Kinetics and pathways of the degradation of PPCPs by carbonate radicals in advanced oxidation processes

Article abstract of DOI:10.1016/j.watres.2020.116231

The carbonate radical (CO₃^??) is a typical secondary radical observed in engineering and natural aquatic systems. This study investigated the degradation kinetics of 20 pharmaceuticals and personal care products (PPCPs) by CO₃^?? and the transformation pathways of a typical PPCP (naproxen) that is susceptible to CO₃^??. CO₃^?? is highly selective for compounds containing aniline and phenolic hydroxyl groups as well as naphthalene rings, such as sulfamethoxazole, sulfamethazine, salbutamol, propranolol, naproxen, and macrolide antibiotics such as azithromycin, for which the second-order rate constants range from 5.6 × 10⁷ M^?1s^?1 to 2.96 × 10⁸ M^?1s^?1. A good linear relationship is observed between the natural logarithms of k_CO3^?? and the negative values of the Hammett Σσ_p⁺ constant for aromatic PPCPs, indicating that electron-donating groups promote the attack of benzene derivatives by CO₃^??. The contribution of CO₃^?? to naproxen degradation is significant in different processes such as UV/H₂O₂, UV/persulfate, UV/chlorine, and UV/monochloramine, in the presence of HCO₃^?, which compensates for the decreased contributions of primary radicals. In particular, the formation of CO₃^?? increases the first-order rate constant of naproxen by 127% in UV/monochloramine in the presence of 50 mM HCO₃^? compared to that without HCO₃^?. Natural organic matter (NOM) exerts a slight scavenging effect on CO₃^??, decreasing the inhibition effect of NOM on the degradation of naproxen by UV/H₂O₂ in the presence of HCO₃^?. The pathways involved in the transformation of naproxen by CO₃^?? include decarboxylation, hydroxylation, ketonization, demethylation and aldolization. In addition, the alteration of the genotoxicity during naproxen degradation by CO₃^?? was negligible.

Full text of DOI:10.1016/j.watres.2020.116231

Journal Pre-proof
Kinetics and pathways of the degradation of PPCPs by carbonate
radicals in advanced oxidation processes
Yujie Zhou , Chunyan Chen , Kaiheng Guo , Zihao Wu ,
Liping Wang , Zhechao Hua , Jingyun Fang
PII:
DOI:
Reference:
S0043-1354(20)30768-5
https://doi.org/10.1016/j.watres.2020.116231
WR 116231
To appear in:
Water Research
Received date:
Revised date:
Accepted date:
19 March 2020
9 July 2020
23 July 2020
Please cite this article as:
Yujie Zhou ,
Chunyan Chen ,
Kaiheng Guo ,
Zihao Wu ,
Liping Wang , Zhechao Hua , Jingyun Fang , Kinetics and pathways of the degradation of
PPCPs by carbonate radicals in advanced oxidation processes, Water Research (2020), doi:
https://doi.org/10.1016/j.watres.2020.116231
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. 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.
© 2020 Published by Elsevier Ltd.

Kinetics and pathways of the degradation of PPCPs by carbonate radicals in advanced
oxidation processes
Yujie Zhou, Chunyan Chen, Kaiheng Guo, Zihao Wu, Liping Wang, Zhechao Hua, Jingyun Fang*
Guangdong Provincial Key Laboratory of Environmental Pollution Control and Remediation
Technology, School of Environmental Science and Engineering, Sun Yat-Sen University, Guangzhou
510275, China
*Corresponding author.
E-mail: fangjy3@mail.sysu.edu.cn (Jingyun Fang)
1

Graphic abstract
Highlights
•−
 CO3 is highly selective for compounds containing aniline, phenolic hydroxyl groups and
naphthalene rings.
+
 A good linear relationship is observed between ln(kCO3•−) and the Hammett Σσp constant.
•−
 The formation of CO3 significantly increases naproxen degradation in UV/NH2Cl but not
UV/H2O2, UV/persulfate and UV/Cl2.
 The transformation products of naproxen by CO3•− are different from HO•.
 The degradation of naproxen by CO3•− does not increase genotoxicity.
2

Abstract
•−
The carbonate radical (CO3 ) is a typical secondary radical observed in engineering and natural
aquatic systems. This study investigated the degradation kinetics of 20 pharmaceuticals and personal
care products (PPCPs) by CO3•− and the transformation pathways of a typical PPCP (naproxen) that is
•−
•−
susceptible to CO3 . CO3 is highly selective for compounds containing aniline and phenolic
hydroxyl groups as well as naphthalene rings, such as sulfamethoxazole, sulfamethazine, salbutamol,
propranolol, naproxen, and macrolide antibiotics such as azithromycin, for which the second-order
rate constants range from 5.6 × 107 M-1s-1 to 2.96 × 108 M-1s-1. A good linear relationship is observed
+
between the natural logarithms of kCO3•− and the negative values of the Hammett Σσp constant for
aromatic PPCPs, indicating that electron-donating groups promote the attack of benzene derivatives by
•−
CO3 . The contribution of CO3•− to naproxen degradation is significant in different processes such as
−
UV/H2O2, UV/persulfate, UV/chlorine, and UV/monochloramine, in the presence of HCO3 , which
•−
compensates for the decreased contributions of primary radicals. In particular, the formation of CO3
increases the first-order rate constant of naproxen by 127% in UV/monochloramine in the presence
−
−
of 50 mM HCO3 compared to that without HCO3 . Natural organic matter (NOM) exerts a slight
•−
scavenging effect on CO3 , decreasing the inhibition effect of NOM on the degradation of naproxen
−
by UV/H2O2 in the presence of HCO3 . The pathways involved in the transformation of naproxen by
•−
CO3 include decarboxylation, hydroxylation, ketonization, demethylation and aldolization. In
addition, the alteration of the genotoxicity during naproxen degradation by CO3•− was negligible.
•−
Keywords: Advanced oxidation process (AOP), carbonate radical (CO3 ), pharmaceuticals and
personal care products (PPCPs), water treatment.
3

1. Introduction
Pharmaceuticals and personal care products (PPCPs) are causes for increasing concern due to
their frequent occurrence in aquatic environments at the ng L−1 to mg L−1 level (Duca and Boldescu,
2009). PPCPs are known to be released into aquatic environments through multiple pathways,
including domestic wastewater, hospital discharges, improper manufacturer disposal and wastewater
treatment plants (Leung et al., 2012; Liu and Wong, 2013). However, most PPCPs are not effectively
removed by conventional water and wastewater treatment processes, such as coagulation,
sedimentation, filtration and biological processes, and may be present in drinking water at a low
concentration range of ng-μg L-1 (Heberer, 2002; Luo et al., 2014).
Advanced oxidation processes (AOPs) effectively remove PPCPs by generating reactive radicals,
•−
•−
2−
such as the hydroxyl radical (HO•, E0 (HO•/H2O) = 1.9 - 2.7 V), sulfate radical (SO4 , E0 (SO4 /SO4 )
= 2.5 - 3.1 V), and chlorine radical (Cl•, E0 (Cl•/Cl−) = 2.47 V) (Buxton et al., 1988; Neta et al., 1977;
−
2−
Xiang et al., 2016). Due to the inevitable presence of HCO3 and CO3 in aquatic systems, the
•−
primary HO•, Cl• and SO4•− radicals are partially transformed to a secondary carbonate radical (CO3 )
through the processes described in eqs (1-3) (Buxton and Elliot, 1986; Huie and Clifton, 1990;
Mertens and Sonntag, 1995).
−
HCO3 /CO32− + HO• → CO3•− + H2O/OH−
(1)
(2)
(3)
kHCO /HO• = 8.5 × 106 M−1s−1/kCO /HO• = 3.9 × 108 M−1s−1
-
-
2
3
3
−
HCO3 /CO32− + Cl• → CO3•− + HCl/Cl−
kHCO /Cl• = 2.2 × 108 M−1s−1/kCO /Cl• = 5.0 × 108 M−1s−1
-
-
2
3
3
−
2−
HCO3 /CO32− + SO4•− → CO3•− + H2SO4/SO4
4

-
6
−1 −1
•− = 4.1 × 106 M−1s−1
-
•− = 9.1 × 10 M s /kCO
2
kHCO
4
4
/SO
/SO
3
3
The concentrations of CO3•− were reported to be 4 and 2 orders of magnitude higher than that of
•−
HO• and SO4 , respectively, in the UV/H2O2 and UV/persulfate (UV/PS) AOPs in the presence of
−
HCO3 /CO32− (Zhang et al., 2015). Meanwhile, in surface water, solar indirect photolysis systems
containing nitrate/nitrite and/or natural organic matter (NOM) produce CO3•− through the reactions of
−
HO• or excited triplet states of dissolved organic matter (3NOM*) and HCO3 /CO32−, for which the
steady-state concentrations were reported to be 2 orders of magnitude higher than that of HO•
(Canonica et al., 2005; Huang and Mabury, 2000; Lam et al., 2003; Sulzberger et al., 1997; Wu and
Linden, 2010). Therefore, CO3•− is widely present in engineering and natural water systems. However,
the specific roles of CO3•− in different processes are unclear and a comparison is needed.
CO3•− is a selective oxidant (E0 = 1.78 V at pH 7), which redox potential is lower than those of
•−
HO• and SO4•− (Neta et al., 1988). The second-order rate constants (k) of CO3 reacting with organic
compounds range from 102 - 109 M−1s−1 (Behar et al., 1970; Cope et al., 1973; Moore et al., 1977).
CO3•− was reported to react rapidly with compounds containing electron-rich groups, such as nitrogen-
and sulfur-containing substrates and aromatic compounds (Neta et al., 1988). For example, CO3•− was
reported to react with phenol and aniline with k of 2.2 × 107 M−1s−1 and 5.0 × 108 M−1s−1, respectively
•−
(Larson and Zepp, 1988; Moore et al., 1977). CO3 possesses high reaction rate constants for
different micropollutants, such as clenbuterol (5.2 × 108 M−1s−1), ketoprofen (3.9 × 108 M−1s−1),
paracetamol (1.7 × 108 M−1s−1), and oxytetracycline (2.9 × 108 M−1s−1) (Liu et al., 2016; Wols et al.,
•−
2015). Therefore, the roles of CO3 in the degradation of structurally diverse PPCPs differ, and the
relationship between reactivity and structure should be clarified.
5

•−
•−
Regarding the reaction mechanism of CO3 , CO3 was reported to react with organic matter
through electron transfer and hydrogen abstraction (Liu et al., 2018; Liu et al., 2016; Moore et al.,
•−
1977). CO3 reacts with phenol and bisphenol A via electron transfer to generate a phenolic radical
cation, which then further transforms to a phenoxyl radical (Gao et al., 2020; Moore et al., 1977).
Similarly, for tertiary amines, electron transfer from the N atom to CO3•− yields an amine radical cation
followed by proton elimination (Dell'Arciprete et al., 2012; Elango et al., 1984), while hydrogen
•−
abstraction yields N-centred radicals for primary amines (Liu et al., 2018). In the present study, CO3
was highly reactive towards compounds containing aniline and phenolic hydroxyl groups, as well as a
naphthalene ring. However, previous studies focused on the degradation pathways of phenols and
•−
amine derivatives, but the pathways underlying the transformation of the naphthalene ring by CO3
remain unknown.
Therefore, the purposes of this study are 1) to investigate the kinetics and structural relationships
•−
of a number of structurally diverse PPCPs degraded by CO3 , 2) to elucidate the transformation
pathways and alterations in the toxicity of a typical PPCP containing a naphthalene ring after a
•−
•−
reaction with CO3 ; and 3) to compare the roles of CO3 in different AOPs, such as UV/H2O2,
UV/PS, UV/chlorine (UV/Cl2), and UV/monochloramine (UV/NH2Cl). Twenty structurally diverse
PPCPs (Table S1) were selected for purpose (1), and naproxen (NPX) was selected as the target
compound for purposes (2) and (3). NPX was selected because it is widely used and frequently
•−
occurs in aquatic systems (Madikizela and Chimuka, 2016), and it reacts with CO3 , but the
transformation mechanism is unknown. Additionally, NPX exhibits adverse effects on aquatic
6

organisms with an EC50 of 21.2 µg L-1 to Vibrio fischeri (Balakrishna et al., 2017), and it may
increase the risk of heart disease and stroke even at trace levels (Manrique-Moreno et al., 2010).
2. Materials and methods
2.1. Materials
All solutions were prepared from reagent-grade chemicals and Milli-Q water (18.2 MΩ cm).
Sources of reagents are provided in Text S1 of the Supplementary Materials. The NH2Cl stock solution
was freshly prepared by slowly adding a sodium hypochlorite (NaOCl) solution to an ammonium
chloride (NH4Cl) solution at a mass ratio of Cl/N of 4:1 with rapid mixing for 40 min (Hand and
Margerum, 1983). The pH of the NH4Cl solution and NaOCl solution was adjusted to approximately
10 with a sodium hydroxide (NaOH) solution before mixing. The engineered bacterium Salmonella
typhimurium TA1535/pSK1002, which served for the umu test, was obtained from Professor Xin Yu,
Xiamen University, China.
2.2. Experimental procedure
All experiments were performed in a UV reactor as reported in Wu et al. (2019) (Text S1). The
reactor was operated in batch mode with rapid mixing provided in the bottom of the reactor, which
stirring rate was set at 200 r/min. The schematic diagram of the UV cylindrical reactor is shown in
Fig. S1.The UV photon flux (I0) entering the solution was determined to be 0.331 μE s−1 using
iodide/iodate chemical actinometry (Bolton and Linden, 2003), and the optical path length at 254 nm
7

was 3.28 cm, as determined by measuring the photolysis kinetics of dilute H2O2 (Fig. S2) (Garoma
and Gurol, 2005). The corresponding average UV fluence rate (Ep0) was about 0.75 mW cm−2.
The basic photochemical experiment was initiated by exposing a 700 mL solution containing 5
−
μM NPX and 50 mM HCO3 to UV irradiation and simultaneously dosing with 50 μM oxidant (i.e.,
H2O2, PS, Cl2, or NH2Cl). Samples were collected at different time intervals and immediately
quenched with sodium thiosulfate at a molar ratio of sodium thiosulfate to oxidant of 4:1. Control
tests of the degradation of NPX by UV photolysis and oxidant alone were also performed. The tests
of 20 PPCPs were conducted in a similar manner, but solutions were spiked with 1 μg/L of each PPCP.
−
The pH was buffered at 8.6 by HCO3 or phosphate buffer in all the tests. Note that, the
−
concentrations of HCO3 range from 38 to 575 mg/L in WWTP effluents and from 61 to 183 mg/L in
−
surface waters (Lai et al., 2017). Higher concentration of HCO3 (50 mM) was used in this study to
•−
create a system only containing CO3 , and to clarify the reaction mechanism of CO3•− with NPX. All
tests were performed at least twice, and the error bars in all the plots represent the maximum and
minimum of the experimental data of the results of the duplicate tests. In addition, NPX, nitrobenzene
(NB), benzoic acid (BA) were used as probe compounds to determine the steady-state concentrations
•−
•−
of CO3 , HO•, and SO4 , respectively, as described in the studies by Fang et al. (2014) and Yang et al.
(2014) (details shown in Text S2).
−
The initial concentrations of NPX, H2O2, and HCO3 were increased to 50 μM, 200 μM and 100
mM, respectively, to identify the degradation products and genotoxicity of NPX after the reaction with
•−
CO3 . Note that HO• was completely quenched and converted to CO3•− under this condition (Fig. S3).
The genotoxicity of NPX (50 μM) after UV/Cl2 (200 μM) treatment was also investigated to compare
8

•−
with that after CO3 treatment. After a certain reaction period, a 500 mL sample was collected and
immediately quenched with sodium thiosulfate, which was then concentrated to 2 mL through solid
phase extraction (SPE) using an HLB cartridge (200 mg, 6 mL, Waters) using a procedure described in
the literature (Michael et al., 2012).
2.3. Analytical methods
The pH was determined using a pH metre (FE20, Mettler Toledo). The concentrations of NaOCl
and NH2Cl were determined using DPD/FAS titration (APHA-AWWA-WEF, 2012). The
concentrations of NPX, NB and BA were analysed using a high-performance liquid chromatograph
(HPLC, Agilent 1260, USA) equipped with a C18 column (Poroshell, 4.6 × 50 mm, 2.7 μm) and a
diode array detector (DAD) (Pan et al., 2018). Ultra-high-performance liquid
chromatography-quadrupole-time of flight-mass spectrometry (UPLC−Q−TOF−MS) with an
electrospray ionization source (ESI) was used to identify the degradation products of NPX (details
shown in Text S3). Twenty PPCPs were quantified on an HPLC (Agilent 1290) connected with an
online SPE system (Spark, Agilent) and coupled with a triple quadrupole mass system with an
electrospray ionization source (AB SCIEX 5500) using the method reported in the study by Guo et al.
(2018).
•−
The genotoxicity of the products of NPX generated by CO3 was analysed using the
SOS/umu-test according to (ISO13829, 2000) and Lv et al. (2014). The umu test evaluates the
genotoxic activity of chemical compounds, using the expression of the SOS repair mechanism
9

associated with the umuC-gene. The detail method of the determination of genotoxicity was provided
in Text S4.
2.4. Determination of the second-order rate constants of CO3•− reacting with PPCPs
•−
The values of k for the reaction of CO3 with PPCPs were determined by competition kinetics
experiments using NPX as a reference compound, which reacts with CO3•− at 5.6 × 107 M−1s−1 (Wols et
al., 2015). The UV/H2O2 AOP was conducted at a H2O2 dosage of 50 μM in the presence of 50 mM
−
HCO3 to create an effective condition for CO3•− production. Note that, 99.7% HO• was converted to
•−
CO3 under this condition (Fig. S4), and the contributions of 0.3% HO• to PPCP degradation were
•−
less than 7.7% (Table S2). The degradation kinetics of PPCPs by CO3 all followed
•−
pseudo-first-order kinetics (Fig. S5). Then the values of k for the PPCPs reacting with CO3 were
calculated using eq (4):
N X
(4)
N X
•−
•−
,PPCP
•−
where kCO
and kCO ,NPX represent the k of CO3 reacting with the selected PPCPs and NPX,
3
3
respectively.
2.5. QSAR analysis
A quantitative structure-activity relationship (QSAR) was employed to explore the relationship
between the structures of aromatic PPCPs and their reactivity towards different radicals. Hammett σ+
constants were calculated to represent the electron-donating (large negative value) or
electron-withdrawing (large positive value) properties of substituents on aromatic compounds
10

(details shown in Text S5) (Hansch et al., 1991; Hansch et al., 1995; Lee and von Gunten, 2012), and
the k (ln k) of CO3•− reacting with PPCPs represented the reactivity.
2.6. Quantum chemistry calculation
All calculations for structural optimizations and molecular orbital distributions were performed
based on density functional theory at the B3LYP/6-31G (2df,p) level and the integral equation
formalism polarizable continuum model (IEFPCM) using the Gaussian 09 program (Frisch et al.,
2010). Structural optimization was conducted without any symmetry constraints, and the minima were
verified by a vibrational analysis. The molecular orbital distribution contained the highest occupied
molecular orbitals (HOMOs).
2.7 Kinetic modelling
The concentrations of different reactive species in UV/H2O2, UV/Cl2 and UV/NH2Cl AOPs
were simulated through kinetic modelling using Kintecus® software v5.50, which can calculate the
concentrations of different species based on all the reactions involved in a specific process and the
corresponding reaction rate constants (Kintecus, 2020). Details of the reactions are shown in Table
S3.
11

3. Results and discussion
•−
3.1. Degradation of 20 PPCPs by CO3
Fig. 1 compares the pseudo first-order rate constants (k′) of 20 PPCPs degraded by UV/H2O2 and
−
UV/H2O2/HCO3 in pure water at pH 8.6. The degradation of all the PPCPs was negligible in H2O2
alone (data not shown), while direct UV photolysis was not effective at degrading most PPCPs, except
for sulfamethoxazole (k' = 0.00113 s−1). All PPCPs were degraded quickly by the UV/H2O2 AOP with
−
k′ in the range of 0.00225 - 0.00527 s−1, which was mainly attributed to HO•. In UV/H2O2/HCO3 , the
degradation of some PPCPs was significantly inhibited, however, the degradation of other PPCPs was
inhibited less significantly or even increased. The specific k′ of HO• (k′HO•) was almost completely
−
•−
inhibited (99.7%) in UV/H2O2/HCO3 (Fig. S4), and the values of CO3 (k′CO •−) ranged from 0 -
3
0.00618 s−1.
To facilitate the discussion, the 20 PPCPs were divided into two groups according to the
•−
•−
contribution of CO3 to the degradation of PPCPs. (1) PPCPs for which CO3 had a negligible
contribution to their degradation were defined as group I (k′CO < 0.00022 s−1), including
•−
3
dimetridazole, metronidazole, ornidazole, tinidazole, ronidazole, bezafibrate, caffeine, atrazine,
nalidixic acid, carbamazepine, atenolol, venlafaxine and mecoprop. (2) PPCPs that showed strong
•−
reactivity with CO3 were designated as group II (k′CO = 0.00084 - 0.00618 s−1), including
•−
3
trimethoprim, sulfamethoxazole, propranolol, azithromycin, sulfamethazine, salbutamol and NPX.
Specifically, the k′ of trimethoprim, propranolol, azithromycin and NPX decreased by 77.3%, 21.3%,
12

46.8% and 69.8%, respectively, while the values of sulfamethoxazole, sulfamethazine and salbutamol
−
increased by 13%, 44.5% and 44.7%, respectively, in the presence of 50 mM HCO3 .
•−
The k values of CO3 for these 20 PPCPs were determined using competition kinetics
experiments (Fig.S6), and the k values were consistent with those available in literatures (Table S4).
In addition, we obtained the k of CO3•− reacting with the following PPCPs from Wols et al. (2015) and
Yan et al. (2019) for a more comprehensive comparison of the value of k of CO3•− reacting with organic
compounds: paracetamol, clenbuterol, sotalol, diclofenac, erythromycin A, gemfibrozil, clofibric acid,
and vanillic acid, p-coumaric acid. These organic compounds are divided into seven categories
according to their functions or functional groups (Fig. 2a): substituted phenols: salbutamol (1.51 × 108
M−1s−1), paracetamol (1.70 × 108 M−1s−1), vanillic acid (5.0 × 108 M−1s−1), and p-coumaric acid (4.60 ×
108 M−1s−1); naphthalene ring derivatives: propranolol (1.92 × 108 M−1s−1) and NPX (5.60 × 107
M−1s−1); aniline derivatives: sulfamethoxazole (2.48 × 108 M−1s−1), sulfamethazine (2.96 × 108 M−1s−1),
clenbuterol (5.20 × 108 M−1s−1), sotalol (2.20 × 108 M−1s−1), and diclofenac (7.80 × 107 M−1s−1);
amines (excluding aniline): trimethoprim (4.91 × 107 M−1s−1), carbamazepine (1.25 × 107 M−1s−1), and
atenolol (9.79 × 106 M−1s−1); N-heterocyclic carbenes: nitroimidazoles (5.09 – 9.11 × 106 M−1s−1),
atrazine (9.39 × 106 M−1s−1), nalidixic acid (7.49 × 106 M−1s−1), and caffeine (6.09 × 106 M−1s−1);
macrolide antibiotics: azithromycin (9.12 × 107 M−1s−1) and erythromycin A (8.0 × 107 M−1s−1); and
lipid regulators: bezafibrate (5.33 × 106 M−1s−1), gemfibrozil (4.10 × 106 M−1s−1), and clofibric acid
(1.0 × 107 M−1s−1). Among these seven groups, substituted phenols, naphthalene ring derivatives,
aniline derivatives, and macrolide antibiotics react faster with CO3•− (> 5.0 × 107 M−1s−1), while other
groups react slowly (< 5.0 × 107 M−1s−1). Thus, the aniline, phenolic hydroxyl group and naphthalene
13

•−
ring are more likely to react with CO3 , and the reaction rates of common amines are slower than
anilines. For example, the k of CO3•− reacting with trimethoprim was lower than those of the aniline
derivatives, even though it contains two amino groups.
Aprimary QSAR was established based on the assumption that the benzene ring was the main site
of attack to better explain the reactivity of HO• and CO3•− towards aromatic PPCPs. Several PPCPs in
each group were selected for the QSAR: atenolol, venlafaxine, bezafibrate and mecoprop in group I,
and trimethoprim and salbutamol in group II. In addition to the PPCPs listed above, six PPCPs with
•−
known kCO were also selected to increase the sample size: clenbuterol, diclofenac, paracetamol,
3
clofibric acid, gemfibrozil and sotalol (Wols et al., 2015). These PPCPs were selected due to their
•−
available Hammett σ+ constants. The QSARs between kCO and kHO• of the aromatic PPCPs and
3
•−
Hammett σ+ constants were established (Table S4 and Excel S1). The natural logarithms of kCO
3
(lnkCO •−) of the 12 PPCPs increased linearly with increasing Hammett σ+ constants, as shown in eq (5)
3
(Fig. 2b). Here, ∑σ+ represents the Hammett σ+ constants of the substituents on the aromatic
compounds.
lnkCO3•− = -3.91∑σ+ + 13.4
R2 = 0.8959
(5)
The reactivity of HO• was maintained at a constant value as the Hammett σ+ constants
decreased, likely due to the closeness of the values to the diffusion controlled limit, while the
•−
•−
reactivity of CO3 increased significantly, indicating that CO3 is more selective towards
•−
electron-rich compounds than HO•. The reactivity of CO3 with aromatic PPCPs depends on the
electron-donating properties of substituents on the aromatic ring.
14

3.2 Roles of CO3•− in different UV-based AOPs
•−
NPX was selected as the target compound to elucidate the roles of CO3 in different AOPs:
−
UV/H2O2, UV/Cl2, UV/NH2Cl and UV/PS AOPs. In the presence of 50 mM HCO3 , the overall k′ of
NPX degradation by the UV/H2O2, UV/PS and UV/Cl2 AOPs changed slightly, while the k′ by
−
UV/NH2Cl increased significantly by 127% compared to that without HCO3 (Fig. 3). The specific k′
•−
of CO3 is expected to significantly contribute to the overall k′, which would compensate for the
decrease in the contributions of primary radicals to NPX degradation. Notably, the values of k′ for
−
NPX degradation by UV and UV/HCO3 were identical.
−
•−
In the absence of HCO3 , k′ is attributable to HO• in UV/H2O2 and HO• and SO4 in UV/PS
(Xie et al., 2015). In UV/Cl2 and UV/NH2Cl, HO• and reactive chlorine species (RCS) are involved.
In particular, ClO• is present in the former reaction system, while reactive nitrogen species (RNS) are
−
•−
involved in the latter system (Wu et al., 2019; Fang et al., 2014). In the presence of HCO3 , CO3 is
−
•−
•−
produced in the reactions between HCO3 and HO•, SO4 , Cl•−, Cl2 , but ClO• was reported to be not
−
2−
reactive toward HCO3 by analogy to CO3 , which reacts with ClO• at a low k of 600 M−1s−1 (Wu et
al., 2017; NIST, 2020). The k for the reaction between CO3•− and NPX (5.60 × 107 M−1s−1) is 2 − 3
orders of magnitude lower than the k values for the reactions including HO• (8.61 × 109 M−1s−1), Cl•
(2.01 × 1010 M−1s−1) and SO4•− (5.64 × 109 M−1s−1) (Lei et al., 2019; Lian et al., 2017; Wols et al.,
•−
2015). However, the concentrations of CO3 in these AOPs may be much higher, as the lifetime of
CO3•− is longer than that HO• due to the much lower radical-radical recombination rate of CO3•− (k = 2
× 107 M−1s−1) than HO• (k = 5.5 × 109 M−1s−1) (Chen et al., 2019; Lai et al., 2017). By using a kinetic
•−
model, the concentrations of CO3 reached 2.2 × 10-11, 9.5 × 10-10 and 5.2 × 10-9 M in the UV/H2O2,
15

−
UV/Cl2 and UV/NH2Cl AOPs in presence of 50 mM HCO3 , respectively (Table S5), and that in
UV/PS was also determined to be 6.8 × 10-11 M by experiment (Figs S7 and S8).
•−
Interestingly, the highest concentration of CO3 was observed in UV/NH2Cl, which coincided
with the significant increase in k′ in UV/NH2Cl. Cl• was the primary radical produced from the
photolysis of NH2Cl. The scavenging rate constants of Cl• were calculated based on the k values of Cl•
−
reacting with H2O/OH−, HCO3 , NPX and oxidants such as NH2Cl and HOCl/OCl− and their
concentrations (Fig. S9). Cl• reacts with H2O and OH− at the scavenging rate constants of 2.5 × 105
and 7.2 × 104 s−1, respectively (Fig. S9) (Guo et al., 2018), to produce ClOH•− at pH 8.6. Then,
ClOH•− was transformed to HO• through a reversible reaction (eq (6)) (Grebel et al., 2010), with the
conversion yield of 67.1% as calculated using the kinetic model.
ClOH•− ↔ HO• + Cl−
k+ = 6.1 × 109 s−1/k- = 4.3 × 109 M−1s−1
(6)
•−
−
However, greater than 95.9% of Cl• would be converted to CO3 by 50 mM HCO3 , which
scavenging rate constant is 1.1 × 107 s−1 and is 2 orders of magnitude higher than H2O/OH−. Thus,
−
−
most of Cl• was converted to CO3•− through the direct reaction of HCO3 in the presence of HCO3 ,
−
and the conversion rate was higher than that of the conversion of Cl• to HO• in the absence of HCO3 .
The direct formation of CO3•− from Cl• inhibited the decrease of the oxidation capacity of Cl• to NPX
abatement when Cl• was converted to HO• with the low conversion yield of 67.1%, resulting in the
−
higher oxidation capacity for NPX abatement in UV/NH2Cl with HCO3 .
Although Cl• is also present in UV/Cl2, the trend is different from UV/NH2Cl. k′ by UV/Cl2
−
slightly changed in the presence of HCO3 . Cl• was primarily consumed by HOCl/OCl− at a
scavenging rate constant of 3.95 × 105 s−1 at pH 8.6 to directly form ClO•, a value that is higher than
16

the scavenging rate constant by H2O/OH− (Fig. S9). ClO• was the predominant radical produced
during NPX degradation in UV/Cl2 with k toward NPX of < 2.30 × 109 M−1s−1 (Guo et al., 2018). In
−
the presence of 50 mM HCO3 , Cl• in UV/Cl2 was still largely converted to ClO•, resulting in the
slightly decreased ClO• concentration from 3.1 × 10-10 M to 2.5 × 10-10 M (Table S5). Meanwhile,
•−
−
•−
CO3 can also be formed through the reactions of HCO3 with HO• and Cl2 with k values of 8.5 ×
106 M−1s−1 and 8.0 × 106 M−1s−1, respectively (Matthew and Anastasio, 2006; NIST, 2020). The
concentration of Cl• in UV/NH2Cl was 1.9 times higher than that in UV/Cl2, while those of HO• and
•−
•−
Cl2 were one order of magnitude higher (Table S5). Thus, the concentration of CO3 in
−
−
UV/NH2Cl/HCO3 was one order of magnitude higher than that in UV/Cl2/HCO3 .
3.3 Effect of water matrices
−
HCO3 concentrations exerted a slight effect on the degradation of NPX by the UV/H2O2 AOP
−
(Fig. 4a). The k′ only decreased by 2 − 11.6% in the presence of 5 - 100 mM HCO3 compared to
−
−
UV/H2O2 in the absence of HCO3 . Specifically, as the HCO3 concentration increased from 0 to 100
mM, k′HO• significantly decreased from 0.00133 s−1 to 0 at 30 mM (Figs 4a and S10), while k′CO
•−
3
increased from 0 to 0.00130 s−1. The modelled concentration of HO• significantly decreased from 1.0
−
× 10-13 M to 4.3 × 10-15 M as the HCO3 concentration increased from 0 to 100 mM in the UV/H2O2
•−
AOP, whereas the concentration of CO3 increased from 9.1 × 10-12 M to 2.3 × 10-11 M. The
•−
−
concentration of CO3 observed in the presence of HCO3 was 2 orders of magnitude higher than
−
that of HO• in the absence of HCO3 obtained from the kinetic model (Table S5). Meanwhile, k for
the reaction between CO3•− and NPX (5.6 × 107 M−1s−1) is also 2 orders of magnitude lower than for
17

the reaction between HO• and NPX (8.61 × 109 M−1s−1) (Lei et al., 2019; Wols et al., 2015). Therefore,
the formation of CO3•− almost completely compensates for the decreased contribution of HO• to NPX
−
degradation, resulting in a slight change in the overall k′ in the presence of HCO3 .
Fig. 4b shows the effect of NOM on the degradation of NPX by the UV/H2O2 and
−
UV/H2O2/HCO3 AOPs. NOM slowed NPX degradation in both AOPs, while the decrease in k′ in
−
UV/H2O2 was greater than in UV/H2O2/HCO3 . Specifically, 1 mg/L NOM decreased the values of k′
−
by 18% and 10.9% in UV/H2O2 and UV/H2O2/HCO3 , respectively. The molar absorption coefficient
(Ɛ254 nm) of NOM is 3.15 (m−1 (mg/L)−1) (Fang et al., 2014), which serves as an inner filter for UV light.
UV light transmittance was reduced by 10.6% in the reactor containing 1 mg/L NOM in the UV/H2O2
system, which reduced the H2O2 photolysis and thus decreased the production of HO• and the
•−
−
subsequent formation of CO3 . Therefore, the decrease in the overall k′ in UV/H2O2/HCO3 was
mainly due to the inner filter of NOM. Furthermore, NOM significantly scavenged HO•, witha k of 2.0
× 104 mg−1 L s−1, while the consumption of CO3•− was negligible, with a k of 135 mg−1 L s−1 (Yan et al.,
2019). Therefore, the scavenging effect of NOM on CO3•− is negligible, resulting in smaller effect of
−
NOM on the degradation of NPX by UV/H2O2/HCO3 .
−
Fig. 4c shows the efficiencies of NPX degradation by the UV/H2O2 and UV/H2O2/HCO3 AOPs
−
in tap water. In tap water, the k′ of NPX degradation in UV/H2O2/HCO3 was similar to the value in
UV/H2O2, which were 0.00132 s−1 and 0.00127 s−1, respectively. However, both values of k′ for NPX
−
degradation decreased by 18.9% and 13% by UV/H2O2 and UV/H2O2/HCO3 , respectively, compared
to pure water. The characteristics of the tap water are presented in Table S6. The concentrations of
−
NOM, Cl− and NO3 were 1.36 mg/L, 11.65 mg/L and 8.91 mg/L, respectively. As mentioned above,
18

NOM acts as an inner filter and a radical scavenger to consume HO•, which should be responsible for
−
the decrease in k′ observed in tap water. In addition, the effects of NO3 and Cl− are negligible. The
−
−
photolysis of NO3 can produce HO•, but the production of HO• by < 5 mM NO3 was negligible (Keen
et al., 2012). Cl− reacts quickly with HO•, but it is a reversible reaction (Jayson et al., 1973).
•−
According to the kinetic model, the steady-state concentrations of HO• and CO3 in the presence of
11.65 mg/L Cl− were similar to the concentrations observed in the absence of Cl− (Table S5). Therefore,
the slight inhibition of NPX degradation in tap water was mainly attributable to NOM.
•−
3.4 Pathways of NPX degradation by CO3
Based on the products identified using UPLC-Q-TOF-MS, the detailed information (i.e.,
elemental formula and structure) for the products obtained during the degradation of NPX by CO3•− is
provided in Table S7. ESI negative mode was used to identify the products generated during the
degradation of NPX. The total ion chromatogram of NPX is shown in Fig. S11 and the MS spectra of
the NPX main products are shown in Fig. S12. Isomers of the degradation products are distinguished
by the peaks at different retention time with the same mass-to-charge ratio (m/z). During the
−
degradation of NPX by UV/H2O2/HCO3 , the major products generated were P186 (C12H10O2, m/z =
185), P202 (C13H14O2, m/z = 201), P204 (C11H8O4, m/z = 203), P216 (C12H8O4, m/z = 215), P218a
(C12H10O4, m/z = 217), P218b (C13H14O3, m/z = 217), P234a (C12H10O5, m/z = 233), P234b (C13H14O4,
m/z = 233), P246a,b (C13H10O5, m/z = 245), P248a,b,c,d (C13H12O5, m/z = 247), P250a,b (C13H14O5, m/z =
249), P252 (C12H12O6, m/z = 251) and P266 (C13H14O6, m/z = 265).
19

Quantum chemical calculations were used to predict the reaction sites for CO3•− in NPX, which
might be beneficial to refine the structure of products. The HOMOs were calculated for NPX (Fig.
S13). The HOMOs of NPX were dominated by the C atoms on the naphthalene ring, suggesting that
the sites of NPX attacked by the radicals are primarily located on aromatic rings. Moreover, O17 and
C22 on the side chains also contributed to the HOMOs. Additionally, the partial charges assigned to
each atom for the NPX molecules are shown in Table S8. Based on the HOMOs and partial charges,
NPX-C10, NPX-C5, NPX-C2 and NPX-C11 located on the naphthalene ring provided the main
contributions to the HOMOs. P202 is an initial intermediate in the process, and thus P202 was
calculated using the same method (Fig. S13 and Table S8), consistent with the result calculated for
NPX. Based on the quantum chemical calculations, the reaction sites for CO3•− in NPX were proposed.
Fig. 5 shows the time-dependent evolutionary profiles of the major products of NPX degraded by
−
UV/H2O2/HCO3 at pH 8.6. The abundances of P202, P204, P216, P218a, P218b, P234b, P246b,
P248a,b,c,d and P266 initially increased and then decreased, whereas the abundances of P186, P234a,
P246a, P250a,b and P252 monotonically increased with the reaction time. Based on the identified
products, the pathways of NPX degradation by CO3•− are proposed in Scheme 1. The main pathways of
•−
NPX transformation by CO3 included decarboxylation, hydroxylation, ketonization, demethylation
and aldolization. First, NPX was transformed to P202 through decarboxylation and hydroxylation
together, or to P248c,d through hydroxylation and demethylation. Then, P202 was further transformed
to P218b, P234b, P250a,b and P266 via hydroxylation, while the demethylation/ketonization of P202
formed P186. The demethylation/ketonization of P234b generated P216, and the
demethylation/aldolization of P234b generated P218a, which was finally converted to P204 by
20

demethylation. Furthermore, P266 was transformed to P252 via demethylation, P250a,b was ketonized
to form P248a,b, which then generated P234a and P246a,b via demethylation and ketonization,
respectively.
Compared with HO• (Dulova et al., 2017; Pan et al., 2018), the initial stages of NPX
•−
transformation by CO3 were similar, namely, demethylation, hydroxylation and decarboxylation.
•−
However, the products formed by HO• and CO3 through these pathways are largely different, and
•−
•−
only P248c was observed at both the HO• and CO3 systems (Table S9). Furthermore, CO3
transformation of NPX subsequently resulted in the formation of a series of quinones, aldehydes and
ketones, including P246a,b, P248a,b, P234a, P186, P216, P218a and P204. The potential explanation for
•−
the differences is the difference in reaction mechanisms between CO3 and HO•. The mechanisms of
•−
CO3 are electron transfer and hydrogen abstraction (Liu et al., 2018; Liu et al., 2016; Moore et al.,
1977), while the mechanisms of HO• with NPX are addition as the major step and hydrogen
abstraction and single electron transfer as the minor steps (Luo et al., 2018). Similar to NPX, the
•−
degradation of bisphenol A by CO3 was also mainly mediated by electron transfer to form
hydroxylated products (Gao et al., 2020), whereas the attack of HO• on bisphenol A is mediated by
•−
addition and hydrogen abstraction (Sharma et al., 2016). The different reaction mechanisms of CO3
and HO• were responsible for the differences in products generated during the degradation of NPX
−
by UV/H2O2 and UV/H2O2/HCO3 , although the kinetics were similar.
21

3.5 Alterations in the genotoxicity of NPX after the CO3•− treatment
The genotoxicity results are presented as the growth factor (G) and the induction ratio (IR) (Text
S4). The IR value is valid when G > 0.5, and the genotoxicity is considered positive when IR > 2,a
sample is considered weakly genotoxic if the IR is between 2 and 3, moderately genotoxic if between
3 and 6 and strongly genotoxic if > 6 (Karci et al., 2014; Lv et al., 2014).
Fig. 6 shows the genotoxicity alteration during the NPX degradation by CO3•− and UV/Cl2. Note
that the G values of all samples were greater than 0.5 (Fig. S14). The initial IR value of NPX is
1.47-1.56, indicating that NPX does not have mutagenicity (Fig. 6). The IR values were 0.96-1.47,
•−
indicating that the products formed during NPX degradation by CO3 have negligible mutagenic
activities. Generally, the toxicity of organic compounds positively correlates with their octanol-water
partition coefficients (Kow) (Yang et al., 2017). The negligible genotoxicity observed during NPX
•−
degradation by CO3 is partially attributed to the high hydrophilicity of the hydroxylated products.
Additionally, the genotoxicity of these naphthoquinone derivatives was also likely to be negligible,
as naphthoquinone derivatives are often used to treat tumours and many of their derivatives have
proven to be effective as antimalarial and anti-acquired immunodeficiency syndrome treatments (Li
•−
et al., 2018). Similar to that by CO3 , the genotoxicity was also negligible during the NPX
transformation by HO• (Li et al., 2020). However, the IR value increased significantly and reached
the maximum of 4.1 at the 50% removal rate of NPX by UV/Cl2, which then decreased but still
higher than 2 when the removal rate of NPX was 95% (Fig. 6). This indicates that the products
formed during the NPX degradation by UV/Cl2 have high mutagenicity, mainly due to the formation
of toxic chlorine containing products (Pan et al., 2018; Xiang et al., 2020; Sieira et al., 2020).
22

4. Conclusions and engineering implications
CO3•− selectively reacts with the compounds containing aniline and phenolic hydroxyl groups as
well as naphthalene rings with the second-order rate constants in the range of 5.6 × 107 M-1s-1 to 2.96 ×
108 M-1s-1. By using a primary QSAR, the reactivity of CO3•− with PPCPs was found to depend on the
−
electron-donating properties of substituents on the aromatic ring. The presence of HCO3 has slight
•−
effects on the degradation of NPX by UV/H2O2, UV/PS and UV/Cl2 due to the formation of CO3 , and
•−
even significantly increased the k′ by 127% in UV/NH2Cl as the radical conversion from Cl• to CO3
outcompeted the consumption of Cl• by side reactions. The overall k′ of NPX by UV/H2O2 changed
−
•−
slightly as the HCO3 concentrations increased from 5 to 100 mM, due to that the formation of CO3
almost completely compensated for the decreased contribution of HO• to NPX degradation. The
scavenging effect of NOM on CO3•− is slight. The reaction of CO3•− with NPX is mediated by electron
transfer and hydrogen abstraction, while the reaction with HO• is mainly mediated by addition.
•−
Importantly, the degradation of NPX by CO3 does not increase genotoxicity. However, due to the
complexity of the composition of the real water environment, more precursor substances are needed
for comprehensive assessment.
−
•−
CO32−/HCO3 is ubiquitous in natural water systems, leading to the inevitable production of CO3
by AOPs during water treatment. The roles of CO3•− in the abatement of micropollutants depend on the
compound. Additionally, the effects of CO3•− on the degradation of PPCPs varied in different AOPs
under similar conditions. The transformation of micropollutants susceptible to CO3•− differed from that
of HO•, which requires further investigation in a future study.
23

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal
relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
This work was financially supported by the Natural Science Foundation of China (21922612,
21677181, 51908564). We would like to acknowledge Tian Lu for Gaussian09 software support.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.org...
24

References
APHA-AWWA-WEF, 2012. Standard Methods for the Examination of Water and Wastewater. 22nd.
American Public Health Association/American Water Works Association/Water Environment
Federation, Washington DC.
Balakrishna, K., Rath, A., Praveenkumarreddy, Y., Guruge, K.S., Subedi, B., 2017. A review of the
occurrence of pharmaceuticals and personal care products in Indian water bodies. Ecotox.
Environ. Safe. 137, 113-120.
Behar, D., Czapski, G., Duchovny, I., 1970. Carbonate radical in flash photolysis and pulse
radiolysis of aqueous carbonate solutions. J. Phys. Chem. 74(10), 2206-2210.
Bolton, J., Linden, K., 2003. Standardization of methods for fluence (UV Dose) determination in
bench-scale UV experiments. J. Environ. Eng. 129, 209-216.
Buxton, G.V., Elliot, A.J., 1986. Rate constant for reaction of hydroxyl radicals with bicarbonate
ions. Int. J. Radiat. Appl. Instrum. Part C Radiat. Phys. Chem. 27(3), 241-243.
Buxton, G.V., Greenstock, C.L., Helman, W.P., Ross, A.B., 1988. Critical Review of rate constants
for reactions of hydrated electrons, hydrogen atoms and hydroxyl radicals (•OH/•O−) in aqueous
solution. J. Phys. Chem. Ref. Data 17(2), 513-886.
Canonica, S., Kohn, T., Mac, M., Real, F.J., Wirz, J., von Gunten, U., 2005. Photosensitizer method
to determine rate constants for the reaction of carbonate radical with organic compounds. Environ.
Sci. Technol. 39(23), 9182-9188.
Chen, F., Xia, L., Zhang, Y., Bai, J., Wang, J., Li, J., Rahim, M., Xu, Q., Zhu, X., Zhou, B., 2019.
Efficient degradation of refractory organics for carbonate-containing wastewater via generation
carbonate radical based on a photoelectrocatalytic TNA-MCF system. Applied Catalysis B:
Environmental 259, 118071.
Cope, V.W., Chen, S., Hoffman, M.Z., 1973. Intermediates in the photochemistry of of
carbonato-amine complexes of cobalt(III). Carbonate(-) radicals and the aquocarbonato complex.
J. Am. Chem. Soc. 95(10), 3116-3121.
Dell'Arciprete, M.L., Soler, J.M., Santos-Juanes, L., Arques, A., Mártire, D.O., Furlong, J.P.,
Gonzalez, M.C., 2012. Reactivity of neonicotinoid insecticides with carbonate radicals. Water
25

Res. 46(11), 3479-3489.
Duca, G., Boldescu, V., 2009. Pharmaceuticals and Personal Care Products in the Environment.
Bahadir, A.M. and Duca, G. (eds), pp. 27-35, Springer Netherlands, Dordrecht.
Dulova, N., Kattel, E., Trapido, M., 2017. Degradation of naproxen by ferrous ion-activated
hydrogen peroxide, persulfate and combined hydrogen peroxide/persulfate processes: The effect
of citric acid addition. Chem. Eng. J. 318, 254-263.
Elango, T.P., Ramakrishnan, V., Vancheesan, S., Kuriacose, J.C., 1984. Reaction of the carbonate
radical with substituted anilines. Proc. Indian Acad. Sci.-Chem. Sci. 93(1), 47-52.
Fang, J., Fu, Y., Shang, C., 2014. The roles of reactive species in micropollutant degradation in the
UV/Free chlorine system. Environ. Sci. Technol. 48(3), 1859-1868.
Frisch, G.W.T.M., Schlegel, H.B., Scuseria, G.E., Robb, M.A., Cheeseman, J.R., Scalmani, G.,
Barone, V., Mennucci, B., Petersson, G.A., Nakatsuji, H., Caricato, M., Li, X., Hratchian, H.P.,
Izmaylov, A.F., Bloino, J., Zheng, G., Sonnenberg, J.L., Hada, M., Ehara, M., Toyota, K., Fukuda,
R., Hasegawa, J., Ishida, M., Nakajima, T., Honda, Y., Kitao, O., Nakai, H., Vreven, T., Jr.
Montgomery, J.A., Peralta, J.E., Ogliaro, F., Bearpark, M., Heyd, J.J., Brothers, E., Kudin, K.N.,
Staroverov, V.N., Keith, T., Kobayashi, R., Normand, J., Raghavachari, K., Rendell, A., Burant,
J.C., Iyengar, S.S., Tomasi, J., Cossi, M., Rega, N., Millam, J.M., Klene, M., Knox, J.E., Cross,
J.B., Bakken, V., Adamo, C., Jaramillo, J., Gomperts, R., Stratmann, R.E., Yazyev, O., Austin,
A.J., Cammi, R., Pomelli, C., Ochterski, J.W., Martin, R.L., Morokuma, K., Zakrzewski, V.G.,
Voth, G.A., Salvador, P., Dannenberg, J.J., Dapprich, S., Daniels, A.D., Farkas, O., Foresman,
J.B., Ortiz, J.V., Cioslowski, J., Fox, D.J., 2010. Gaussian, Gaussian 09, revision C.01; Gaussian,
Inc.: Wallingford CT.
Gao, J., Duan, X., O Shea, K., Dionysiou, D.D., 2020. Degradation and transformation of bisphenol
A in UV/Sodium percarbonate: Dual role of carbonate radical anion. Water Res. 171, 115394.
Garoma, T., Gurol, M., 2005. Modeling aqueous ozone/UV process using oxalic acid as probe
chemical. Environ. Sci. Technol. 39, 7964-7969.
Grebel, J.E., Pignatello, J.J., Mitch, W.A., 2010. Effect of halide ions and carbonates on organic
contaminant degradation by hydroxyl radical-based advanced oxidation processes in saline waters.
26

Environ. Sci. Technol. 44(17), 6822-6828.
Guo, K., Wu, Z., Yan, S., Yao, B., Song, W., Hua, Z., Zhang, X., Kong, X., Li, X., Fang, J., 2018.
Comparison of the UV/chlorine and UV/H2O2 processes in the degradation of PPCPs in simulated
drinking water and wastewater: Kinetics, radical mechanism and energy requirements. Water Res.
147, 184-194.
Hand, V.C., Margerum, D.W., 1983. Kinetics and mechanisms of the decomposition of dichloramine
in aqueous solution. Inorg. Chem. 22(10), 1449-1456.
Hansch, C., Leo, A., Hoekman, D., 1995. Exploring QSAR: Hydrophobic, electronic, and steric
constants. Hansch, C., Leo, A. and Hoekman, D.H. (eds), American Chemical Society,
Washington, DC.
Hansch, C., Leo, A., Taft, R.W., 1991. A survey of Hammett substituent constants and resonance and
field parameters. Chem. Rev. 91(2), 165-195.
Heberer, T., 2002. Occurrence, fate, and removal of pharmaceutical residues in the aquatic
environment: a review of recent research data. Toxicol. Lett. 131(1), 5-17.
Huang, J., Mabury, S., 2000. A new method for measuring carbonate radical reactivity toward
pesticides. Environ. Toxicol. Chem. 19, 1501-1507.
Huie, R.E., Clifton, C.L., 1990. Temperature dependence of the rate constants for reactions of the
•−
sulfate radical, SO4 , with anions. J. Phys. Chem. 94(23), 8561-8567.
ISO13829, 2000. Water Quality—Determination of the Genotoxicity of Water and Waste Water
Using the umu-Test, Geneva.
Jayson, G.G., Parsons, B.J., Swallow, A.J., 1973. Some simple, highly reactive, inorganic chlorine
derivatives in aqueous solution. Their formation using pulses of radiation and their role in the
mechanism of the Fricke dosimeter. J. Chem. Soc., Faraday Trans. 1 69(0), 1597-1607.
Karci, A., Arslan-Alaton, I., Bekbolet, M., Ozhan, G., Alpertunga, B., 2014. H2O2/UV-C and
Photo-Fenton treatment of a nonylphenol polyethoxylate in synthetic freshwater: Follow-up of
degradation products, acute toxicity and genotoxicity. Chem. Eng. J. 241, 43-51.
Keen, O.S., Love, N.G., Linden, K.G., 2012. The role of effluent nitrate in trace organic chemical
oxidation during UV disinfection. Water Res. 46(16), 5224-5234.
27

Kintecus, 2020. Kintecus–Enzyme & Combustion Chemical Kinetics and Equilibrium Software.
kinetics. (Accessed May 29, 2020).
Lai, W.W., Hsu, M., Lin, A.Y., 2017. The role of bicarbonate anions in methotrexate degradation via
UV/TiO2: Mechanisms, reactivity and increased toxicity. Water Res. 112, 157-166.
Lam, M., Tantuco, K., Mabury, S., 2003. PhotoFate: a new approach in accounting for the
contribution of indirect photolysis of pesticides and pharmaceuticals in surface waters. Environ.
Sci. Technol. 37, 899-907.
Larson, R., Zepp, R., 1988. Reactivity of the carbonate radical with aniline derivatives. Environ.
Toxicol. Chem. 7, 265-274.
Lee, Y., von Gunten, U., 2012. Quantitative structure–activity relationships (QSARs) for the
transformation of organic micropollutants during oxidative water treatment. Water Res. 46(19),
6177-6195.
Lei, Y., Cheng, S., Luo, N., Yang, X., An, T., 2019. Rate constants and mechanisms of the reactions
of Cl• and Cl2•– with trace organic contaminants. Environ. Sci. Technol. 53(19), 11170-11182.
Leung, H.W., Minh, T.B., Murphy, M.B., Lam, J.C.W., So, M.K., Martin, M., Lam, P.K.S.,
Richardson, B.J., 2012. Distribution, fate and risk assessment of antibiotics in sewage treatment
plants in Hong Kong, South China. Environ. Int. 42, 1-9.
Li, K., Wang, B., Zheng, L., Yang, K., Li, Y., Hu, M., He, D., 2018. Target ROS to induce apoptosis
and cell cycle arrest by 5,7-dimethoxy-1,4-naphthoquinone derivative. Bioorg. Med. Chem. Lett.
28(3), 273-277.
Li, M., Wen, Q., Chen, Z., Tang, Y., Yang, B., 2020. Comparison of ozonation and UV based
oxidation as pre-treatment process for ultrafiltration in wastewater reuse: Simultaneous water
risks reduction and membrane fouling mitigation. Chemosphere 244, 125449.
Lian, L., Yao, B., Hou, S., Fang, J., Yan, S., Song, W., 2017. Kinetic study of hydroxyl and sulfate
radical-mediated oxidation of pharmaceuticals in wastewater effluents. Environ. Sci. Technol.
51(5), 2954-2962.
Liu, J., Wong, M., 2013. Pharmaceuticals and personal care products (PPCPs): a review on
environmental contamination in China. Environ. Int. 59, 208-224.
28

Liu, T., Yin, K., Liu, C., Luo, J., Crittenden, J., Zhang, W., Luo, S., He, Q., Deng, Y., Liu, H., Zhang,
D., 2018. The role of reactive oxygen species and carbonate radical in oxcarbazepine degradation
via UV, UV/H2O2: Kinetics, mechanisms and toxicity evaluation. Water Res. 147, 204-213.
Liu, Y., He, X., Duan, X., Fu, Y., Fatta-Kassinos, D., Dionysiou, D.D., 2016. Significant role of UV
and carbonate radical on the degradation of oxytetracycline in UV-AOPs: Kinetics and
mechanism. Water Res. 95, 195-204.
Luo, S., Gao, L., Wei, Z., Spinney, R., Dionysiou, D.D., Hu, W., Chai, L., Xiao, R., 2018. Kinetic
and mechanistic aspects of hydroxyl radical‒mediated degradation of naproxen and reaction
intermediates. Water Res. 137, 233-241.
Luo, Y., Guo, W., Ngo, H.H., Nghiem, L.D., Hai, F.I., Zhang, J., Liang, S., Wang, X.C., 2014. A
review on the occurrence of micropollutants in the aquatic environment and their fate and
removal during wastewater treatment. Sci. Total Environ. 473-474, 619-641.
Lv, L., Jiang, T., Zhang, S., Yu, X., 2014. Exposure to mutagenic disinfection byproducts leads to
increase of antibiotic resistance in pseudomonas aeruginosa. Environ. Sci. Technol. 48(14),
8188-8195.
Madikizela, L.M., Chimuka, L., 2016. Determination of ibuprofen, naproxen and diclofenac in
aqueous samples using a multi-template molecularly imprinted polymer as selective adsorbent for
solid-phase extraction. J. Pharmaceut. Biomed. 128, 210-215.
Manrique-Moreno, M., Suwalsky, M., Villena, F., Garidel, P., 2010. Effects of the nonsteroidal
anti-inflammatory drug naproxen on human erythrocytes and on cell membrane molecular models.
Biophys. Chem. 147(1), 53-58.
Matthew, B., Anastasio, C., 2006. A chemical probe technique for the determination of reactive
halogen species in aqueous solution: Part 1 – bromide solutions. Atmospheric Chemistry and
Physics Discussions 6, 899-940.
Mertens, R., Sonntag, C., 1995. Photolysis (λ = 254 nm) of tetrachloroethene in aqueous solutions. J.
Photochem. Photobiol. A-Chem. 85, 1-9.
Michael, I., Hapeshi, E., Osorio, V., Perez, S., Petrovic, M., Zapata, A., Malato, S., Barceló, D.,
Fatta-Kassinos, D., 2012. Solar photocatalytic treatment of trimethoprim in four environmental
29