Chemical Engineering Journal 427 (2022) 131995

Contents lists available at ScienceDirect

Chemical Engineering Journal

journal homepage: www.elsevier.com/locate/cej

Enhanced atrazine degradation in the Fe(III)/peroxymonosulfate system via

accelerating Fe(II) regeneration by benzoquinone
Xiaowan Li a, Jun Ma b, Yufei Gao a, Xitao Liu c, *, Yi Wei a, d, Zongjun Liang d
a
Gansu Key Laboratory for Environmental Pollution Prediction and Control, College of Earth and Environmental Sciences, Lanzhou University, Lanzhou, Gansu 730000,
PR China
b
Development Research Center of the Ministry of Water Resources, Beijing 100038, PR China
c
State Key Laboratory of Water Environment Simulation, School of Environment, Beijing Normal University, Beijing 100875, PR China
d
Gansu Membrane Science and Technology Research Institute Co., Ltd, Lanzhou, Gansu 730020, PR China

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

Keywords: Fe(II)-involved advanced oxidation processes could effectively degrade refractory contaminants, but the slow
Peroxymonosulfate transformation of Fe(III) to Fe(II) limits their wide application. Herein we reported that p-benzoquinone (BQ)
Fe(II) regeneration could accelerate Fe(III)/Fe(II) cycle and promote atrazine (ATZ) degradation in the Fe(III)/peroxymonosulfate
Benzoquinone
(Fe(III)/PMS) system through reconstructing active Fe(II)/PMS system. Under the identical experimental con­
Ferryl ion
ditions, ATZ degradation efficiency in the BQ/Fe(III)/PMS system (99.8%) was superior to that in the Fe(III)/
Degradation pathway
PMS system (23.2%), suggesting the indispensable role of BQ in Fe(II) regeneration and PMS activation. Reactive
species sulfate radical (SO4•− ), hydroxyl radical (•OH) and ferryl ion (Fe(IV)) were involved in the BQ/Fe(III)/
PMS system based on alcohol quenching experiments, ESR tests and the selective transformation from methyl
phenyl sulfoxide (PMSO) to methyl phenyl sulfone (PMSO2). ATZ degradation mechanism was proposed based
on LC-MS detection, among which dealkylation and dechlorination-hydroxylation were the dominant trans­
formation pathways. The BQ/Fe(III)/PMS system showed satisfactory catalytic performance at pH 2.5–7.0 and
low Fe(III) concentration, which expanded the pH range and avoided the addition of a large amount of Fe. More
importantly, the increase of PMS concentration not only enhanced the oxidative degradation of ATZ and its two
main chlorinated byproducts (atrazine-desethyl (DEA) and atrazine-desisopropyl (DIA)), but also facilitated the
removal of toxic BQ. Cl− inhibited the degradation of ATZ and caused the accumulation of DEA and DIA, and the
higher Cl− , the more obvious this effect. This study provided a new approach to improve the Fe(III)/PMS system
by boosting Fe(III)/Fe(II) cycle with quinons.

1. Introduction desired to develop feasible and effective treatment methods to remove

ATZ from water.
Atrazine (ATZ) is a triazine herbicide developed by Geigy Chemical Although membrane filtration, physical adsorption, biodegradation
Company in 1952 and is commonly used for the prevention and control and other processes can be used to remove ATZ, they also have problems
of broad-leaved weeds such as corn, sorghum and sugarcane [1]. ATZ is such as high cost, long cycle, low treatment efficiency and potential
one of the most widely used herbicides in agriculture owing to its low secondary pollution. Persulfate (PS)-based advanced oxidation pro­
cost and decent weeding effect, but it also causes water pollution cesses (AOPs), widely concerned in the field of environmental pollution
worldwide because of its high persistence, low biodegradability and control and remediation, have shown broad application prospect in the
long-term use [2]. Although the European Union (EU) has banned the treatment of micro-pollutants such as ATZ. PS (including (PDS) and
use of ATZ in 2004, it is still used in the USA, China, Mexico and other peroxymonosulfate (PMS)) usually needs to be activated by external
countries [3]. At present, the USA and China set the ATZ limit in energy or catalyst to produce active species and degrade target pollut­
drinking water to 3 and 2 μg L− 1, respectively, while the EU is more ants due to its limited oxidation ability [4]. The literature on PS acti­
stringent and no >0.1 μg L− 1 is allowed [3]. Therefore, it is still highly vation has paid much attention to transition metals (Co(II), Fe(II), Cu

* Corresponding author.
E-mail address: liuxt@bnu.edu.cn (X. Liu).

https://doi.org/10.1016/j.cej.2021.131995
Received 26 May 2021; Received in revised form 16 August 2021; Accepted 20 August 2021
Available online 28 August 2021
1385-8947/© 2021 Elsevier B.V. All rights reserved.
X. Li et al. Chemical Engineering Journal 427 (2022) 131995

(II), Mn(II), etc.) [5], among which Fe(II) has been proven to effectively phosphide may have metal leaching problem, and cysteine/polyphenol
trigger the generation of reactive species, such as sulfate radical (SO4•− ), will additionally increase the total organic carbon of the reaction system
hydroxyl radical (•OH) and/or ferryl ion (Fe(IV) (Eqs. (1)–(6)) [6]. [15]. In this regard, using reducing agent originally existing in the water
Meanwhile, Fe(II) has the advantages of large quantity, high efficiency to regenerate Fe(II) may be a feasible strategy of improving efficiency
and non toxicity, thus the homogeneous Fe(II)/PS system has shown while reducing external pollution.
good development prospects in contaminant elimination, wastewater Quinones are a group of organic compounds with unique biological
treatment and soil remediation [7]. However, the Fe(II)/PS system also activities, which widely exist in soil, surface water and even atmospheric
has some drawbacks, such as the difficulty of Fe (II) regeneration, the aerosols and play an important role in biological and chemical processes
narrow range of pH and the large amount of iron sludge, which limit its by acting as electron-shuttle [16]. Quinones have a wide range of
practical application to a certain extent [8]. sources. In addition to being present in humic acid and traditional
Chinese medicine, they can also be generated by the degradation of
Fe(II) + S2O82− → Fe(III) + SO4•− + SO42− (1)
certain pollutants. For example, Zhou et al. found that quinone in­
Fe(II) + HSO5− → Fe(III) + SO4•− + OH− (2) termediates were formed when PMS was used to oxidize and degrade
phenols, and the substituent groups and their positions on phenols have
− •
Fe(II) + HSO5 → Fe(III) + OH + SO4 2−
(3) significant effects on the formation rate and type of quinone in­
termediates [17]. Besides, quinones are often used to synthetize in­
Fe(II) + S2O8 2−
(+H2O) → Fe OIV 2+
+ 2 SO4 2− +
(+2H ) (4)
termediates in industrial production, causing them to ubiquitously exist
Fe(II) + HSO5− → FeIVO2+ + SO42− + H+ (5) in chemical, pharmeceutical, printing and dyeing wastewater. Quinones
are inherent environmental pollutants, if they can be used to degrade
Side reactions: other pollutants, then the pollution control strategy of “treating waste
IV 2+
Fe O + Fe(II)/H2O/S2O82− /HSO5− → Fe(III) (6) with waste” can be realized. In recent years, quinones have indeed been
reported to activate PS to degrade organic pollutants under alkaline
Notably, Fe(II) regeneration is the rate-limiting step of the Fe(II)/PS conditions. Fang et al. found that benzoquinone (BQ) can activate PDS to
system, thus the slow conversion from Fe (III) to Fe(II) is the root cause degrade 2,4,4’-trichlorobiphenyl, and the internal mechanism was that
of the above shortcomings. As shown in Eqs. (1)–(5), reactive species are BQ first reacted with hydroquinone (HQ, the self-condensation or
rapidly produced by a series of chain reactions between Fe(II) and PS, decomposition of BQ) to generate semiquinone radicals, and then PDS
but the difficulty of Fe(II) regeneration hinders the continuous and was decomposed into SO4•− [18]. BQ can also activate PMS, but unlike
efficient operation of the Fe(II)/PS system. Despite both Fe(III) and Fe PDS, the active substance was considered to be singlet oxygen (1O2)
(II) have the ability to activate oxidants, Fe(III) is less effective than Fe instead of SO4•− [19]. Zhang et al. used 2,6-dimethyl-1,4-benzoquinone,
(II) in decomposing PS and generating reactive species [1,79]. There­ 2,6-dichloro-1,4-benzoquinone and tetrafluoro-1,4-benzoquinone as
fore, seeking ways to promote the regeneration of Fe(II) is highly model quinones, and found that PMS could be activated by halogenated
desired. Adding reducing agents is a simple way to accelerate the Fe and methylated benzoquinone to form 1O2 and degrade sulfamethoxa­
(III)/Fe(II) cycle and improve the degradation efficiency by rebuilding zole, and the process was highly pH and quinone dependent [20]. In
the Fe(II)/PS system and accelerating the formation of active species. addition to directly activating PS and producing active species to
Hydroxylamine (HA) is one of the most studied reducing agents in recent degrade pollutants, quinones also have the potential to promote the
years due to its strong ability of reducing Fe(III) to Fe(II) and low re­ conversion from Fe(III) to Fe(II). Chen et al. reported that BQ can
action rates with SO4•− and •OH. Its introduction can significantly improve the efficiency of the Fenton reaction by reducing Fe(III) and
enhance the efficiency of the reaction system: adding 2 mM HA to the Fe regenerating Fe(II) [21]. Duesterberg et al. also confirmed that quinones
(II)/PMS system can increase the degradation efficiency of benzoic acid can reduce Fe(III) and accelerate the redox cycle of Fe(III)/Fe(II) [22].
from less than 5% to >80% [10]. HA can not only promote Fe(III) Inspired by these attempts and facts, we expect the combination of
reduction in homogeneous system, but also accelerate Fe(II) regenera­ BQ and Fe(III) could boost Fe(II) regeneration and enhance PMS acti­
tion in heterogeneous system. According to Li et al., HA promoted the Fe vation, reactive species formation and contaminant degradation. Here­
(III)/Fe(II) cycle on the Fe3O4 surface, and the degradation efficiency of in, in this work, the stimulating effect of BQ on the Fe(III)/PMS system
atrazine in the Fe3O4/PMS/HA system was 38 times that in the Fe3O4/ was confirmed using herbicide atrazine (ATZ) as the target pollutant due
PMS system [11]. In addition to HA, homogeneous reductants such as to its potential threat to human health and ecological environment [23].
ascorbic acid (AA), cysteine and polyphenols have been proved to Subsequently, the underlying promotion mechanism of the constructed
improve the conversion rate of Fe (III) to Fe(II) [12]. Lei et al. introduced BQ/Fe(III)/PMS system and the degradation pathway of ATZ was
AA into the Fe(III)/PDS system and found that the system can not only explored. Crucial factors (including solution pH, Fe(III) dose and PMS
rapidly and continuously degrade organic pollutants, but also have the concentration) that affected the catalytic performance of the BQ/Fe(III)/
advantages of high removal efficiency and wide pH range [13]. This is PMS system were evaluated, and the influences of different Cl− con­
mainly due to the reduction and chelating ability of AA. With the centrations on the degradation of ATZ and the formation of two main
development of research, heterogeneous reductants such as zero valent chlorinated byproducts were emphatically studied. Ultimately, the po­
metal, metal sulfur/phosphide, boron/phosphorus powder were also tential of different phenols and quinones in accelerating Fe(III)/Fe(II)
used for Fe(II) regeneration [12]. Zhou et al. explored the role of sulfur cycle for PMS activation was also examined.
conversion by constructing a pyrite (FeS2) activated PMS system and
found that sulfide is the main electron donor for PMS activation, which 2. Materials and methods
can mediate Fe(II) regeneration on the FeS2 surface [14]. Zhou et al.
successfully prepared metal-free black-red phosphorus (BRP) based on 2.1. Chemicals and reagents
commercial red phosphorus (RP). Compared with RP, the ability of BRP
to reduce Fe(III), regenerate Fe(II) and activate PMS was significantly The details of chemicals and reagents are provided in Text S1.
improved due to the enhanced electron transfer from BRP to Fe(III), and
the constructed Fe(III)/PMS/BRP system had excellent recyclability and 2.2. Degradation procedure
strong versatility for natural water [15]. Although the addition of
reducing agents can boost Fe(II) regeneration and improve system per­ Batch experiments were conducted in triplicates in 50 mL plastic
formance, it may also cause secondary pollution risks. For instance, the centrifuge tubes under constant temperature water bath shaker at 150
commonly used HA is highly toxic and not recyclable, metal sulfur/ rpm and 30 ± 0.5 ◦ C. Typically, Fe(III) and PMS with desired

2
X. Li et al. Chemical Engineering Journal 427 (2022) 131995

concentrations (1 mL each) were orderly spiked in 17 mL of 50 μM ATZ 3. Results and discussion

solution, then the degradation process was initiated by adding 1 mL BQ
of preset concentration. At set intervals, 1.8 mL of samples were 3.1. Promoting effect of BQ on Fe(III)/PMS system
collected, filtered with 0.22 μm nylon filters and quenched by 0.1 mL
EtOH before analysis. Note that all experiments but pH effect were ATZ degradation in the Fe(III)/PMS system with/without BQ are
conducted at initial pH 3.0 with HClO4 serving as the adjuster. The shown in Fig. 1(A). Under the condition of 0.20 mM Fe(III), 0.25 mM
quenching experiments employed excess EtOH or TBA as scavengers, PMS and pH 3.0, the Fe(III)/PMS system showed poor catalytic perfor­
which were pre-added before the reaction started. For the Fe(IV) veri­ mance, with only 23.2% of ATZ being degraded in 120 min. This result
fication experiment, except that the PMSO solution was substituted for might be ascribed to that although Fe(III) could react with PMS to
the ATZ solution, other experiment steps were the same as the above. regenerate Fe(II) and form SO5•− via Eq. (7) [27], the conversion of Fe
(III) to Fe(II) is relatively slow [10] and the oxidation capability of SO5•−
2.3. Analytic methods (redox potential of 1.1 V at pH 7) is far inferior to SO4•− and •OH [28].
When 0.10–0.50 mM BQ was added into the Fe(III)/PMS system, the
The concentrations of ATZ, BQ, DEA, DIA, PMSO and PMSO2 were all catalytic oxidation process was accelerated in varying degrees. Espe­
analyzed on high performance liquid chromatography (HPLC, Waters cially when BQ concentration was equal to or exceeded 0.25 mM, the
e2695, USA) equipped with a UV detector and a reverse-phase C18 final ATZ degradation efficiency in the BQ/Fe(III)/PMS system was even
column (4.6 × 250 mm, 5 μm, CMW, China). The detailed HPLC pa­ close to 100%, much higher than 23.2% in the Fe(III)/PMS system
rameters are shown in Table S1. The electron spin resonance (ESR) tests Comparison between the BQ/Fe(III)/PMS system and other treatment
with DMPO as spin-trapping agent were conducted on a JEOL FA-200 methods is shown in Table S2. It can be seen that the BQ/Fe(III)/PMS
Spectrometer (JEOL, Japan) to identify reactive species in the BQ/Fe system can obtain decent degradation efficiency at higher ATZ con­
(III)/PMS system. The concentration of regenerative Fe(II) and residual centration (50 μM) and lower PMS dosage (0.25 mM). For water con­
PMS were measured on Hach DR6000 UV–vis spectrophotometer by taining both Fe(III) and BQ, this system can give full play to its
1,10-Phenanthroline method [24] and iodide spectrophotometry advantages and achieve the purpose of “treating waste with waste”. For
method [25], respectively. The degradation intermediates of ATZ were the enhancement caused by BQ, two possible reasons were proposed.
detected using a Shimadzu LC-20AD UPLC coupled with a 4500 triple First, according to Zhou et al., BQ/PMS system in alkaline condition
quadrupole mass spectrometry (AB SCIEX, USA) with a waters Sym­ could generate singlet oxygen (1O2), which made main contribution to
metry C18 column (2.5 μm, 2.1 × 100 mm). The detailed analysis pa­ sulfamethoxazole degradation [19]. However, Fig. S1 shows that the
rameters please see our previous study [26]. BQ/PMS system at pH 3 (experimental condition of this study) failed to
effectively degrade ATZ, suggesting BQ did not improve the efficiency of
the Fe(III)/PMS system by activating PMS. Second, Chen et al. reported
that BQ could act as electron shuttle and strengthen the catalytic

Fig. 1. (A) ATZ degradation efficiency, (B) A510 nm value and (C) PMS utilization efficiency in the BQ/Fe(III)/PMS system with different BQ concentrations; (D) ATZ
degradation efficiency in the Fe(III)/PMS, Co(II)/PMS and Cu(II)/PMS systems with/without BQ. Reaction condition: (A–C) [ATZ] = 50 μM, [Fe(III)] = 0.20 mM,
[PMS] = 0.25 mM, pH = 3; (D) Fe(III)/PMS system with/without BQ: [ATZ] = 50 μM, [Fe(III)] = 0.20 mM, [PMS] = 0.25 mM, [BQ] = 0.25 mM, pH = 3; Co(II)/PMS
system with/without BQ: [ATZ] = 50 μM, [Co(II)] = 5 μM, [PMS] = 0.25 mM, [BQ] = 0.50 mM, pH = 3; Cu(II)/PMS system with/without BQ: [ATZ] = 10 μM, [Cu
(II)] = 10 μM, [PMS] = 0.25 mM, [BQ] = 0.50 mM, pH = 3.

3
X. Li et al. Chemical Engineering Journal 427 (2022) 131995

performance of Fenton system by reducing Fe(III) and regenerating Fe reactivity towards PMS because it can only act as an electron acceptor,
(II) [21]. In addition, Duesterberg et al. also confirmed that quinones while it is Cu(I) that activates PMS and induces the generation of SO4•−
have a reducing effect on Fe(III) and accelerate the cycle of Fe(III)/Fe(II) and •OH [30]. But Cu(I) also has the defect of instability, which leads to
[22]. Inspired by these reports, we preliminarily conjuctured that the unsatisfactory oxidation ability [31].
enhancement might be attributed to that BQ could boost the cycle of Fe
(III)/Fe(II), accelerate the regeneration of Fe(II) and then make it keep 3.2. Reaction mechanism and degradation pathway
high catalytic reactivity through reconstructing the Fe(II)/PMS system.
3.2.1. Activation mechanism
Fe(III) + HSO5− → Fe(II) + SO5•− + H+ (7)
Considering BQ could reduce Fe(III) to Fe(II) and then reconstruct
To further confirm this conjucture, the concentration of regenerative the Fe(II)/PMS system, the BQ/Fe(III)/PMS system is likely to have the
Fe(II) in the BQ/Fe(III)/PMS system was measured by 1,10-Phenanthro­ same free radicals as the Fe(II)/PMS system, i.e., generally recognized
line method and the relevant results are shown in Fig. 1(B), where the SO4•− and •OH. Thus main reactive species in the BQ/Fe(III)/PMS sys­
value of A510 nm meant the absorbance at 510 nm. In general, the higher tem were identified by alcohol quenching experiments and ESR tests.
BQ concentration, the greater A510 nm value, which suggested that BQ Ethanol (EtOH) was selected to scavenge both SO4•− and •OH [32],
could accelerate Fe(III)/Fe(II) cycle and Fe(II) regeneration was signif­ while tert-butyl alcohol (TBA) was used for quenching •OH [33]. Results
icantly dependent on added BQ. It should be noted that except for 0.10 showed that ATZ degradation efficiency significantly dropped from
and 0.50 mM BQ, all cases had a low A510 nm value period followed by a 99.8% to 11.9% and 35.6% in the presence of 0.1 M EtOH and TBA,
fast accumulation of regenerative Fe(II). And with the increase of BQ respectively (Fig. 2(A)). This phenomenon preliminarily indicated that
dose, the low tide period gradually shortened and the accumulation was like the Fe(II)/PMS system, the BQ/Fe(III)/PMS system simultaneously
significantly strengthened. These results could be explained in the generated SO4•− and •OH. Subsequently, ESR tests with DMPO as the
following way. When 0.10 mM BQ was added to the Fe(III)/PMS system, spin trapping agent were performed to further identify the formation of
the amount of regenerative Fe(II) was limited, Fe(II) was regenerated by SO4•− and •OH. No peak was observed in the Fe(III)/PMS system (data
BQ while being consumed by PMS, causing a low A510nm value during not shown), while distinct signals corresponding to DMPO-SO4•− and
the whole reaction. As BQ concentration increased, A510 nm value DMPO-•OH adducts were presented in Fig. 2(B), providing clear evi­
showed the trend of “gliding first and then taking off”. Take 0.25 mM BQ dence for the coexistence of SO4•− and •OH in the BQ/Fe(III)/PMS
for example, A510 nm value in the first 40 min was relatively low and system.
stable, but gradually rose to 0.141, 0.202, 0.266 and 0.302 at 60, 80, Very recently, Wang at al. held different views on the reactive species
100 and 120 min, respectively. The stable A510nm value in the early in the Fe(II)/PDS system, and proposed that Fe(IV) rather than SO4•−
period might be due to the balance between the production and con­ was the dominant intermediate based on the selective transformation
sumption of Fe(II), while the soaring A510nm value in the later stage was from PMSO to PMSO2 [6]. Subsequently, the authors found that both
related to the gradual degradation of ATZ and consumption of PMS. In SO4•− and Fe(IV) were involved in nZVI/PS system, and PS activation
particular, the greater BQ was added, the faster ATZ was degraded, the induced by nZVI through electron transfer led to SO4•− formation while
more PMS was decomposed, and the earlier A510nm value rose, which Fe(II) released from nZVI activated PS and then produced Fe(IV) [34].
was consistent with the results of gradually advanced take-off point, i.e., Our previous study have reached similar conclusions that SO4•− , •OH
80, 40, 20 min in the presence of 0.20, 0.25 and 0.30 mM BQ, respec­ and Fe(IV) coexisted in Fe0/PMS system [26]. Considering that BQ could
tively. However, when further increasing BQ concentration to 0.5 mM, reduce Fe(III) to Fe(II) and rebuild the Fe(II)/PMS system, whether Fe
Fe(II) regeneration was much greater than its consumption, leading to (IV) was generated in the BQ/Fe(III)/PMS system was confirmed
the gliding period disappeared and entered the ascent stage directly. In through the degradation of PMSO and the formation of PMSO2. Note
addition, the residual PMS after the reaction was also detected and it can that PMS itself could oxide PMSO to PMSO2 [35], but PMSO loss was
be seen that more and more PMS was consumed as added BQ rose (as lower than that in the BQ/Fe(III)/PMS system (as shown in Fig. 2(C)).
shown in Fig. 1(C)). For instance, only 48.01 μM PMSO was degraded in sole PMS system
In addition to the Fe(III)/PMS system, the Cu(II)/PMS and Co(II)/ within 120 min, while 77.76 μM in the BQ/Fe(III)/PMS system due to
PMS systems were also constructed to explore whether BQ had a pro­ the successive Fe(II) regeneration, PMS activation and reactive species
moting effect on them (Fig. 1(D)). Like the Fe(III)/PMS system, the formation. Similar phenomenon was observed in the case of PMSO2
presence of BQ also accelerated ATZ degradation in the Cu(II)/PMS formation. As shown in Fig. 2 (C), PMS oxidized PMSO and produced
system, with ATZ degradation efficiency being increased from 13.6% to 3.3, 4.6, 5.4, 4.2, 6.9, 7.9, 8.2 and 8.4 μM PMSO2 at given reaction time
36.7%. This enhancement probably because BQ induced the production in sole PMS system, while 6.4, 9.6, 13.2, 16.4, 20.4, 21.7, 21.1 and 22.3
of reactive Cu(I) by promoting Cu(II)/Cu(I) cycle [29], and then led to μM PMSO2 were formed in the BQ/Fe(III)/PMS system, respectively.
PMS activation, reactive species formation and ATZ degradation. The yield of PMSO2 (i.e., the molar ratio of PMSO2 formation to PMSO
However, different from the above two systems, no enhancement but degradation) in the BQ/Fe(III)/PMS was always higher than that in the
inhibition was observed in the Co(II)/PMS system, this may be because PMS system (Fig. 2 (D)), indicating that Fe(IV) existed in the BQ/Fe(III)/
BQ not only failed to boost Co(III)/Co(II) cycle, it also acted as a PMS system. However, the yield of PMSO2 was not high in the BQ/Fe
pollutant and competed active species with ATZ. The opposite effect of (III)/PMS system, and maintained between 25 and 30% in the latter
BQ on Cu(II)/Cu(I) and Co(III)/Co(II) cycles might be related to their stage of the reaction. This may be because SO4•− , •OH and Fe(IV) were
standard reduction potential. The standard reduction potential of Co all involved in the BQ/Fe(III)/PMS system, and Fe(IV) only accounted
(III)/Co(II) cycle is 1.92 V, which is much higher than 0.15 V of Cu(II)/ for part. Besides, it has been reported that Fe(IV) tended to be an
Cu(I) cycle. Therefore, compared with Cu(II)/Cu(I) cycle, it was more alternative reactive substance at near-neutral pH [6], whereas the BQ/
difficult for BQ to reduce Co(III) to reactive Co(II), resulting in no Fe(III)/PMS system in this study was conducted at pH 3. It should be
continuous available Co(II) to activate PMS for ATZ degradation. It noted that both PMSO degradation and PMSO2 formation did not change
should be noted that the standard reduction potential of Fe(III)/Fe(II) much after 60 min of reaction, which might be due to the the exhaustion
(0.771 V) is larger than that of Cu(II)/Cu(I) cycle (0.15 V), theoretically, of PMS.
Cu(II)/Cu(I) cycle is more likely to occur. However, in this study, the Based on the above results, we proposed the activation mechanism of
promotion effect of BQ on the Fe(III)/PMS system was much greater the BQ/Fe(III)/PMS system (Fig. 3). First, acting as electron-shuttle, BQ
than that on the Cu(II)/PMS system. This may be attributed to the poor transmitted electrons to Fe(III), initiated the cycle of Fe(III)/Fe(II) and
catalytic performance of the Cu(II)/PMS system. Cu(II) exhibits poor led to the regeneration of Fe(II). Then two reaction pathways might have
happened. On the one hand, PMS was activated by the regenerated Fe(II)

4
X. Li et al. Chemical Engineering Journal 427 (2022) 131995

Fig. 2. (A) ATZ degradation efficiency in the BQ/Fe(III)/PMS system with 0.1 M TBA and EtOH, (B) ESR spectra of DMPO-•OH and DMPO-SO4•− in the BQ/Fe(III)/
PMS system; (C) PMSO degradtion and PMSO2 formation and (D) The yield of PMSO2 in sole PMS and BQ/Fe(III)/PMS systems. Reaction condition: [ATZ] = 50 μM,
[PMSO] = 200 μM, [Fe(III)] = 0.20 mM, [PMS] = 0.25 mM, [BQ] = 0.25 mM, pH = 3.

SO4•− + OH− → SO42− + •OH (9)

3.2.2. Degradation pathway and toxicity assessment

Reactive species SO4•− , •OH and Fe(IV) were proven to coexist in the
BQ/Fe(III)/PMS system, and they contributed to ATZ degradation and
accounted for intermediates formation. In this regard, degradation in­
termediates of ATZ were detected by LC-MS, and three dechlorinated
and five dealkylated byproducts were identified, respectively. Their
chemical names, abbreviations, m/z values and structures are shown in
Table S3, and the LC-MS spectra of eight identified degradation by-
products are provided in Fig. S2. These intermediates indicated that
ATZ degradation was mainly controlled by dealkylation and
dechlorination-hydroxylation, accompanied by alkylic-oxidation and
deamination-hydroxylation (as shown in Fig. 4). As an electrophilic
oxidant, SO4•− tends to attack the electron-rich part of the target
pollutant, thus the C-N bond of alkylic lateral chain of ATZ can be easily
removed to generate dealkylated products. Dealkylation can happen
Fig. 3. The proposed activation mechanism of the BQ/Fe(III)/PMS system. alone to produce DEA (m/z, 188), DIA (m/z, 174) and DEIA (m/z, 146),
or it can co-occur with other reactions. For example, deamination-
and induced the formation of SO4•− and •OH via Eqs. (2) and (3). Be­ hydroxylation replacing amino groups with hydroxyl groups occurred
sides, •OH can also be produced through the reactions of SO4•− with after the dealkylation reaction and made DEA form COIT (m/z, 189). In
H2O/OH− (Eqs. (8) and (9)). On the other hand, the reaction of Fe(II) addition, alkylic-oxidation resulting from H-abstraction
and PMS can yield Fe(IV) (Eq. (5)), which was finally converted to Fe by SO4•− /•OH can lead to the formation of CDAT (m/z, 188) [1]. Note
(III) via side reactions with Fe(II), H2O or PMS (Eq. (6)). It should be that the C-N bond length of ethyl group is longer than that of isopropyl,
noted that in this process, Fe(II) was continuously regenerated by BQ thus the deethyl reaction is more likely to happen than the deisopropyl
and then was consumed by PMS, generating multiple reactive species reaction [36]. This is the reason why DEA concentration was always
and consequently degrading ATZ. much higher than that of DIA in the BQ/Fe(III)/PMS system. Compared
with the dealkylation reaction, dechlorination-hydroxylation replacing
SO4•− + H2O → SO42− + •OH + H+ (8) hydroxyl group with chlorine atom is more likely to occur because the
length of C-Cl bond is the longest (1.734 Å) and the cleavage of C-Cl

5
X. Li et al. Chemical Engineering Journal 427 (2022) 131995

Fig. 4. Potential degradation pathways of ATZ in the BQ/Fe(III)/PMS system. Reaction condition: [ATZ] = 50 μM, [Fe(III)] = 0.20 mM, [PMS] = 0.25 mM, [BQ] =
0.25 mM, pH = 3.

bond is the easiest [37]. In this study, dechlorinated byproducts OAIT products possess a much lower bioaccumulation factor than ATZ, with
(m/z, 170), OOAT (m/z, 129) and OAAT (m/z, 128) were detected, being decreased by 92.6% (OAAT), 91.6% (CDAT), 91.3% (OAIT),
confirming that ATZ underwent dechlorination-hydroxylation. Howev­ 89.4% (OOAT), 80.5% (DEIA), 73.6% (DIA), 59.8% (DEA) and 46.8%
er, there are two points to be noted. One was that these three in­ (COIT), respectively. Fig. 5(C) shows that ATZ itself is a “developmental
termediates were produced under the joint action of dechlorination- toxicant”, while the BQ/Fe(III)/PMS process not only can lead to the
hydroxylation and dealkylation/deamination-hydroxylation, further formation of both “developmental non-toxicant” (i.e., OAIT, DIA and
confirming that dealkylation can occur in coordination with other pro­ CDAT) and “developmental toxicant” (OAAT, DEIA, OOAT, DEA and
cesses. The second was that the initial dechlorinated byproduct 2-hy­ COIT). The mutagenicity result in Fig. 5(D) shows that ATZ itself is
droxy-4-ethylamino-6-isopropylamino-s-triazine (HEIT, produced only mutagenicity negative and the mutagenicity of most by-products (except
through the dechlorination-hydroxylation of ATZ) was not detected in COIT) can be decreased after the BQ/Fe(III)/PMS process. The overall
this study probably because it was rapidly oxidized by the subsequent toxicity assessment results suggested that the degradation of ATZ was
reactions. accompanied by the formation of toxic by-products, therefore, we not
Moreover, the Toxicity Estimation Software Tool (T.E.S.T.) was used only need to focus on the degradation of parent ATZ but also pay
to assess the acute toxicity, bioaccumulation factor, developmental attention to the elimination of by-products to guarantee the water safety
toxicity and mutagenicity of ATZ and the eight degradation by-products after treatment.
through quantitative structure–activity relationship (QSAR) [38]. Fig. 5
(A) shows that the LD50 of ATZ for rat is 1370.80 mg kg− 1, which is 3.3. Crucial factors influencing the catalytic performance
much lower than those of COIT (3331.73 mg kg− 1), OOAT (2693.30 mg
kg− 1) and OAIT (2586.51 mg kg− 1). The LD50 value of DEA and DIA is Based on the above discussion, BQ could indeed accelerate the
equivalent to that of ATZ, which is consistent with the “comparable conversion of Fe(III) to Fe(II), and achieve the goal of ATZ degradation
toxicity of these three” reported in literature [39]. It is undeniable that by reconstructing the Fe(II)/PMS system and generating reactive spe­
more toxic DEIA and CDAT can also be produced after the BQ/Fe(III)/ cies, but there were also some issues: One was how the catalytic per­
PMS process. It can be seen from Fig. 5(B) that all degradation by- formance of the BQ/Fe(III)/PMS system was affected by crucial reaction

6
X. Li et al. Chemical Engineering Journal 427 (2022) 131995

Fig. 5. (A) Acute toxicity, (B) Bioaccumulation factor, (C) Developmental toxicity and (D) Mutagenicity of ATZ and eight degradation by-products.

parameters (including solution pH, Fe(III) dose, PMS concentration)? species of Fe. Thus, ATZ degradation experiments at eight different pH
The other one was whether this system could remove part of the highly levels in the BQ/Fe(III)/PMS system were performed. Note that the pKa
toxic BQ while degrading ATZ? Therefore, a series of experiments were values of PMS are less than 0 and 9.4 [40], thus PMS mainly existed as
conducted to address these problems. HSO5− within the pH investigated (from 2.5 to 7.0). However, the form
of Fe and the degradation of ATZ varied with solution pH. Fig. 6(A)
3.3.1. Solution pH shows the final ATZ degradation efficiency, it can be seen that the
Solution pH can not only effect the form of PMS, but also mediate the optimal efficiency was achieved at pH 2.5 and 3.0, but ATZ degradation

Fig. 6. ATZ degradation efficiency in the BQ/Fe(III)/PMS system (A) under different pH conditions and (B) at various Fe(III) amounts. Reaction condition: (A) [ATZ]
= 50 μM, [Fe(III)] = 0.20 mM, [PMS] = 0.25 mM, [BQ] = 0.25 mM; (B) [ATZ] = 50 μM, [PMS] = 0.25 mM, [BQ] = 0.25 mM, pH = 3 .

7
X. Li et al. Chemical Engineering Journal 427 (2022) 131995

process at pH 2.5 was slower than that at pH 3.0 (See the curve of ATZ elimination. Fe (III) widely exists in natural and actual water bodies,
degradation efficiency with reaction time in Fig. S3). This “inverse H+- which indicates that the system has certain practical applicability.
dependency” phenomenon might be due to the following two aspects: However, it should be noted that the optimal Fe(III) concentration was
one was that the formation of H-bond between H+ and O–O group of 0.2 mM, too low or excessive Fe(III) content was detrimental to the
PMS was more obvious at acidic pH [40], making PMS positively catalytic degradation process: When Fe(III) dose was as low as 0.005
charged and hindered its interaction with Fe(II), the other was that Fe mM, ATZ degradation was slower than other cases, which might be
(II) was dominant in strong acid condition, while more reactive Fe(OH)2 because when Fe(III) was added in a low concentration, only a small
tended to form as pH rose [41]. When solution pH value was above 3.0, amount of Fe(II) was regenerated by BQ, resulting in slow PMS activa­
obvious inhibition of ATZ degradation was observed, which probably tion and low ATZ degradation. Once Fe(III) concentration increased
because Fe species intended to precipitate in the form of oxyhydroxides from 0.20 to 0.25 or even 0.50 mM, ATZ degradation process slowed
and decrease the availability of Fe(II). Although ATZ degradation was down. This might be explained by the fact that a large amount of Fe(II)
suppressed, its efficiency could still reach 60% or more probably due to was regenerated when added Fe(III) was too high, and excessive Fe(II)
the continuous regeneration of Fe(II) and production of reactive species could quench SO4•− , •OH and Fe(IV) via Eqs. (10)–(12) [6]{Cheng, 2017
(89.8%, 79.9%, 76.0%, 72.2%, 75.0% and 63.0% at pH 3.5, 4.0, 4.5, 5.0, #25;Wang, 2018 #19}, causing these reactive species to be consumed in
6.0 and 7.0, respectively). vain. Therefore, it is necessary to pay close attention to Fe(III) concen­
tration to make it more conducive to the degradation of contaminants.
3.3.2. Fe(III) amount Nevertheless, the BQ/Fe(III)/PMS system is still very promising, i.e.,
To investigate how different Fe(III) levels in water affected the cat­ even if the Fe(III) concentration is as low as 0.005 mM, it has the po­
alytic degradation process, Fe(III) concentration ranging from 0.005 mM tential to continuously generate reactive species and degrade pollutants.
to 0.50 mM was introduced to the BQ/Fe(III)/PMS system. Note that
0.005 mM Fe(III) was determined according to GB5749-2006 “Stan­ Fe(II) + SO4•− → Fe(III) + SO42− (10)
dards for drinking water quality in China”, where the Fe content in •
Fe(II) + OH → Fe(III) + OH −
(11)
drinking water is not allowed to exceed 0.3 mg L− 1 (0.005 mM Fe(III)
corresponding to 0.28 mg L− 1 Fe(III)). The purpose of setting 0.50 mM Fe(II) + FeIVO2+ + (2H+) → 2 Fe(III) + H2O (12)
Fe(III) was to evaluate whether the BQ/Fe(III)/PMS system can still
obtain satisfactory degradation efficiency under relatively high Fe(III)
amount (compared to the optimal concentration of 0.20 mM Fe(III)). As
3.3.3. PMS dose
shown in Fig. 6(B), the BQ/Fe(III)/PMS system could effectively degrade
As the source of SO4•− and •OH, PMS concentration had a marked
ATZ regardless of Fe(III) concentration, which meant that as long as Fe
impact on the catalytic oxidation process. The influence of PMS con­
(III) was contained in the reaction system, BQ could continuously reduce
centration on contaminant degradation usually showed the trend of
it to reactive Fe(II) and rebuild the Fe(II)/PMS system for contaminants
increasing first and then decreasing, similar results were obtained in our

Fig. 7. (A) ATZ degradation efficiency, (B) DEA concentration, (C) DIA concentration and (D) BQ degradation efficiency in the BQ/Fe(III)/PMS system at different
PMS doses. Reaction condition: [ATZ] = 50 μM, [Fe(III)] = 0.20 mM, [BQ] = 0.25 mM, pH = 3.

8
X. Li et al. Chemical Engineering Journal 427 (2022) 131995

study. As shown in Fig. 7(A), the optimal PMS concentration was 0.20 chlorinated intermediates formation in the BQ/Fe(III)/PMS system were
mM, with ATZ degradation efficiency reaching up to 99.7%. Although explored. Cl− concentration range was selected between 0 and 1.0 mM
regenerative Fe(II) was enough, too low PMS dose (i.e., 0.05 and 0.10 according to Cl− concentration of different actual water bodies reported
mM) also failed to produce sufficient reactive species, the unchanged in literature (Table S4). It has been reported that Cl− can not only
ATZ degradation in later stage reaction was the proof. ATZ degradation directly react with PMS to generate Cl2 and HOCl through Eqs. (13)–(16)
trend with excessive PMS concentration (i.e., 0.25 and 0.50 mM) was [45], but also quench SO4•− and •OH to form less reactive Cl• and Cl2•−
almost the same as that with 0.20 mM, suggesting that the limiting through Eqs. (17)–(22) [46]. Thus, when Cl− was contained in the BQ/
factor was regenerative Fe(II) rather than PMS dose. Fe(III)/PMS system, SO4•− , •OH, Fe(IV), Cl•, Cl2•− , Cl2 and HOCl
In addition to the parent pollutant ATZ, chlorinated byproducts DEA simultaneously existed and worked together to attack ATZ. As shown in
and DIA identified in the BQ/Fe(III)/PMS system have comparable Fig. 8(A), the final ATZ degradation efficiency gradually decreased with
toxicity to ATZ [42], thus their concentration changes were also inves­ the increase of Cl− concentration, dropping from ~100% to 96.4%,
tigated via HPLC analysis. In general, DEA concentration (Fig. 7(B)) was 91.3%, 86.8%, 83.1% and 76.1% in the presence of 0–1.0 mM Cl− . But
far greater than that of DIA (Fig. 7(C)) during the whole reaction, meanwhile, as Cl− concentration rose, ATZ degradation was accelerated
probably because deethyl is more likely to occur than deisopropyl [43]. and the time to reach the maximum degradation was gradually
When PMS dose was 0.05 and 0.10 mM, the concentration of DEA and advanced, after which ATZ degradation was kept constant. The accel­
DIA fluctuated during the first 30 and 40 min, respectively, but changed eration and then stabilization of ATZ degradation process was probably
little in the subsequent reaction. This phenomenon was consistent with due to the abundant production and rapid consumption of reactive
the ATZ degradation results in Fig. 7(A), further proving that no species via PMS activation by both Cl− and regenerative Fe(II). Subse­
excessive reactive species for the elimination of ATZ and its in­ quently, the concentrations of DEA and DIA were detected and their
termediates. In the cases of 0.20, 0.25 and 0.50 mM PMS, the concen­ changes were shown in Fig. 8(B) and Fig. 8(C), respectively. For
tration of DEA and DIA generally showed the trend of increasing first example, except for 1.0 mM Cl− , DEA concentration in other cases
and then decreasing, and as the PMS concentration increased, the time showed a trend of “increasing first, then decreasing and finally stabi­
point where DEA and DIA reached the maximum concentration was lizing”. This was because DEA accumulated rapidly and reached to its
gradually advanced. In detail, 0.50 mM PMS made DEA concentration maximum level due to the removal of ethyl group of ATZ, and then the
quickly accumulate to the maximum value of 13.88 μM at 30 min, faster formed DEA was further oxidized by SO4•− and •OH, causing its con­
than that of 0.20 mM (60 min) and 0.30 mM (40 min) kSO4•− /DEA. This centration to decline until it no longer changed. For different Cl− con­
accumulation was due to the deethylation reaction occurring on ATZ. As centrations, the maximum concentration of DEA was almost the same,
the reaction proceeding, DEA concentration rapidly dropped and even to but the time to reach its maximum value gradually advanced with the
undetectable level, probably because the generated DEA was further increase of Cl− concentration, shifting from 60 min (0.2 mM Cl− ) to 40
oxidized by SO4•− and •OH (kSO4•− /DEA = (6.42 ± 0.12) × 108 M− 1 s− 1 min (0.4 mM Cl− ) and 30 min (0.6 and 0.8 mM Cl− ). This forward shift
and k•OH/DEA = (1.14 ± 0.09) × 109 M− 1 s− 1) [42]. For DIA, although its was closely related to ATZ degradation and reactive species production,
concentration change was not as obvious as DEA, it also followed a i.e., the higher Cl− concentration was added, the more reactive species
similar rule, that is, the higher PMS dose was added, the faster time point were generated, and the faster ATZ was degraded and DEA was formed.
was reached, and the more formed DIA was removed (kSO4•− /DIA = Simliar phenomenon was also obtained in the case of DIA, but the dif­
(1.70 ± 0.30) × 109 M− 1 s− 1 and k•OH/DIA = (2.22 ± 0.44) × 109 M− 1 ference was that the trend of DIA was not as obvious as DEA, which was
s− 1) [42]. probably due to the yield of DIA was much lower than that of DEA (Fig. 8
The BQ/Fe(III)/PMS system has been confirmed to degrade ATZ and (D), the deep reason of lower yield of DIA was that deisopropylation is
its intermediates, so the other question is coming: Although BQ could harder to occur than deethylation). It should be noted that in the case of
promote the regeneration of Fe(II), it is also a highly toxic organic 1.0 mM Cl− , the concentration of DEA and DIA gradually increased and
pollutant, whether it could also be removed while acting as electron reached a relatively stable value as the reaction progressed, which might
mediator in the BQ/Fe(III)/PMS system? For this problem, considering be attributed to that most of the produced active species were used to
that PMS was the source of reactive species, thus BQ removal at different degrade ATZ, and there were not much left to oxidize DEA and DIA.
PMS concentrations were investigated. As shown in Fig. 7(D), in all
cases, the loss of BQ occurred, and BQ removal efficiency increased with Cl− + HSO5− → HOCl + SO42− (13)
the increase of PMS concentration. Specifically, when PMS concentra­ 2Cl− + HSO5− + 2H+ → Cl2 + HSO4− + H2O (14)
tion was 0.05, 0.10, 0.20, 0.25 and 0.50 mM, BQ removal efficiency in
120 min was 4.5%, 5.7%, 11.4%, 21.0% and 40.0%, respectively. For Cl2 + H2O → H+ + HOCl + Cl− (15)
low PMS concentration (i.e., 0.05 and 0.10 mM), ATZ degradation
HOCl → H + OCl + −
(16)
varied greatly but BQ removal was not much different, which might be
attributed to the insufficient reactive species. In the case of higher PMS SO4 •−
+ Cl → SO4 − 2−
+ Cl •
(17)
concentration, ATZ was usually rapidly and completely degraded in the
first 60 min, while BQ showed a low removal efficiency. However, in the
•
OH + Cl → ClOH − •−
(18)
subsequent reaction process, BQ removal sped up due to the attack of ClOH•− + H+ → Cl• + H2O (19)
reactive species, such as SO4•− and •OH. These results showed that like
ATZ, BQ can also be removed by the Fe(III)/BQ/PMS process. Therefore,
•
Cl + Cl → Cl2 − •−
(20)
BQ played two roles in this system, one was to act as an electron
Cl2 •−
+ Cl2•− −
→ 2Cl + Cl2 (21)
mediator for Fe(II) regeneration, and the other was to be removed as a
pollutant. •
Cl + Cl → Cl2 •
(22)

3.3.4. Cl− concentration

In addition to reaction parameters, environmental factors such as
coexisting ions may also have an important impact on the reaction 3.4. Potential of different phenols/quinones for Fe(II) regeneration
system. Among them, Cl− is widely present in water bodies and previous
studies have reported that the presence of Cl− will induce the production To evaluate the potential of phenols/quinones to accelerate Fe(III)/
of chlorinated byproducts during the catalytic degradation process [44]. Fe(II) cycle, seven typical substances including phenol, p-cresol, 4-
Thus, the effects of different Cl− concentrations on ATZ degradation and methoxyphenol, hydroquinone (HQ), p-benzoquinone (BQ), methyl-p-

9
X. Li et al. Chemical Engineering Journal 427 (2022) 131995

Fig. 8. (A) ATZ degradation efficiency, (B) DEA concentration, (C) DIA concentration and (D) The yield of DEA or DIA in the BQ/Fe(III)/PMS system at different Cl−
concentrations. Reaction condition: [ATZ] = 50 μM, [Fe(III)] = 0.20 mM, [PMS] = 0.25 mM, [BQ] = 0.25 mM, pH = 3.

Fig. 9. (A) ATZ degradation efficiency in the Fe(III)/PMS system with seven phenols/quinones, (B) A510 nm value in the Fe(III)/PMS system with hydroquinone and
three quinones; Electron cloud density maps of (C) benzoquinone (BQ), (D) methyl-p-benzoquinone (MBQ) and (E) methoxybenzoquinone (MOBQ). Reaction
condition: [ATZ] = 50 μM, [Fe(III)] = 0.20 mM, [PMS] = 0.25 mM, [phenols/quinones] = 0.25 mM, pH = 3.

10
X. Li et al. Chemical Engineering Journal 427 (2022) 131995

benzoquinone (MBQ) and methoxy-p-benzoquinone (MOBQ) were existence of quinones and Fe(III) provides convenient conditions for the
introduced into the Fe(III)/PMS system for ATZ degradation. Their practical application of the BQ/Fe(III)/PMS system, because only adding
structural formulas were shown in Table S5. As shown in Fig. 9(A), only PMS into the water body can achieve the purpose of pollutants degra­
23.2% ATZ was degraded by the Fe(III)/PMS process in 120 min due to dation, which also saves the economic cost to a certain extent. Never­
that the slow regeneration of Fe(II) from Fe(III) hindered the successive theless, some issues still need attention in the actual application process.
PMS activation and ATZ degradation. However, with the addition of One is whether the quinones in the surface water are sufficient: Rely on
quinones, ATZ degradation efficiency increased to 26.4% (MOBQ), the inherent quinones in water for catalytic degradation as much as
97.6% (MBQ) and 99.8% (BQ), respectively. Notably, compared with possible, so as to reduce the extra addition of quinones and the pollution
BQ, MBQ could accelerate the catalytic degradation faster, and ATZ load of water body. If it is necessary to add extra quinones, the type and
degradation in the MBQ/Fe(III)/PMS system was completed only in 20 content of quinones should be determined according to the water quality
min. HQ, an important electron transfer mediator capable of mediating characteristics. The other one is the impact of water quality parameters
Fe(III)/Fe(II) cycle [47], was also proven to induce Fe(II) regeneration and water safety assessment. Although the BQ/Fe(III)/PMS system can
and promote ATZ degradation (45.9%) in this study. In contrary, the achieve decent degradation efficiency in the pH range of 2.5–7.0, the pH
other three phenols not only failed to promote the catalytic oxidation of surface water is usually maintained in a near-neutral or weak- alkaline
process, but also inhibited the degradation of ATZ. This may be due to condition. Whether to adjust the pH of water body and the possible
the structural differences between quinones and phenols, making these problems after pH adjustment are the key points to be considered. In
phenols unable to participate in electron-transfer reaction with Fe like addition, the effects of coexisting substances on the degradation of
quionones, and can only act as pollutants and compete with ATZ for target pollutants and the type/content of formed intermediates as well
reactive species. as the safety assessment (including water quality and toxicity) should
Furtherly, Fe(II) regeneration in the Fe(III)/PMS system containing also be focused on.
0.25 mM quinones or HQ was evaluated and shown in Fig. 9(B). The
more Fe(II) regeneration, the greater A510nm value. It can be seen that 4. Conclusion
A510nm values of three quinones always followed the order of MBQ >
MOBQ > BQ during the reaction process, preliminarily suggesting MBQ In this study, quinons were proven to accelerate Fe(II) regeneration
and MOBQ had higher ability to regenerate Fe(II) compared with BQ. and enhance ATZ degradation in the Fe(III)/PMS system. In the presence
However, it should be noted that this order was not completely consis­ of 0.25 mM BQ, 99.8% of ATZ was degraded in 120 min under the
tent with that of MBQ > BQ > MOBQ in ATZ degradation (see Fig. 9(A)). condition of 0.20 mM Fe(III), 0.25 mM PMS and pH 3.0, much higher
The paradox was mainly reflected between MOBQ and BQ, and we than that without BQ (23.2%). SO4•− , •OH and Fe(IV) were identified to
speculated that the substituted functional groups may be the root of this coexist in the BQ/Fe(III)/PMS system. ATZ degradation mechanism was
contradiction. Theoretically, electron cloud density on cyclohexadiene clarified on the basis of LC-MS detection, and dealkylation and
dione of MOBQ and MBQ was higher than that of BQ due to the electron- dechlorination-hydroxylation were proposed to be the main trans­
donating ability of − OCH3 and − CH3 groups. The electron cloud density formation pathways. The BQ/Fe(III)/PMS system achieved >60% ATZ
maps of the three quinones in Fig. 9(C–E) indeed show that the sub­ degradation efficiency in the pH range of 2.5 to 7.0, and maintained
stitutions of − OCH3 and − CH3 groups have changed the electron cloud satisfactory catalytic performance even if Fe(III) concentration was as
density distribution of the molecule, and as a stronger electron donating low as 0.005 mM (0.28 mg L− 1) due to continuous Fe(II) regeneration
group, the presence of − OCH3 group has a more significant effect on it. and PMS activation. ATZ degradation, DEA and DIA oxidation as well as
Based on these, we speculated that the same concentration of MOBQ BQ removal were enhanced with the increase of PMS concentration. The
tended to regenerate more Fe(II) than BQ, that was, the optimal MOBQ presence of Cl− suppressed ATZ degradation and caused chlorinated
concentration should be lower than BQ. To confirm this conjecture, byproducts accumulation, and generally followed the rule of “the higher
0.02–0.25 mM MOBQ was introduced into the Fe(III)/PMS system. As Cl− was added, the less ATZ was degraded, and the more DEA and DIA
shown in Fig. S4(A), ATZ degradation gradually increased when MOBQ were produced”. Although BQ was proved to enhance the efficiency of
enlarged from 0.02 mM to 0.15 mM (32.3%, 38.5%,68.4% and 76.6% at the Fe(III)/PMS system via accelerating Fe(III)/Fe(II) cycle, the inter­
0.02, 0.05, 0.10 and 0.15 mM, respectively), suggesting that 0.15 mM mediate products and transformation pathways of BQ during the process
was the optimum amount for Fe(II) regeneration and PMS activation. need further study. The structure–activity relationship between qui­
However, ATZ degradation efficiency dramatically dropped to 26.4% nones with different functional groups and Fe(II) regeneration should
when increased MOBQ concentration to 0.25 mM, which might be also be clarified, and whether these quinones can boost the surface Fe(II)
attributed to that a large amount of Fe(II) was regenerated by 0.25 mM generation of heterogeneous Fe-based catalysts requires special
MOBQ (corresponding to MOBQ > BQ in A510 nm value), but excessive Fe attention.
(II) could quench reactive species (Eqs. (7)–(9)) and make them to be
consumed in vain, consequently inhibiting the degradation of ATZ Declaration of Competing Interest
(corresponding to BQ > MOBQ in ATZ degration). Coincidentally, HQ
also showed the contradictory result of “high Fe(II) regeneration but The authors declare that they have no known competing financial
limited ATZ degradation” (see Fig. 9(A) and Fig. 9(B)), and similar interests or personal relationships that could have appeared to influence
phenomenon to MOBQ was observed in the Fe(III)/PMS system at the work reported in this paper.
different HQ concentrations, with the optimum HQ addition amount
being 0.10 mM (Fig. S4(B)). These results further confirmed that low Acknowledgments
concentration made little contribution to ATZ degradation due to
limited enhancement for Fe(III)/Fe(II) cycle, while high dose showed This study was supported by the National Key Research and Devel­
negative effect due to the side reactions between excessive regenerated opment Program of China (Project No. 2019YFC1805602), the National
Fe(II) and reactive species. Natural Science Foundation of China (Project No. 41671487) and the
Key Research and Development Program of Gansu Province (Project No.
3.5. Application prospect of the BQ/Fe(III)/PMS system 20YF8GA113).

Considering that quinones widely exist in surface water, which Appendix A. Supplementary data
usually also contains Fe(III), the BQ/Fe(III)/PMS system may be more
suitable for the restoration of contaminated surface water. The natural Supplementary data to this article can be found online at https://doi.

11
X. Li et al. Chemical Engineering Journal 427 (2022) 131995

org/10.1016/j.cej.2021.131995. [24] A.E. Harvey, J.A. Smart, E.S. Amis, Simultaneous spectrophotometric
determination of iron(II) and total iron with 1,10-phenanthroline, Anal Chem. 27
(1) (1955) 26–29.
References [25] C. Liang, C.F. Huang, N. Mohanty, R.M. Kurakalva, A rapid spectrophotometric
determination of persulfate anion in ISCO, Chemosphere 73 (2008) 1540–1543.
[1] J.A. Khan, X. He, N.S. Shah, H.M. Khan, E. Hapeshi, D. Fatta-Kassinos, D. [26] X. Li, X. Liu, C. Lin, Z. Zhou, M. He, W. Ouyang, Catalytic oxidation of
D. Dionysiou, Kinetic and mechanism investigation on the photochemical contaminants by Fe0 activated peroxymonosulfate process: Fe(IV) involvement,
degradation of atrazine with activated H2O2, S2O82− and HSO5− , Chem. Eng. J. 252 degradation intermediates and toxicity evaluation, Chem. Eng. J. 382 (2020),
(2014) 393–403. 123013.
[2] M.Y. Huang, Q. Zhao, R.Y. Duan, Y. Liu, Y.Y. Wan, The effect of atrazine on [27] J. Cao, L. Lai, B. Lai, G. Yao, X. Chen, L. Song, Degradation of tetracycline by
intestinal histology, microbial community and short chain fatty acids in Pelophylax peroxymonosulfate activated with zero-valent iron: performance, intermediates,
nigromaculatus tadpoles, Environ. Pollut. 288 (2021), 117702. toxicity and mechanism, Chem. Eng. J. 364 (2019) 45–56.
[3] E. Hu, H. Cheng, Y. Hu, Microwave-induced degradation of atrazine sorbed in [28] P. Neta, R.E. Huie, A.B. Ross, Rate constants for reactions of inorganic radicals in
mineral micropores, Environ. Sci. Technol. 46 (9) (2012) 5067–5076. aqueous solution, J. Phys. Chem. Ref. Data. 17 (3) (1988) 1027–1284.
[4] W. Huang, S. Xiao, H. Zhong, M. Yan, X. Yang, Activation of persulfates by [29] H. Lee, H.-J. Lee, J. Seo, H.-E. Kim, Y.K. Shin, J.-H. Kim, C. Lee, Activation of
carbonaceous materials: a review, Chem. Eng. J. 418 (2021), 129297. oxygen and hydrogen peroxide by copper(II) coupled with hydroxylamine for
[5] G.P. Anipsitakis, D.D. Dionysiou, Radical generation by the interaction of transition oxidation of organic contaminants, Environ. Sci. Technol. 50 (15) (2016)
metals with common oxidants, Environ. Sci. Technol. 38 (13) (2004) 3705–3712. 8231–8238.
[6] Z. Wang, J. Jiang, S. Pang, Y. Zhou, C. Guan, Y. Gao, J. Li, Y.I. Yang, W. Qiu, [30] Q. Ye, H. Xu, J. Zhang, Q. Wang, P. Zhou, Y. Wang, X. Huang, X. Huo, C. Liu, J. Lu,
C. Jiang, Is sulfate radical really generated from peroxydisulfate activated by iron Enhancement of peroxymonosulfate activation for antibiotics removal by nano
(II) for environmental decontamination? Environ. Sci. Technol. 52 (19) (2018) zero valent tungsten induced Cu(II)/Cu(I) redox cycles, Chem. Eng. J. 382 (2020),
11276–11284. 123054.
[7] S. Wang, J. Wu, X. Lu, W. Xu, Q. Gong, J. Ding, B. Dan, P. Xie, Removal of [31] X. Yuan, A.N. Pham, G. Xing, A.L. Rose, T.D. Waite, Effects of pH, Chloride, and
acetaminophen in the Fe2+/persulfate system: Kinetic model and degradation Bicarbonate on Cu(I) Oxidation Kinetics at Circumneutral pH, Environ. Sci.
pathways, Chem. Eng. J. 358 (2019) 1091–1100. Technol. 46 (3) (2012) 1527–1535.
[8] D. Li, D. Chen, Y. Yao, J. Lin, F. Gong, L. Wang, L. Luo, Z. Huang, L.i. Zhang, Strong [32] H. Zhang, X. Liu, J. Ma, C. Lin, C. Qi, X. Li, Z. Zhou, G. Fan, Activation of
enhancement of dye removal through addition of sulfite to persulfate activated by a peroxymonosulfate using drinking water treatment residuals for the degradation of
supported ferric citrate catalyst, Chem. Eng. J. 288 (2016) 806–812. atrazine, J. Hazard. Mater. 344 (2018) 1220–1228.
[9] A. Rastogi, S.R. Al-Abed, D.D. Dionysiou, Sulfate radical-based ferrous- [33] S. Yang, P. Wu, J. Liu, M. Chen, Z. Ahmed, N. Zhu, Efficient removal of bisphenol a
peroxymonosulfate oxidative system for PCBs degradation in aqueous and by superoxide radical and singlet oxygen generated from peroxymonosulfate
sediment systems, Appl. Catal. B: Environ. 85 (2009) 171–179. activated with Fe0-montmorillonite, Chem. Eng. J. 350 (2018) 484–495.
[10] J. Zou, J. Ma, L. Chen, X. Li, Y. Guan, P. Xie, C. Pan, Rapid acceleration of ferrous [34] Z. Wang, W. Qiu, S. Pang, Y. Gao, Y. Zhou, Y. Cao, J. Jiang, Relative contribution of
iron/Peroxymonosulfate oxidation of organic pollutants by promoting Fe(III)/Fe ferryl ion species (Fe(IV)) and sulfate radical formed in nanoscale zero valent iron
(II) cycle with hydroxylamine, Environ. Sci. Technol. 47 (20) (2013) activated peroxydisulfate and peroxymonosulfate processes, Water Res. 172
11685–11691. (2020), 115504.
[11] J. Li, Y. Wan, Y. Li, G. Yao, B. Lai, Surface Fe(III)/Fe(II) cycle promoted the [35] Z. Wang, W. Qiu, S.-Y. Pang, Y. Zhou, Y. Gao, C. Guan, J. Jiang, Further
degradation of atrazine by peroxymonosulfate activation in the presence of understanding the involvement of Fe(IV) in peroxydisulfate and
hydroxylamine, Appl. Catal. B: Environ. 256 (2019), 117782. peroxymonosulfate activation by Fe(II) for oxidative water treatment, Chem. Eng.
[12] H. Zhou, H. Zhang, Y. He, B. Huang, C. Zhou, G. Yao, B. Lai, Critical review of J. 371 (2019) 842–847.
reductant-enhanced peroxide activation processes: trade-off between accelerated [36] N.a. Ta, J. Hong, T. Liu, C. Sun, Degradation of atrazine by microwave-assisted
Fe3+/Fe2+ cycle and quenching reactions, Appl. Catal. B: Environ. 286 (2021), electrodeless discharge mercury lamp in aqueous solution, J. Hazard. Mater. 138
119900. (1) (2006) 187–194.
[13] Y. Lei, H. Zhang, J. Wang, J. Ai, Rapid and continuous oxidation of organic [37] C. Chen, S. Yang, Y. Guo, C. Sun, C. Gu, B. Xu, Photolytic destruction of endocrine
contaminants with ascorbic acid and a modified ferric/persulfate system, Chem. disruptor atrazine in aqueous solution under UV irradiation: products and
Eng. J. 270 (2015) 73–79. pathways, J. Hazard. Mater. 172 (2-3) (2009) 675–684.
[14] Y. Zhou, X. Wang, C. Zhu, D.D. Dionysiou, G. Zhao, G. Fang, D. Zhou, New insight [38] Z. Cai, X. Hao, X. Sun, P. Du, W. Liu, J. Fu, Highly active WO3@anatase-SiO2
into the mechanism of peroxymonosulfate activation by sulfur-containing aerogel for solar-light-driven phenanthrene degradation: mechanism insight and
minerals: role of sulfur conversion in sulfate radical generation, Water Res. 142 toxicity assessment, Water Res. 162 (2019) 369–382.
(2018) 208–216. [39] F.J. Beltrán, M. González, F.J. Rivas, P. Alvarez, Aqueous UV radiation and UV/
[15] H. Zhou, J. Peng, J. Li, J. You, L. Lai, R. Liu, Z. Ao, G. Yao, B. Lai, Metal-free black- H2O2 oxidation of atrazine first degradation products: Deethylatrazine and
red phosphorus as an efficient heterogeneous reductant to boost Fe3+/Fe2+ cycle deisopropylatrazine, Environ. Toxicol. Chem. 15 (6) (1996) 868–872.
for peroxymonosulfate activation, Water Res. 188 (2021), 116529. [40] T. Zhang, H. Zhu, J.P. Croué, Production of sulfate radical from peroxymonosulfate
[16] R.-A. Doong, H.-C. Chiang, Transformation of carbon tetrachloride by thiol induced by a magnetically separable CuFe2O4 spinel in water: efficiency, stability,
reductants in the presence of quinone compounds, Environ. Sci. Technol. 39 (19) and mechanism, Environ. Sci. Technol. 47 (2013) 2784–2791.
(2005) 7460–7468. [41] L. Chen, J. Ma, X. Li, J. Zhang, J. Fang, Y. Guan, P. Xie, Strong enhancement on
[17] Y. Zhou, J. Jiang, Y. Gao, S.Y. Pang, Y. Yang, J. Ma, J. Gu, J. Li, Z. Wang, L. fenton oxidation by addition of hydroxylamine to accelerate the ferric and ferrous
H. Wang, L.P. Yuan, Y. Yang, Activation of peroxymonosulfate by phenols: iron cycles, Environ. Sci. Technol. 45 (9) (2011) 3925–3930.
important role of quinone intermediates and involvement of singlet oxygen, Water [42] J.A. Khan, X. He, N.S. Shah, M. Sayed, H.M. Khan, D.D. Dionysiou, Degradation
Res. 125 (2017) 209–218. kinetics and mechanism of desethyl-atrazine and desisopropyl-atrazine in water
[18] G. Fang, J. Gao, D.D. Dionysiou, C. Liu, D. Zhou, Activation of persulfate by with •OH and SO4•− based-AOPs, Chem. Eng. J. 325 (2017) 485–494.
quinones: free radical reactions and implication for the degradation of PCBs, [43] Y. Ji, C. Dong, D. Kong, J. Lu, Q. Zhou, Heat-activated persulfate oxidation of
Environ. Sci. Technol. 47 (2013) 4605–4611. atrazine: implications for remediation of groundwater contaminated by herbicides,
[19] Y. Zhou, J. Jiang, Y. Gao, J. Ma, S.Y. Pang, J. Li, X.T. Lu, L.P. Yuan, Activation of Chem. Eng. J. 263 (2015) 45–54.
peroxymonosulfate by benzoquinone: a novel nonradical oxidation process, [44] C. Fang, D. Xiao, W. Liu, X. Lou, J. Zhou, Z. Wang, J. Liu, Enhanced AOX
Environ. Sci. Technol. 49 (2015) 12941–12950. accumulation and aquatic toxicity during 2,4,6-trichlorophenol degradation in a
[20] H. Zhang, L. Qiao, J. He, N.a. Li, D. Zhang, K. Yu, H. You, J. Jiang, Activating Co(II)/peroxymonosulfate/Cl system, Chemosphere 144 (2016) 2415–2420.
peroxymonosulfate by halogenated and methylated quinones: performance and [45] G.P. Anipsitakis, D.D. Dionysiou, M.A. Gonzalez, Cobalt-mediated activation of
mechanism, RSC Adv. 9 (47) (2019) 27224–27230. peroxymonosulfate and sulfate radical attack on phenolic compounds. implications
[21] R. Chen, J.J. Pignatello, Role of quinone intermediates as electron shuttles in of chloride ions, Environ. Sci. Technol. 40 (3) (2006) 1000–1007.
fenton and photoassisted fenton oxidations of aromatic compounds, Environ. Sci. [46] T. Soltani, A. Tayyebi, B.-K. Lee, Quick and enhanced degradation of bisphenol A
Technol. 31 (8) (1997) 2399–2406. by activation of potassium peroxymonosulfate to SO4•− with Mn-doped BiFeO3
[22] C.K. Duesterberg, T.D. Waite, Kinetic modeling of the oxidation of p- nanoparticles as a heterogeneous Fenton-like catalyst, Appl. Surf. Sci. 441 (2018)
hydroxybenzoic acid by fenton’s reagent: implications of the role of quinones in 853–861.
the redox cycling of iron, Environ. Sci. Technol. 41 (2007) 4103–4110. [47] C. Jiang, S. Garg, T.D. Waite, Hydroquinone-mediated redox cycling of iron and
[23] X. Li, X. Liu, C. Lin, H. Zhang, Z. Zhou, G. Fan, J. Ma, Cobalt ferrite nanoparticles concomitant oxidation of hydroquinone in oxic waters under acidic conditions:
supported on drinking water treatment residuals: an efficient magnetic comparison with iron-natural organic matter interactions, Environ. Sci. Technol.
heterogeneous catalyst to activate peroxymonosulfate for the degradation of 49 (2015) 14076–14084.
atrazine, Chem. Eng. J. 367 (2019) 208–218.

12