has neglect impact on reaction rate. Thus the equation above can be

157 rewritten as

158 -d[ANT]/dt=kobs[ANT]m (2)

159 kobs=kapp[ClO2]n (3)

160 At each sampling time of the batch experiments, sample was quenched and analyzed by

161 UPLC to determine the residual concentration of ANT. By plotting natural logarithm of

162 normalized concentration of ANT versus time (Fig. 1) and kobs versus ClO2 dosage (Fig. 1 Insert),

163 kobs and kapp can be obtained respectively.

164
165 Fig. 1. Pseudo-first-order kinetic plot at different ClO2 dosage (Insert: kobs at different ClO2 dosage)

166 [ANT]0=100 μM, pH=7.00±0.10

167 Based on pseudo-first-order kinetics, natural logarithm of ANT normalized concentration

168 showed good correlation with sampling time, indicating that the reaction satisfied first-order to

169 ANT, i.e. m=1 (Fig. 1). For the constant initial concentration of ANT (100 μM), kobs appeared to

170 be increasing linearly while ClO2 dosage increased from 5.0 mM (7.0×10-2 min-1) to 15.0 mM

171 (3.6×10-1 min-1), which indicated that the reaction was also first-order to ClO2, i.e. n=1 (Fig. 1

172 Insert). Therefore, the ANT degradation should be summarized as the following second-order

173 kinetic model and kapp can be calculated to be 29 M-1min-1, or 4.8×10-1 M-1s-1.

174 -d[ANT]/dt=kapp[ANT][ClO2] (5)

175 Compared to other disinfectants such as chlorine (kapp=1.0×104 M-1s-1) and ozone

176 (kapp=6.5×104 M-1s-1), ClO2 exhibited an extremely low reaction rate with ANT [16, 19]. This

177 phenomenon was obviously conflict with the fact that the ORP of ClO2 solution is higher than

178 chlorine solution, which will be discussed latter.

179 3.2. Effects of pH

180 The pH of the solution may have different effect on the degradation process of

181 pharmaceuticals by ClO2. Pharmaceuticals with amino groups, like phenylurea, have high

182 reactivity at alkaline conditions [31]. Lower pH is in favor of the degradation process of
183 polycyclic aromatic hydrocarbons (PAHs) and carbamazepine. In other cases, such as diclofenac,

184 neither acidic nor alkaline condition can accelerate the reaction [36].

185 Experiments were conducted to determine the effects of pH on the reaction rate. Fig. 2

186 indicated that pH has slight effects on ANT (100 μM) oxidation of by ClO2 (10 mM). While pH

187 value raised from 4.21 to 8.10, kobs slowly increased from 2.0×10-1 min-1 to 2.7×10-1 min-1,

188 confirming a higher reactivity at slight alkaline condition. However, when pH was higher than

189 9.69, kobs steeply nosedived to 1.7×10-1 min-1, which is even lower than the acidic condition. The

190 kobs value at each pH could be readily calculated according to Eq. (2).

191
192 Fig. 2. kobs of ANT degradation process at different pH value.

193 [ANT]0=100 μM, [ClO2]0=10mM

194 Table 1

195 Kinetic parameters of ANT degradation at different ClO2 dosage and pH value. [ANT]0=100 μM

[ClO2]0(mM) pH kobs(min-1) T1/2(min) R2

5.0 7.04 7.0×10-2 9.9 0.9935

7.5 7.04 1.3×10-1 5.3 0.9966

10.0 7.04 2.2×10-1 3.2 0.9981

12.5 7.04 2.8×10-1 2.5 0.9975

15.0 7.04 3.6×10-1 1.9 0.9976

10.0 4.21 2.0×10-1 3.5 0.9970

 10.0 5.87 2.2×10-1 3.2 0.9964

10.0 8.10 2.7×10-1 2.6 0.9977

10.0 9.69 1.7×10-1 4.1 0.9914

196 It can be inferred that reaction order remained unchanged within pH range of 4-10, because

197 the existence form of ANT (pKa=1.4) and ClO2 were stable. According to its electrochemical

198 properties, higher pH value results in higher ORP of ClO2 solution, makes it easier to oxidize

199 substrate, which explains the high reactivity at alkaline condition[29, 38]. However, ClO2 may

200 involve in a disproportionation reaction as Eq. (6) when pH is higher than 9, causing a sharp

201 decline of ClO2 concentration, which is adverse to the degradation process of ANT [22, 38].

202 2ClO2+2OH-→ClO2-+ClO3- (pH>9) (6)

203 3.3. Structural change during ANT degradation

204 Transformation of major chemical groups and other structural components were revealed by

205 FT-IR spectra, as shown in Fig. 3. In the original spectrum of blank ANT without ClO2 oxidation,

206 absorption peak at 3091 cm-1 represented unsaturated C-H stretching vibrations such as benzene

207 ring, while peaks at 2966 cm-1 to 2870 cm-1 represented saturated C-H stretching vibrations of

208 methyl on pyrazolone ring. In area of double bond stretching vibrations, absorptions of both C=O

209 and C=C were recognized, meanwhile, N-N stretching vibrations were observed at 1666 cm-1.

210 Absorption peaks at 1580 cm-1 and 1456 cm-1 represented the vibration of aromatic ring skeleton.

211 Two types of C-H absorption of methyl in different chemical environment were also perceived,

212 1456 cm-1 and 1427 cm-1 peaks represent methyl connecting to a nitrogen atom, 1367 cm-1 peaks

213 possibly stand for methyl connecting to an ethylene. In fingerprint region, only 813 cm-1 peaks

214 could be clearly distinguished, which indicated the out-of-plane wagging vibration of C-H on

215 tri-substituted C=C.

216 By comparing shift and transformation of characteristic absorptions between ANT and its

217 oxidation products, vulnerable chemical groups under ClO2 attack could be recognized, changes in

218 compound structure could be speculated, as shown in Fig. 4. As the initial concentration ratio of

219 ANT to ClO2 changed from 1:10 to 1:100, no significant variations were detected with

220 characteristic peaks of aromatic ring skeleton (1580 cm-1 and 1456 cm-1) while the peak at 3091
221 cm-1 was slightly weakened, which indicated substitution on benzene ring but denied ring-opening

222 reaction of it. Neither N-N (1666 cm-1) nor N-CH3 (1456 cm-1 and 1427 cm-1) were changed,

223 indicating that those groups showed poor reactivity with ClO2. Furthermore, peaks at 1652 cm-1,

224 1456 cm-1 and 1367 cm-1 were slightly weakened, which confirmed the cleavage of C=C bond.

225 Notably, strong absorption peaks at 1417 cm-1 and 1359 cm-1 emerged, those peaks stands for C-H

226 bonds of acetyl, meanwhile, peak for carbonyl group (1592 cm-1) strengthened slightly, which

227 indicated that while carbonyl group on C5 was removed, a new carbonyl group created on C3.

228 Evidences of chlorine atom substitution were also discovered, for peaks of C-H on tri-substituted

229 C=C (813 cm-1) disappeared and –Cl (651 cm-1) emerged.

230
231 Fig. 3. Infrared spectrum analysis of ANT degradation products.

232 [ANT]0=100 μM, pH=7.00±0.10

233
234 Fig. 4. Characteristic absorption bonds and vulnerable chemical groups of ANT under ClO2 attack

235 3.4. Identification of ANT degradation products

236 Preliminary experiment ([ANT]0=1 mM, [ClO2]0=15 mM, pH=7.04) had revealed that the

237 oxidation process of ANT by ClO2 involved at least five main degradation products (P1-P5), as
238 shown in Fig. 5. The vertical axis of Fig. 5, Area/AreaANT,0 refers to the HPLC signals of each

239 products normalized to the initial ANT signal, which could be calculated by the ratio of their peak

240 area. For different products, the concentration variation patterns were distinctive during

241 degradation process. At the beginning of the reaction, the concentration of P1 and P2 increased

242 rapidly while ANT concentration dropped under the minimum detection limit of UPLC after 20

243 minutes. The concentration of P1 gradually dropped down, while the concentration of P3, P4 and

244 P5 slowly increased, which indicated that P1 might be an intermediate product and could be

245 further transformed into P2-P5.

246
247 Fig. 5. Variation pattern of different degradation products.

248 [ANT]0=10 μM, [ClO2]0=1 mM, pH=7.00±0.10

249 All the degradation products were identified by an Agilent GC-MSD system, as shown in Fig.

250 6 and Fig. 7. P1 (tR=24.7 min, m/z=MANT+34) was found in the reaction system when

251 [ANT]0:[ClO2]0=1:2, 1:1 and 1:5, the isotope abundance ratio of molecular ion peak was

252 222:223:224=100:18:32, which indicated that it was the single-chlorine-substituted product of

253 ANT, its formula (C11H11N2OCl) and structure could be confirmed with the help of NIST

254 Database on Aglient MS Workstation. P2 (tR=16.4min, m/z=MANT-24) was found when

255 [ANT]0:[ClO2]0=1:1, 1:5 and 1:10, possible candidate was C9H12N2O. P3 (tR=12.6min,

256 m/z=MANT-26), P4 (tR=22.9min, m/z=MANT+58) and P5 (tR=19.0min, m/z=MANT+8) were

257 detected when [ANT]0:[ClO2]0=1:5 and 1:10, which means they could only be generated when

258 ClO2 was excess. The isotope abundance ratio of P4 was 246:247:248=100:11:63, suggesting the
259 existence of two chlorine ion, which may be substituent groups of benzene ring, the formula

260 should be C10H12N2OCl2. The degree of unsaturation of P3 (C10H14N2) was calculated to be five,

261 suggesting the existence of alkylene radical. P5 (C10H13N2Cl) was confirmed to be a

262 single-chlorine-substitute product and the chlorine substituent group was attached to the benzene

263 ring.

264

265

266

267
268 Fig. 6. Total ion chromatograms of degradation products.

269 [ANT]0 =1 mM, [ANT]0:[ClO2]0=2:1, 1:1, 1:5 and 1:10

270

271

272

273

274
275

276
277 Fig. 7. Mass spectrograms of main degradation products

278 3.5. Degradation pathway of ANT

279 Degradation pathway of ANT was proposed according to the results of GC-MS and FT-IR, as

280 shown in Fig. 8.

281
282 Fig. 8. Proposed degradation pathway of ANT with ClO2

283 According to previous study on degradation pathways of ANT with O3[19, 21] and Cl2 [17],

284 N1 atom on pyrazolone ring showed negligible reactivity with ClO2, while N2 atom was
285 vulnerable under electrophilic reaction of ClO2 for its high electron cloud density. The first step of

286 ANT oxidation involved a single-electron-transfer reaction, lone pair electrons of N2 atom was

287 attacked by ClO2 to form ClO2- and ANT·+ radical [30, 39]. During the oxidation process, ClO2

288 was sequentially reduced to ClO, HClO and Cl- [40-42], and substitution reaction happened

289 between HClO with ANT·+ radical to form ANT-Cl.

290 Further degradation of ANT-Cl played an important role in oxidation process and formed

291 P2-P5, the degradation process could also be classified as single-electron-transfer reaction of ClO2,

292 main active sites included C-N bond and C=C double bond of pyrazolone ring [43, 44]. Electron

293 cloud density of C=C was strengthened after introduced a chlorine atom, for its electron donating

294 conjugative effect was relatively stronger than electrophilic inductive effect, which significantly

295 increased the trends of double bond cleavage under electrophilic attack of ClO2 [45, 46]. Another

296 vulnerable chemical bond was C-N bond connecting to N1 atom, the carbonyl group was removed

297 and formed P2 while N-N bond maintained intact, which is distinguished from chlorination and

298 photo-degradation of ANT [18]. According to the discussion in Fig. 5, P1 could be further

299 degraded into P3, P4 and P5, the process includes the removal of carbonyl group by excess ClO2

300 and substitution on benzene ring.

301 The above-mentioned degradation pathway provided a credible explanation about the reason

302 why ClO2 performed lower reactivity with ANT than chlorine and O3. Due to the p-π and π-π

303 conjugated system, the electron cloud density of possible reactive moieties such as N atoms and

304 C=C double bond were lower. ANT provides few active sites and could hardly lead to

305 electrophilic attack unless electron donating substitute groups change electronic cloud distribution.

306 Although the ORP of ClO2 solution was higher than chlorine solution, it cannot find the active

307 moieties that react with ClO2, therefore, the reaction rate of ClO2 was even lower than that of

308 chlorine[17].
309
310 Fig. 9. Comparison of ANT degradation pathways between different disinfectants

311 ANT exhibited different degradation pathways with chlorine[16, 17], ClO2 and ozone[19, 21],

312 as shown in Fig. 9. Substitution was the main reaction between chlorine and ANT, then the further

313 degradation included de-carbonyl reaction and cleavage of N-N bond. As for ClO2 and ozone,

314 direct oxidation was the major way to degrade ANT. Due to its structural characteristics, C=C

315 double bond was the most vulnerable chemical group under direct oxidation, the cleavage of this

316 bond was the beginning of ring-opening and de-carbonyl process. Different degradation pathways

317 may cause the variations in toxicity, to evaluate the toxicity changes during the reaction, further

318 study is needed.

319 4. Conclusions
320 The reaction kinetics and degradation pathway of ANT by ClO2 were investigated in this

321 study. ANT could be slowly but completely oxidized by ClO2, the degradation process followed

322 second order kinetics. The second-order rate constant (kapp) was calculated to be 4.8×10-1 M-1s-1 at

323 neutral pH. Higher pH value was slightly in favor of ANT degradation, but strong alkaline

324 environment might lead to disproportionate and significantly restrained the reaction. Five main

325 degradation products were discovered with the assistance of FT-IR and GC-MS. C=C bond and

326 N-C bond were recognized as main active moieties of ANT by ClO2. Compared with the rapid

327 chlorine-substitution reaction by free chlorine, ANT degradation process by ClO2 was much
328 slower due to its low reactivity under electrophilic attack. Oxidation process was slow until

329 hydrogen atom on C4 was substituted by chlorine atom (ANT-Cl), which changed electronic cloud

330 distribution of C=C bond. Degradation pathway of ANT was confirmed, including

331 single-electron-transfer reaction, substitution, ring-opening reaction and de-carbonyl reaction.

332 Acknowledgments
333 This work was supported by the Beijing Natural Science Foundation (No. 8152022) and the

334 National Nature Science Foundation of China (No. 51178046).

335 References
336 [1] T.A. Ternes, Occurrence of drugs in German sewage treatment plants and rivers, Water research 32
337 (1998) 3245-3260.
338 [2] N. Nakada, K. Komori, Y. Suzuki, C. Konishi, I. Houwa, H. Tanaka, Occurrence of 70
339 pharmaceutical and personal care products in Tone River basin in Japan, Water Sci Technol 56 (2007)
340 133-140.
341 [3] R.J. Zhang, J.H. Tang, J. Li, Z.N. Cheng, C. Chaemfa, D.Y. Liu, Q. Zheng, M.K. Song, C.L. Luo, G.
342 Zhang, Occurrence and risks of antibiotics in the coastal aquatic environment of the Yellow Sea, North
343 China, Science of the Total Environment 450 (2013) 197-204.
344 [4] L. Wang, G.G. Ying, J.L. Zhao, X.B. Yang, F. Chen, R. Tao, S. Liu, L.J. Zhou, Occurrence and risk
345 assessment of acidic pharmaceuticals in the Yellow River, Hai River and Liao River of north China,
346 Science of the Total Environment 408 (2010) 3139-3147.
347 [5] Q. Sui, J. Huang, S.B. Deng, G. Yu, Q. Fan, Occurrence and removal of pharmaceuticals, caffeine
348 and DEET in wastewater treatment plants of Beijing, China, Water research 44 (2010) 417-426.
349 [6] S.D. Kim, J. Cho, I.S. Kim, B.J. Vanderford, S.A. Snyder, Occurrence and removal of
350 pharmaceuticals and endocrine disruptors in South Korean surface, drinking, and waste waters, Water
351 research 41 (2007) 1013-1021.
352 [7] C.I. Kosma, D.A. Lambropoulou, T.A. Albanis, Investigation of PPCPs in wastewater treatment
353 plants in Greece: Occurrence, removal and environmental risk assessment, Science of the Total
354 Environment 466 (2014) 421-438.
355 [8] M. Park, D. Reckhow, M. Lavine, E. Rosenfeldt, B. Stanford, M.H. Park, Multivariate Analyses for
356 Monitoring EDCs and PPCPs in a Lake Water, Water Environ Res 86 (2014) 2233-2241.
357 [9] G.H. Dai, J. Huang, W.W. Chen, B. Wang, G. Yu, S.B. Deng, Major Pharmaceuticals and Personal
358 Care Products (PPCPs) in Wastewater Treatment Plant and Receiving Water in Beijing, China, and
359 Associated Ecological Risks, B Environ Contam Tox 92 (2014) 655-661.
360 [10] C. Tixier, H.P. Singer, S. Oellers, S.R. Muller, Occurrence and fate of carbamazepine, clofibric
361 acid, diclofenac, ibuprofen, ketoprofen, and naproxen in surface waters, Environ Sci Technol 37 (2003)
362 1061-1068.
363 [11] C. Alfonsin, A. Hospido, F. Omil, M.T. Moreira, G. Feijoo, PPCPs in wastewater - Update and
364 calculation of characterization factors for their inclusion in LCA studies, J Clean Prod 83 (2014)
365 245-255.
366 [12] X. Zhao, Z.L. Chen, X.C. Wang, J.M. Shen, H. Xu, PPCPs removal by aerobic granular sludge
367 membrane bioreactor, Appl Microbiol Biot 98 (2014) 9843-9848.
368 [13] M. Beretta, V. Britto, T.M. Tavares, S.M.T. da Silva, A.L. Pletsch, Occurrence of pharmaceutical
369 and personal care products (PPCPs) in marine sediments in the Todos os Santos Bay and the north
370 coast of Salvador, Bahia, Brazil, J Soil Sediment 14 (2014) 1278-1286.
371 [14] W.A. Mitch, D.L. Sedlak, Characterization and fate of N-nitrosodimethylamine precursors in
372 municipal wastewater treatment plants, Environ Sci Technol 38 (2004) 1445-1454.
373 [15] A.J. Watkinson, E.J. Murby, D.W. Kolpin, S.D. Costanzo, The occurrence of antibiotics in an
374 urban watershed: From wastewater to drinking water, Science of the Total Environment 407 (2009)
375 2711-2723.
376 [16] M.Q. Cai, L. Feng, J. Jiang, F. Qi, L.Q. Zhang, Reaction kinetics and transformation of antipyrine
377 chlorination with free chlorine, Water research 47 (2013) 2830-2842.
378 [17] M. Cai, L. Zhang, F. Qi, L. Feng, Influencing factors and degradation products of antipyrine
379 chlorination in water with free chlorine, Journal of Environmental Sciences 25 (2013) 77-84.
380 [18] C. Tan, N. Gao, Y. Deng, Y. Zhang, M. Sui, J. Deng, S. Zhou, Degradation of antipyrine by UV,
381 UV/H(2)O(2) and UV/PS, Journal of hazardous materials 260 (2013) 1008-1016.
382 [19] M. Favier, R. Dewil, K. Van Eyck, A. Van Schepdael, D. Cabooter, High-resolution MS and MS(n)
383 investigation of ozone oxidation products from phenazone-type pharmaceuticals and metabolites,
384 Chemosphere 136 (2015) 32-41.
385 [20] A. Duran, J.M. Monteagudo, I. Sanmartin, A. Carrasco, Solar photo-Fenton mineralization of
386 antipyrine in aqueous solution, Journal of environmental management 130 (2013) 64-71.
387 [21] H.F. Miao, M. Cao, D.Y. Xu, H.Y. Ren, M.X. Zhao, Z.X. Huang, W.Q. Ruan, Degradation of
388 phenazone in aqueous solution with ozone: influencing factors and degradation pathways,
389 Chemosphere 119 (2015) 326-333.
390 [22] M.M. Huber, S. Korhonen, T.A. Ternes, U. von Gunten, Oxidation of pharmaceuticals during
391 water treatment with chlorine dioxide, Water research 39 (2005) 3607-3617.
392 [23] S. Navalon, M. Alvaro, H. Garcia, Reaction of chlorine dioxide with emergent water pollutants:
393 Product study of the reaction of three β-lactam antibiotics with ClO2, Water research 42 (2008)
394 1935-1942.
395 [24] G. Hey, R. Grabic, A. Ledin, J. la Cour Jansen, H.R. Andersen, Oxidation of pharmaceuticals by
396 chlorine dioxide in biologically treated wastewater, Chemical Engineering Journal 185-186 (2012)
397 236-242.
398 [25] M.C. Dodd, C.H. Huang, Aqueous chlorination of the antibacterial agent trimethoprim: Reaction
399 kinetics and pathways, Water research 41 (2007) 647-655.
400 [26] S.D. Richardson, M.J. Plewa, E.D. Wagner, R. Schoeny, D.M. DeMarini, Occurrence, genotoxicity,
401 and carcinogenicity of regulated and emerging disinfection by-products in drinking water: A review
402 and roadmap for research, Mutation Research/Reviews in Mutation Research 636 (2007) 178-242.
403 [27] S.D. Richardson, Disinfection by-products and other emerging contaminants in drinking water,
404 TrAC Trends in Analytical Chemistry 22 (2003) 666-684.
405 [28] S.D. Richardson, A.D. Thruston, T.V. Caughran, P.H. Chen, T.W. Collette, K.M. Schenck, B.W.
406 Lykins, C. Rav-Acha, V. Glezer, Identification of New Drinking Water Disinfection by - Products from
407 Ozone, Chlorine Dioxide, Chloramine, and Chlorine, Water, Air, and Soil Pollution 123 (2000) 95-102.
408 [29] H. Roques, Chemical water treatment : principles and practice, VCH1996.
409 [30] Y. Wang, H. Liu, Y. Xie, T. Ni, G. Liu, Oxidative removal of diclofenac by chlorine dioxide:
410 Reaction kinetics and mechanism, Chemical Engineering Journal 279 (2015) 409-415.
411 [31] F.-X. Tian, B. Xu, T.-Y. Zhang, N.-Y. Gao, Degradation of phenylurea herbicides by chlorine
412 dioxide and formation of disinfection by-products during subsequent chlor(am)ination, Chemical
413 Engineering Journal 258 (2014) 210-217.
414 [32] P. Wang, Y.L. He, C.H. Huang, Oxidation of fluoroquinolone antibiotics and structurally related
415 amines by chlorine dioxide: Reaction kinetics, product and pathway evaluation, Water research 44
416 (2010) 5989-5998.
417 [33] Y. Lee, U. von Gunten, Oxidative transformation of micropollutants during municipal wastewater
418 treatment: Comparison of kinetic aspects of selective (chlorine, chlorine dioxide, ferrateVI, and ozone)
419 and non-selective oxidants (hydroxyl radical), Water research 44 (2010) 555-566.
420 [34] P. Wang, Y.L. He, C.H. Huang, Reactions of tetracycline antibiotics with chlorine dioxide and free
421 chlorine, Water research 45 (2011) 1838-1846.
422 [35] S. Navalon, M. Alvaro, H. Garcia, Chlorine dioxide reaction with selected amino acids in water,
423 Journal of hazardous materials 164 (2009) 1089-1097.
424 [36] Y. Wang, H. Liu, G. Liu, Y. Xie, X. Liu, Kinetics for diclofenac degradation by chlorine dioxide in
425 aqueous media: Influences of natural organic matter additives, Journal of the Taiwan Institute of
426 Chemical Engineers 56 (2015) 131-137.
427 [37] M.-Q. Cai, L.-Q. Zhang, L. Feng, Influencing factors and degradation behavior of
428 propyphenazone and aminopyrine by free chlorine oxidation, Chemical Engineering Journal 244 (2014)
429 188-194.
430 [38] M. Wu, J. Liu, S. You, L. Wang, J. Huang, Y. Tian, Effects and Kinetics of Chlorine Dioxide for
431 Removal of Benzo[a]pyrene in Water, Environ Eng Sci 29 (2012) 133-138.
432 [39] Y. Wang, H. Liu, G. Liu, Y. Xie, Oxidation of diclofenac by aqueous chlorine dioxide:
433 identification of major disinfection byproducts and toxicity evaluation, The Science of the total
434 environment 473-474 (2014) 437-445.
435 [40] A. Ison, I.N. Odeh, D.W. Margerum, Kinetics and mechanisms of chlorine dioxide and chlorite
436 oxidations of cysteine and glutathione, Inorg Chem 45 (2006) 8768-8775.
437 [41] C.W. Pan, D.M. Stanbury, Kinetics of the Initial Steps in the Aqueous Oxidation of Thiosulfate by
438 Chlorine Dioxide, J Phys Chem A 118 (2014) 6827-6831.
439 [42] A.K. Horvath, I.N. Nagypal, Kinetics and mechanism of the oxidation of sulfite by chlorine
440 dioxide in a slightly acidic medium, J Phys Chem A 110 (2006) 4753-4758.
441 [43] M.M. Huber, S. Canonica, G.Y. Park, U. Von Gunten, Oxidation of pharmaceuticals during
442 ozonation and advanced oxidation processes, Environ Sci Technol 37 (2003) 1016-1024.
443 [44] J.M. Monteagudo, A. Duran, J. Latorre, A.J. Exposito, Application of activated persulfate for
444 removal of intermediates from antipyrine wastewater degradation refractory towards hydroxyl radical,
445 Journal of hazardous materials 306 (2016) 77-86.
446 [45] J. Wenk, M. Aeschbacher, E. Salhi, S. Canonica, U. von Gunten, M. Sander, Chemical Oxidation
447 of Dissolved Organic Matter by Chlorine Dioxide, Chlorine, And Ozone: Effects on Its Optical and
448 Antioxidant Properties, Environ Sci Technol 47 (2013) 11147-11156.
449 [46] K. Wakigawa, A. Gohda, S. Fukushima, T. Mori, T. Niidome, Y. Katayama, Rapid and selective
450 determination of free chlorine in aqueous solution using electrophilic addition to styrene by gas
451 chromatography/mass spectrometry, Talanta 103 (2013) 81-85.

452
 Highlights
 Antipyrine can be slowly but completely oxidized by chlorine dioxide.
 The degradation process followed second-order reaction kinetics.
 Higher pH (pH<9) was slightly in favor of the degradation process.
 ANT-Cl was the major intermediate products of ANT oxidation by ClO2.
 Single-electron-transfer reaction, substitution, and de-carbonyl reaction were the main
degradation pathways.