Degradation of Diclofenac Sodium by the UV/Chlorine Process: Reaction Mechanism, Inuencing Factors and Toxicity Evaluation

23 This study examined the reaction mechanism, influencing factors and toxicity of diclofenac 24 sodium (DS) degradation by ultraviolet (UV)/chlorine process. The UV/chlorine was capable of 25 eliminating DS from water. The DS degradation during the UV/chlorine process followed a 26 pseudo-first-order kinetic model that was influenced by chlorine dosage, solution pH, humic acid 27 and bicarbonate concentrations. The free chlorine affects not only DS elimination, but the 28 contribution of various active species as well. Increasing free chlorine dosage from 1 to 7 mg·L -1 29 increased the first-order rate constant of NaClO, ·OH and reactive chlorine species (RCS) from 30 0.00063, 0.00328 and 0.00203 min − 1 to 0.00233, 0.0101 and 0.0974 min − 1 , respectively, and 31 increased the contribution of RCS from 8.20% to 75.71%, while the contribution of UV, NaClO, 32 and ·OH were declined from 76.02%, 2.54% and 13.24% to 14.63%, 1.81%, and 7.85%, 33 respectively. The contribution of RCS became increasingly prominence with the increment of free 34 chlorine dosage. The k obs,UV/chlorine,DS value decreased from 0.0797 to 0.0445 min -1 as pH increased 35 from 5.0 to 8.0. The presence of bicarbonate and natural organic matter both exerted an inhibitory 36 effect on DS degradation. Eleven intermediate products were identified and the degradation 37 pathway involved C-N cleavage, condensation, hydroxylation, and decarboxylation was proposed. 38 The UV/chlorine process effectively reduced acute toxicity and was superior to chlorination. The 39 genotoxicity induced by a chlorinated solution treated by the UV/chlorine process exhibited 40 negative genotoxicity. These results show that the UV/chlorine process is capable for the 41 degradation and detoxification of DS.


Abstract 23
This study examined the reaction mechanism, influencing factors and toxicity of diclofenac 24 sodium (DS) degradation by ultraviolet (UV)/chlorine process. The UV/chlorine was capable of 25 eliminating DS from water. The DS degradation during the UV/chlorine process followed a 26 pseudo-first-order kinetic model that was influenced by chlorine dosage, solution pH, humic acid 27 and bicarbonate concentrations. The free chlorine affects not only DS elimination, but the 28 contribution of various active species as well. Increasing free chlorine dosage from 1 to 7 mg·L -1 29 increased the first-order rate constant of NaClO, ·OH and reactive chlorine species (RCS) from

Experimental setup
108 All experiments were conducted in a 1-L photochemical reactor (Fig. S1). The ultraviolet 109 radiation was provided by a 4 W low-pressure mercury lamp (working at 254 nm). The incident 110 light irradiance was 3.0 μW·cm -2 at 17.5 cm from the UV lamp tube, as measured by a UV 111 radiometer (TN-2365, Taina Electronics Co. Ltd., Ningbo, China).

112
Unless specifically mentioned, experiments were performed in batch mode at a temperature 113 of 25 °C and the initial solution pH was 6.82. Solutoin pH was adjusted by adding 0.1 mol·L -1 114 NaOH or HNO3. A 1-L solution containing 550 μg·L -1 DS was first prepared in the beaker, then a 115 desired dosage of free chlorine (0, 1.0, 3.0, 5.0 or 7.0 mg·L -1 ) was added and the obtained solution 116 was mixed using the magnetic stirrer. The UV lamp was preheated for 2 min and then initiated the 117 reaction. 5 mL samples were taken out periodically, and the residual free chlorine was quenched 118 immediately by adding an excessive amount of sodium thiosulfate. 6 pure water and acetonitrile (50% pure water (0.05% trifluoroacetic acid) and 50% acetonitrile for 127 DS, 65% pure water and 35% acetonitrile for NB) at a flow rate of 1.0 mL/min. The inertsil 128 reverse-phase ODS-SP column (250 mm × 4.6 mm × 5 μm; GL Sciences, Inc., Tokyo, Japan)

147
The withdrawn samples were dechlorinated by sodium thiosulfate immediately for acute toxicity 7 testing. To provide suitable conditions for the bacteria, the pH was controlled between 6.0 and 8.0 149 and the temperature maintained at 15℃. The acute toxicity was expressed as the relative inhibition 150 ratio of luminescence (%).

151
The SOS/umu assay was performed according to a previously reported method with some 152 modifications (Elisabeth et al. 1998;Li et al. 2018;Oda et al. 1985). In the present study,

194
The values of kobs during treatment with free chlorine (3 mg·L -1 ) and UV were determinated 195 to be 0.00075 and 0.01883 min -1 , respectively, while a much higher kobs was obtained to be 0.0721 196 min -1 for the UV/chlorine process (3 mg·L -1 ). The UV/chlorine treatment resulted in a much 197 higher kobs which was 96.13 and 3.83 times higher than that for chlorine and UV alone, where kobs, UV/chlorine,DS (min -1 ) represents the determinated pseudo-first-order reaction rate constant 229 during the UV/chlorine process, and kobs,UV,DS (min -1 ) represents the pseudo-first-order reaction 230 rate constant for UV photodegradation, kobs, chlorine,DS (min -1 ) represents the pseudo-first-order 231 reaction rate constant for free chlorine oxidation, k•OH,DS (M -1 s -1 ) represents the second order 11 futher investigation. kobs,UV/chlorine,DS, kobs,UV,DS and kobs,chlorine,DS can be determinated through the 237 UV/chlorine treatment, direct UV photolysis and free chlorine oxidation, respectively. 238 Furthermore, k•OH,DS (M -1 s -1 ) can be determined by UV/H2O2 oxidation as described in Text S1.

352
To obtain the detailed information about DS degradation in the UV/chlorine process, the 353 intermediates were analyzed using liquid chromatography-mass spectrometry and eleven 354 intermediate products were identified (Table 1, Fig. S2 Table S1. According to the toxicity assessment levels of the globally harmonized 398 classification and labeling system for chemicals (Nations U 2019) (Table S2), the acute and ChV 399 toxicity levels of DS and its degradation intermediates were determined and presented in Fig. 7

416
As presented in Fig. 7(d), during UV irradiation, the acute toxicity (luminescence inhibition 19 rate) of the reaction solution initially decreased from 18% (0 min) to 9% (5 min), but afterwards 418 increased and reached a maximum inhibition rate of 26% at 30 min and finally decreased to 25% 419 at 75 min. The slight increase of acute toxicity suggested that UV alone was not effective for 420 reducing the acute toxicity of the reacted solution. This result was well consistent with previous 421 finding in which DS solution was treated by chlorine dioxide (Wang et al. 2014).

422
During the UV/chlorine process, the luminescence inhibition rates were higher than those 423 resulting from UV treatment alone in the early stage of the reaction (i.e., at sampling times of 5 424 min, 10 min and 30 min) except at 20 min (10%). However, the inhibition rate gradually 425 decreased from 31.1% to 8% with prolonging the reaction time from 30 min to 75 min. The 426 UV/chlorine treatment was superior to UV irradiation for the control of acute toxicity. As shown 427 in Fig. 7(d), the relative inhibition rate after the UV/chlorine treatment decreased to 8% at 90 min, 428 which was much lower than that of 25% achieved by UV irradiation alone.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.