Journal of Hazardous Materials 377 (2019) 290–298

Contents lists available at ScienceDirect

Journal of Hazardous Materials

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

Comparison of pre-oxidation between O3 and O3/H2O2 for subsequent T

managed aquifer recharge using laboratory-scale columns
Hyun-Chul Kima, Se Hee Parkb, Jin Hyung Nohb, Jaewon Choic, Seunghak Leed,
⁎
Sung Kyu Maengb,
a
Research Institute for Advanced Industrial Technology, College of Science and Technology, Korea University, 2511 Sejong-ro, Sejong, 30019, Republic of Korea
b
Department of Civil and Environmental Engineering, Sejong University, 209 Neungdongro, Gwangjin-gu, Seoul, 05006, Republic of Korea
c
Water Quality & Safety Research Center, K-water Institute, Shintanjin-Ro 200, Daeduck-Gu, Daejeon, 34350, Republic of Korea
d
Water Cycle Research Center, National Agenda Research Division, Korea Institute of Science and Technology, 5, Hwarang-ro 14-gil, Seongbuk-gu, Seoul, 02792, Republic
of Korea

G R A P H I C A L A B S T R A C T

A R T I C LE I N FO A B S T R A C T

Keywords: A hybrid process of managed aquifer recharge with pre-oxidation was investigated as part of a multiple-barrier
Hydrogen peroxide approach for safe water production. This study evaluated O3 and O3/H2O2 for the pre-oxidation of urban surface
Managed aquifer recharge water prior to managed aquifer recharge (MAR) and compared their effectiveness with respect to trace organic
Ozone contaminants (TrOCs), biostability, and trihalomethane formation potential. The combination of pre-oxidation
Pre-oxidation
and MAR was performed using long-term column studies, and the results confirmed the removal of 64 and 56%
dissolved organic carbon by using O3 and O3/H2O2, respectively. MAR combined with O3 and O3/H2O2
achieved > 50% removal of dissolved organic carbon with the first 5 days of residence time. O3 alone showed
better performance in alleviating trihalomethane formation potential during chlorination compared to using O3/
H2O2. The pre-oxidation of urban surface water was effective in attenuating selected TrOCs (35 - > 99% re-
moval), and subsequent MAR achieved > 99% removal of selected TrOCs within the first 5 days, regardless of
pretreatment methods examined in this study. The results of this study provide an understanding of the effects of
O3 and O3/H2O2 as pre-oxidation processes on urban surface water prior to MAR, as well as the resulting impact
on MAR.

⁎
Corresponding author.
E-mail address: smaeng@sejong.ac.kr (S.K. Maeng).

https://doi.org/10.1016/j.jhazmat.2019.05.099
Received 15 March 2019; Received in revised form 20 May 2019; Accepted 29 May 2019
Available online 30 May 2019
0304-3894/ © 2019 Elsevier B.V. All rights reserved.
H.-C. Kim, et al.
Fig. 1. Schematic of the experimental setup.
1. Introduction
Non-regulated trace organic contaminants (TrOCs) are ubiquitous in
urban surface water and associated with the possible ecological impact
(e.g., endocrine disruption) to biota within the environment [1]. TrOCs
presented in urban surface water are mainly derived from municipal
wastewater effluent, and many attempts have been made to reduce
their release to the environments [2–4]. Managed aquifer recharge
(MAR) is emerging as alternative solutions to be effective barriers for
TrOCs, bacteria, and disinfection byproduct precursors. A wide range of
methods are available for recharging water in the urban environment
via MAR, including riverbank filtration, aquifer recharge and recovery,
and soil aquifer treatment [5]. MAR basically offers several advantages
over direct surface water intake resulting from using physical, chemical
and biological degradation processes of aquifers [6,7]. Regarding or-
ganic removal, many studies have demonstrated the effectiveness of
MAR in alleviating the production of disinfection byproducts (DBPs)
and attenuating TrOCs [8–11].
MAR is an effective treatment in reducing many TrOCs, but some
compounds persist during soil passage, such as carbamazepine and
primidone [12]. Therefore, MAR alone cannot remove all TrOCs due to
the persistent characteristics of certain chemical compounds even with
long residence time (e.g., > 1 year) [13]. To improve MAR perfor-
mance, additional advanced water systems are recommended and
combined with MAR [14]. In recent years, there has been growing in-
terest in multiple barriers that have synergies in which two or more
coupled water treatment processes can function as a hybrid process for
the removal of TrOCs. O3-based oxidation has mainly been used as a
pretreatment prior to MAR, whereas a wide range of post-treatment
strategies (e.g., granular activated carbon adsorption, membrane fil-
tration, UV irradiation, and chlorination) have been combined with
MAR to enhance the removal of recalcitrant organic compounds
[15–17]. Pre-oxidation, such as with ozone prior to MAR to enhance the
removal of TrOCs, has been reported in previous studies [4,12,18,19].
Ozone is an effective treatment for the removal of TrOCs that consist of
double bonds, aromatic structures or amine groups [20]. There are
291
Journal of Hazardous Materials 377 (2019) 290–298
possible additional benefits such as the reduction in DBPs [11,21],
enhanced biodegradation by providing more dissolved oxygen (DO)
[18] and being an efficient barrier to transformation products from
TrOCs [3] that occur due to the hybridization with MAR. Peroxone is an
advanced oxidation process and is a combination of O3 and H2O2. The
addition of H2O2 during the ozone process forms the OH radical, which
has very high oxidation power and is nonselective, and is therefore
effective in decomposing chemical byproducts and parent compounds
[22]. Peroxone has been studied widely and is a well-established pro-
cess in drinking water treatment and water reuse applications [23,24].
To the best of the authors’ knowledge, O3 has been used for most pre-
oxidation studies for MAR, and only limited information is available for
O3/H2O2.
As mentioned above, some attempts have been made to improve
MAR performance by combining it with ozone in the past. To our
knowledge, the effect of pre-oxidation between O3 alone and O3/H2O2
on the performance of MAR with respect to biostability, bulk organic
matter, DBPs and TrOCs have not been addressed in previous studies. It
is important to investigate the effects of O3 and O3/H2O2, as the addi-
tion of H2O2 could enhance the degradation efficiency with OH radicals
for compounds that are relatively resistant to O3 oxidation (e.g., caf-
feine). It is easy to adapt the O3/H2O2 process for use with water uti-
lities if treatment plants already use an ozonation system. The main
objective of this study was to evaluate O3 alone and O3/H2O2 as pre-
oxidation agents for urban surface water prior to MAR with respect to
the removal of TrOCs, as well as to examine disinfection byproduct
formation potentials (DBPFPs) in long-term column studies. Moreover,
only a few studies have attempted to investigate the impact of residence
of time (i.e., hydraulic retention time (HRT)) in MAR with pre-oxida-
tion. Hübner et al. used a pilot-scale slow sand filter (HRT < 2 days)
for in the first phase of infiltration in MAR and compared it with a soil
column (HRT 5–6 days), which was designed to monitor longer re-
sidence times [3]. However, most MAR sites are operating at HRTs
longer than 5 days. The secondary goal was to evaluate the effects of
residence time up to 30 days in MAR combined with pre-oxidation (O3
and O3/H2O2). The performance of MAR was determined with respect
H.-C. Kim, et al.
to the fate of bulk organic matter characteristics and biostability.
2. Material and methods
2.1. Column experimental setup
Two soil columns containing silica sands were prepared to simulate
MAR (Fig. 1). Urban surface water from the Tancheon River (Seoul,
South Korea) was periodically collected and used as feed water in this
study. The columns were acclimated using untreated urban surface
river water under the same conditions for approximately 90 d. The
acclimation of soil matrixes continued until the level of effluent dis-
solved organic carbon (DOC) was stable. The feed water sample was
divided into two aliquots and was introduced into the soil columns after
two different pretreatments (O3 and O3/H2O2); one was ozonated at a
dose of 1.0 mg O3/mg DOC, and for the other, the identical ozone dose
was applied with added H2O2 (0.3 × 1.0 mg O3/mg DOC). In this study,
we selected the mass ratio of ozone to DOC based on the judgement of
previous studies [19,25]. Ozone was generated (Lab-I Ozone tech Co.
Ltd., South Korea) with pure oxygen (> 99.9%) and sparged into
deionized water to prepare stock solution on a daily basis. The ozone
concentration was calculated using the light absorbance of the stock
solution at 258 nm (ε = 3000 M−1 cm−1) [26], which ranged between
66 and 68 mg/L during the entire course of the experiment. The ozone
stock solution was spiked into the feed water sample to achieve the
ratio of 1.0 mg O3/mg DOC and no ozone remained after 1 h of contact
time in the feed tank. The hydraulic loading rate varied from 1.8 to
0.19 m/d, complying with each residence time (1, 5, 15, and 30 days).
In this study, the term “residence time” means the time span over which
the dose travels through the soil columns. However, each residence
time has a deviation of one day since it was difficult to control residence
time in columns for exactly 1, 5, 15 and 30 days. Moreover, the changes
in DOC in the feed bottles were minimized by changing the feed water
within every 48 h.
2.2. Dissolved organic matter characteristics
The water samples were filtered with 0.45 μm filters (Whatman,
USA) prior to measuring the water quality. The DOC was determined
(TOC-V CPN, Shimadzu, Japan) and the UV254 absorbance was de-
termined using a DR5000 spectrophotometer (Hach, USA). Specific UV
absorbance (SUVA) was calculated from the UV254 divided by the DOC
of the water sample. The molecular weight (MW) distribution of dis-
solved organic matter was determined by liquid chromatography with
online organic carbon detection (LC−OCD) (DOC-Labor, Germany)
[27]. The change in fluorescence characteristics of dissolved organic
matter was investigated via three-dimensional fluorescence spectra-
excitation-emission matrix (EEM) using a spectrofluorophotometer (RF-
5301PC, Shimadzu, Japan). The fluorescence EEM was obtained by
repeatedly measuring the emission (Em) spectra within a range of
280–600 nm, with excitation (Ex) wavelengths from 220 to 400 nm
spaced at 10 nm intervals in the excitation domain [28,29]. The PAR-
AFAC was carried out as previously described [30–32].
2.3. Disinfection byproduct formation potential
All samples were adjusted to pH 7 ± 0.2 and were buffered with
phosphate buffer solution. The samples were chlorinated with NaOCl
(1 g/L) at a dose of 5 mg-Cl2 per mg-DOC and were incubated at 25 °C
for 24 h in headspace-free amber bottles. Thereafter, free chlorine that
remained after the incubation was quenched with 10% Na2SO3 prior to
gas chromatography analysis. US EPA Method 551.1 was used to de-
termine the concentrations of trihalomethanes (THMs), in which me-
thyl-t-butyl ether was used as the primary extraction solvent. The vo-
latile compounds in the extracted samples were concentrated using an
Agilent-7697A headspace sampler and were injected into an Agilent-
292
Journal of Hazardous Materials 377 (2019) 290–298
7890A gas chromatograph equipped with an electron capture detector.
2.4. Trace organic contaminants
We selected 11 TrOCs based on their occurrence in municipal
wastewater effluent [33,34] and the analytical availability. Based on
the acidity, 7 ionic and 4 neutral chemicals were included in the se-
lected TrOCs. Similarly, they can be classified into 9 hydrophilic and 2
hydrophobic chemicals based on the polarity [8,35]. Stock solutions of
the 11 TrOCs (gemfibrozil, diclofenac, bezafibrate, ibuprofen, keto-
profen, naproxen, carbamazepine, pentoxifylline, phenacetine, feno-
profen, and caffeine) (Sigma-Aldrich, USA) were prepared in HPLC-
grade methanol (JT Baker, USA) and were stored at −20 °C before use.
A diluted mixture of selected compounds was prepared using their stock
solutions and spiked at 1 μg/L per each compound into the river water
samples. Selected compounds in this study were preconcentrated using
the column switching method utilizing a concentration column (Hy-
persil Gold aQ; Thermo Fisher Scientific) and an analytical column
(Hypersil Gold C18). Mass spectrometry measurements we.re per-
formed on a high-resolution full scan Orbitrap Exactive (Thermo-Fisher,
Bremen, Germany). The mass spectrometry resolution and mass accu-
racy were set at 50,000 and 5 ppm, respectively. Quan Browser (ver
2.1) was used for qualitative and quantitative analysis.
2.5. Assimilable organic carbon and biodegradable dissolved organic carbon
Assimilable organic carbon (AOC) measurements were performed
using flow cytometry (FCM) and indigenous bacteria, as described in
our prior work [11]. The microorganisms isolated from the biological
activated carbon process were cultured and used as an inoculum. The
volume inoculated did not exceed 5 μL per mL of each sample. The
initial cell number of the mixture was approximately 10,000 cells/mL,
and cell growth was allowed in the dark for 72 h at 35 °C. The yields of
indigenous bacteria were converted into AOC concentrations using a
conversion factor of 107 cells/μg-AOC [36]. The yield of indigenous
bacteria was determined by measuring the bacterial cells that were
detected during the incubation. Biodegradable dissolved organic carbon
(BDOC) was also assessed for comparison with AOC and to provide a
better understanding of biodegradability [37,38]. The BDOC measure-
ment was conducted based on a previously reported method [39]. The
experiment was performed using a 1 L amber glass bottle containing
0.45–1.25 mm silica sand that was acclimated with heterotrophs [29].
The samples were filtered with 0.2 μm filters (Whatman, USA), were
placed located in the glass bottles containing biosand media and were
incubated in the dark for 5 days at 20 °C.
2.6. Total cell counts and adenosine triphosphate
The column effluent was collected for microbiological analysis.
Staining, followed by fluorescence-based cytometric analysis, was per-
formed as described previously [40]; the procedures are described in
detail elsewhere [35]. All water samples were immediately stained with
a mixture of SYBR® Green I (Sigma Aldrich, USA) and propidium iodide
(Sigma Aldrich, USA) to differentiate enteric bacteria with intact cel-
lular membranes from disrupted microbial cells and inorganic particles.
The analysis was performed using a Cube 6 flow cytometer (Partec,
Germany) equipped with a 20-mW blue diode pumped solid-state laser
emitting at 488 nm. All collected fluorescence data were processed
using the FCS Express 4 Cytometry software (De Novo Software, USA).
Total cell counts were broadly classified into two groups based on
different fluorescence intensities and sideward scatter signals detected
by FCM (high nucleic acid content (HNA) and low nucleic acid content
(LNA) bacteria) [41]. Total and free ATP contents were determined
using the BacTiter-Glo ™ reagent (Promega, USA) and a GloMax® 20/20
luminometer (Promega, USA). Water samples were filtered using
0.1 μm PES syringe filters (Sartorius, Germany) to determine the free
H.-C. Kim, et al. Journal of Hazardous Materials 377 (2019) 290–298

Table 1
Characteristics of urban surface water after pre-oxidation with O3 and O3/H2O2
(Contact time between ozone and urban surface water > 60 min, Tancheon
River, Seoul, South Korea, n = 44).
Unit Urban surface O3 O3 + H2O2
water

pH 8.3 ± 0.2 7.9 ± 1.5 7.8 ± 0.2

Conductivity μs/cm 396 ± 78 275 ± 78 258 ± 78
Dissolved organic mg/L 3.8 ± 0.8 3.6 ± 0.4 3.3 ± 0.3
carbon
UV254 cm−1 0.08 ± 0.01 0.05 ± 0.01 0.05 ± 0.01
Specific UV L/mg·m 1.7 ± 0.1 1.3 ± 0.1 1.5 ± 0.1
absorbance
Turbidity NTU 20.1 ± 10.2 0.4 ± 0.2 0.2 ± 0.2
Dissolved oxygen mg/L 9.8 ± 0.7 12.5 ± 1.2 12.3 ± 1.8
Oxidation reduction mV 108 ± 24 378 ± 41 390 ± 37
potential
Fig. 2. Impact of O3 and O3/H2O2 on dissolved organic carbon (DOC), biode-
gradable dissolved organic carbon (BDOC), and assimilable organic carbon
ATP content. The cell-bound ATP (intracellular ATP) was calculated by (AOC) (n = 4).
subtracting the free ATP (extracellular ATP) from the total ATP using
the entire sample. surface water increased by 2- and 1.8-fold after using O3 and O3/H2O2,
respectively. Moreover, the AOC of the urban surface water was
3. Results and discussion 276 μg/L and increased to 528 μg/L (1.9-fold increase) and 381 μg/L
(1.4-fold increase) after treatment with O3 and O3/H2O2, respectively.
3.1. Changes in dissolved organic matter characteristics in the feed water The bioavailable dissolved organic carbon fraction determined by the
via O3 alone and O3/H2O2 fraction of AOC in DOC increased from 6.7% in urban surface water to
13.5 and 10% with use of O3 and O3/H2O2, respectively. The increases
Table 1 shows the impact of O3 and O3/H2O2 on urban surface water in AOC for O3 and O3/H2O2 are consistent with the increase in low-MW
characteristics (Tancheon river, Seoul, South Korea). The treatment of dissolved organic matter, which seemed to contribute more for AOC
urban surface water with O3 and O3/H2O2 increased both DO and constituents [37,38] and the reduction in humic substances (hardly
oxidation reduction potential (ORP). A slight change in pH was also biodegradable) as determined by LC−OCD.
found, but it was not significant. The DOC and SUVA of the urban
surface water were 3.8 mg/L and 1.7 L/mg·m, respectively. O3 alone 3.2. Changes in dissolved organic matter characteristics by combining MAR
decreased DOC by 5%, indicating that the O3 dose used in this study did with O3 and O3/H2O2
not have a significant effect on changes in DOC. However, the change in
dissolved organic matter characteristics was observed via SUVA, which MAR combined with O3 alone achieved 64% removal of DOC in
was reduced by 24%. A higher removal of DOC (13%) and the reduction surface water after 30 days of residence time. The reduction in DOC was
in SUVA (12%) were achieved by O3/H2O2. The reduction of SUVA associated with the decreased aromaticity and low-molecularization of
using O3/H2O2 was lower than that of using ozone alone. This was due carbonaceous organic matter upon pre-oxidation of urban surface water
to the difference in oxidation mechanism and potential between O3 and subsequent assimilation of degraded organic matter in micro-
molecules and OH radicals [42]. O3 molecules selectively react with the organisms during MAR (as discussed in Section 3.1). A similar trend
aromatic moieties in organic molecules via direct oxidation pathways, was also observed when MAR was combined with O3/H2O2, which
whereas O3/H2O2 promotes nonselective oxidation processes by pro- showed 56% removal of DOC in urban surface water. The higher re-
ducing free radicals with a strong oxidation potential. The degradation duction of DOC compared to previous studies (23% in [19]; 34% in
of the aromatic structure and the conversion of high MW organic [18]; 40% in [21]) was mainly due to characteristics of the urban
fractions to low MW ones could result in a significant decrease in UV surface water that led the higher formation of biodegradable compo-
absorbance [43,44]. The decrease in UV absorbance was observed for nents in dissolved organic matter via the pre-oxidation treatment.
both O3 and O3/H2O2, and the higher SUVA value after O3/H2O2 was SUVA is strongly correlated with the extent of electron-rich sites
due to the higher removal of DOC compared to using O3 alone. Changes such as aromatic functional groups and double-bonded carbon groups
in dissolved organic matter fractions determined by LC−OCD are in an organic molecule. SUVA values in column effluents were higher
shown in the supporting information, in Fig. S1. O3 alone decreased than for the feed water within the first 5 days, probably due to the
humic substances by 32% and building blocks by 28%, which led to a accumulation of UV-causing organic materials. SUVA tended to de-
significant increase in low-MW organic substances (low-MW neutrals crease slightly when the retention time was increased from 5 to 30 days,
and acids), mainly including alcohols, aldehydes, ketones, and amino indicating a slow reduction in the π-electron system, such as a decrease
acids [45]. Likewise, O3/H2O2 also showed a similar trend in decreasing in the number of aromatic rings or conjugated bonds in a chain struc-
high-MW organic fractions, but the increase in low-MW organic frac- ture, or the conversion of a linear to a nonlinear ring system. A similar
tions, which are believed to be more readily assimilable, was lower than observation was previously reported for aquifer recharge pilot tests
that observed for O3 alone. This may be attributed to a slightly higher conducted on the South Platte River in Colorado, USA [47]. Moreover,
amount of organic matter that was mineralized upon the introduction of Hübner et al. also reported that SUVA decreased during ozonation,
O3/H2O2, producing OH radicals with a strong oxidation potential. OH increased during infiltration, and decreased in final effluents from
radicals are generally considered to react unselectively with organic columns [18]. In summary, it appears that aromatic organic fractions
matter when compared to O3 molecules. Westerhoff et al. corroborated were recalcitrant toward biological degradation at an early stage of the
such difference in the reactivity toward organic matter between O3 and aquifer recharge, but they were slowly decomposed under anoxic and
OH radicals [46]. oligotrophic conditions.
Fig. 2 shows the enhanced biodegradability of carbonaceous organic In this study, we also characterized the fate of fluorescent organic
matter upon the pre-oxidation of urban surface water. BDOC in urban matter during MAR combined with pre-oxidation (Fig. 3). The five

293
H.-C. Kim, et al.
Fig. 3. Monitoring 5 components determined by fluorescence excitation emis-
sion matrix coupled with parallel factor analysis in columns effluents (n = 4).
Fig. 4. Changes in dissolved organic matter fractions determined by liquid
chromatography with online organic carbon detection in column effluents.
fluorescent components were identified using PARAFAC modeling (Fig.
S3), and changes to each of the five components varied during soil
passage. The degradation rate constants of 5 components as a function
294
Journal of Hazardous Materials 377 (2019) 290–298
of the residence times (1, 5, 10, and 30 days) are also shown in Table
S1. Component 5 (C5) was assigned to tryptophan-like substances, and
the other components were associated with humic-like substances (C1,
C2, C3, and C4). The maximum fluorescence intensity (Fmax) of the
individual components was used to represent their relative concentra-
tions. The largest change was observed for C5, which was significantly
decreased within the first 5 days of retention. Among the humic-like
components, only C1 gradually decreased as the residence time in-
creased, and for the others, a negligible change in their intensities was
observed, regardless of the applied pretreatment methods. Fluorescent
decay rate constants for C1, C2, C3, C4, and C5 for O3 alone were 0.07,
0.01, 0.02, 0.02, and 0.1 d−1, respectively and were similar to that of
O3/H2O2 (Table S1). Our result clearly showed that the nondegradable-
or slowly degradable characteristics of humic-like substances (C1, C2,
C3, and C4) are consistent with our observations for the fractional
dissolved organic components determined by LC−OCD (Fig. 4). The
results of the chromatographic analysis easily distinguished degradable
low-MW organic components from nondegradable or slowly degradable
humic substances and building blocks. The building blocks are derived
from the decomposition of humic-like substances, and thus, the physi-
cochemical characteristics are considered to be similar to those of
humus. More than 84% of low-MW acids and neutrals (< 350 Da) was
removed within the first 5 days of retention, regardless of pre-oxidation
methods.
Both pre-oxidation treatments were effective in removing biode-
gradable organic matter within the first 5 days of retention (Table S2).
A consistent tendency in the removal of biodegradable organic fractions
while traveling through the soil matrixes was also observed for the
BDOC and AOC measurements of column effluents, indicating a re-
duction in biological stability. (Fig. 5a and 5b). The dissolved organic
matter removed in the columns was higher with O3 alone than when
combined with O3/H2O2. The pre-oxidation of urban surface water with
O3 reduced BDOC in the column effluent to 0.18 mg/L within the first
15 days. The reduction was further decreased to 0.12 mg/L after 15
more days of traveling through the columns, whereas the pre-oxidation
of O3/H2O2 resulted in 0.27 mg/L of BDOC after a residence time of 30
days. O3/H2O2 achieved a higher removal of DOC through the miner-
alization of organic matter, but produced a smaller amount of bioassi-
milable organic fractions (see Table 1 and Fig. 2). The combination of
ozone alone with MAR in urban surface water achieved a slightly higher
removal of AOC compared to the combination of O3 and H2O2. This
result demonstrates that the pre-oxidation of O3 alone can be used in
combination with MAR, resulting in greater reliability for compliance
with AOC goals with less chemical cost (e.g., chlorine and H2O2), and
therefore lower bacterial regrowth potential in column effluents.
3.3. Bacteriological characteristics in MAR combined with O3 alone and
O3/H2O2
The impacts O3 and O3/H2O2 on the redox conditions in MAR were
estimated by DO and ORP as a function of residence time. Based on the
criteria proposed by [48] and [49], the soil columns examined in this
study remained oxic within the first 5 days of residence time, and they
became more anoxic after 15 days (Fig. 5c and d). The declines in DO
and ORP were observed as residence time increased, which were as-
sociated with the respiration of microorganisms in the columns. No
matter which pre-oxidation was used for urban surface water prior to
MAR, the total cell counts (HNA and LNA bacteria) in the column ef-
fluents gradually decreased as residence time increased (Fig. 6). The
HNA cell population was highly correlated with the cell-bound ATP
(R = 0.93) (Fig. S4), which was consistent with a previous study
(R=0.79, n=260, [41]). Meanwhile, the LNA cells derived from the
HNA cells are considered inactive, dormant, or inviable compared to
the HNA cells [50]. Total ATPs in the column effluents after 30 days of
residence were reduced by 85 and 90% for O3 and O3/H2O2, respec-
tively. As the energy metabolism and biochemical activity of
H.-C. Kim, et al.
Fig. 5. Monitoring biodegradable dissolved organic carbon (BDOC) (a), assimilable organic carbon (AOC) (b), dissolved oxygen (c), and oxidation reduction potential
(d) in columns effluents (n = 4).
microorganisms decreased, the distribution of inactive LNA cells in the
effluent became higher with residence time due to the lack of sub-
strates. The lack of BDOC and AOC resulted in more LNA in columns as
the residence time increased. Additionally, it is important to note that
the residual of H2O2 must be monitored when O3/H2O2 is considered
for pre-oxidation of MAR in real treatment plants. A recent study re-
vealed that H2O2 residuals from advanced oxidation processes could
have a critical impact on the behavior of some aerobic bacteria asso-
ciated with sand systems that showed different responses, either sen-
sitive or tolerant [51]. In our study, residual H2O2 was not detected in
the feed water after O3/H2O2 and showed no effect on the performance
of the columns.
3.4. Trihalomethane formation potential and trace organic contaminants
Fig. 7 shows the impacts of O3 and O3/H2O2 on trihalomethane
formation potential (THMFP) in MAR as a function of residence time.
The pre-oxidation of urban surface water with O3 and O3/H2O2 resulted
in an alleviated formation of CHCl3 (Fig. S5). This was attributed to the
reduction in both humic substances and building blocks among the
fractional components of dissolved organic matter (Fig. S1) and was
consistent with our prior work [11]. MAR combined with O3 decreased
the formation potential of CHCl3 by 75%. This was mainly associated
with the reduction in the reaction sites of humic molecules, which have
been implicated as an important precursor of the haloform reaction. A
similar trend was observed when using O3/H2O2 as a pretreatment, but
295
Journal of Hazardous Materials 377 (2019) 290–298
the formation potential of CHCl3 was higher compared to when using
O3 alone.
More than 71% of selected TrOCs (except carbamazepine) was re-
moved through the artificial recharge of untreated surface water after a
residence time of 1 day (Fig. 8). Specifically, ionic and hydrophilic
species were effectively removed during soil passage, which is con-
sistent with our previous observation [8]. Meanwhile, an insignificant
change in the level of hydrophobic neutral carbamazepine was found
due to its recalcitrance toward biological degradation, corresponding to
previous studies [36,52]. The use of O3 and OH radicals is considered to
be effective in attenuating carbamazepine [53]. Pre-oxidation of urban
surface water with O3 and O3/H2O2 achieved 35 to > 99% removal of
the selected TrOCs. Carbamazepine, naproxen, and gemfibrozil were
almost completely attenuated, whereas only 35% of caffeine was re-
moved upon ozonation. The oxidation rate constant of carbamazepine
with O3 is 3 × 105 M−1 s−1 at ambient temperature, and this rate
constant was comparable to that of naproxen and gemfibrozil with O3.
In contrast, the rate constant of caffeine with O3 is 5.7 × 102 M−1 s−1
[56]. Moreover, it has been reported that caffeine contains two fused
rings (pyrimidinedion and imidazole), and these structures are resistant
to O3 [57]. The rate constant of OH radicals is approximately seven
orders of magnitude higher than that with O3 [56], which was asso-
ciated with the better performance of using O3 with added H2O2 (caf-
feine was decreased by 72% when using O3/H2O2). No matter which
pre-oxidation treatment was applied, subsequent MAR achieved > 99%
removal of selected TrOCs. As thoroughly discussed in the previous
H.-C. Kim, et al. Journal of Hazardous Materials 377 (2019) 290–298

Fig. 6. Changes of high nucleic acid (HNA) and low nucleic acid (LNA) bacteria
in column effluents (n = 4).

Fig. 8. Attenuation of selected TrOCs from urban surface water using O3, O3/
H2O2, and MAR alone (column only).

sections, the pre-oxidation of urban surface water using either O3 or

O3/H2O2 was promising to decay the refractory moieties of TrOCs,
consequently improving the attenuation of the TrOCs.

4. Conclusions

C with chromatography and spectroscopy showed the applicability

of using O3 alone and O3/H2O2 decreased the aromaticity of humic-like
substances and increased the biodegradability of refractory organic
matter in urban surface water. The biodegradation of dissolved organic
matter in columns was enhanced by 64 and 56% by using O3 and O3/
H2O2, respectively. MAR combined with O3 alone and O3/H2O2 showed
good removal of assimilable organic fractions, which was defined by
determining BDOC and AOC in urban surface water samples, indicating
the reduction of biological stability. BDOC and AOC decreased by in-
creasing the residence time due to the consumption of slowly biode-
Fig. 7. Monitoring of trihalomethane formation potential (THMFP) in columns
gradable organic matter in the soil columns. For the residence time over
(n = 4).
the first 15 days, the column environment was changed to anoxic
conditions. LNA cells dominated in the effluent and initiated a slight
decrease in recalcitrant humic-like substances, which were hardly de-
graded in the early stage of aquifer recharge. MAR showed better

296
H.-C. Kim, et al.
performance in reducing the formation of CHCl3 by coupling with O3
compared to when using O3/H2O2 as a pre-oxidation agent, which was
associated with the selective oxidation tendency in the degradation of
organic matter with O3 molecules. The pre-oxidation of urban surface
water with O3 alone and O3/H2O2 was also effective in attenuating
selected TrOCs. The extent of attenuation varied from 35 to > 99%
depending on the use of H2O2. However, subsequent MAR
achieved > 99% removal of selected TrOCs within the first 5 days of
retention, regardless of pre-oxidation methods. The use of O3 alone is
more cost-effective than the use of both O3 and H2O2 due the additional
cost of H2O2. Unless there is a reduction in the TrOCs that are more
effectively removed by O3/H2O2, it is recommended that O3 alone is
used for the pre-oxidation processes of MAR.
Acknowledgments
This research was supported by the Future Research Program
(2E27030) funded by the Korea Institute of Science and Technology.
This research was also partially supported by the National Research
Foundation of Korea grant funded by the Korean government (MSIT)
(No. 2018R1A2A2A05022776).
Appendix A. Supplementary data
Supplementary material related to this article can be found, in the
online version, at doi:https://doi.org/10.1016/j.jhazmat.2019.05.099.
References
[1] B. Petrie, R. Barden, B. Kasprzyk-Hordern, A review on emerging contaminants in
wastewaters and the environment: current knowledge, understudied areas and re-
commendations for future monitoring, Water Res. 72 (2015) 3–27.
[2] A. de Wilt, K. van Gijn, T. Verhoek, A. Vergnes, M. Hoek, H. Rijnaarts,
A. Langenhoff, Enhanced pharmaceutical removal from water in a three step bio-
ozone-bio process, Water Res. 138 (2018) 97–105.
[3] U. Hübner, B. Seiwert, T. Reemtsma, M. Jekel, Ozonation products of carbamaze-
pine and their removal from secondary effluents by soil aquifer treatment–indica-
tions from column experiments, Water Res. 49 (2014) 34–43.
[4] W. Linlin, Z. Xuan, Z. Meng, Removal of dissolved organic matter in municipal
effluent with ozonation, slow sand filtration and nanofiltration as high quality pre-
treatment option for artificial groundwater recharge, Chemosphere 83 (2011)
693–699.
[5] D. Page, E. Bekele, J. Vanderzalm, J. Sidhu, Managed Aquifer Recharge (MAR) in
Sustainable Urban Water Management, Water 10 (2018) 239.
[6] J. Greskowiak, H. Prommer, G. Massmann, C.D. Johnston, G. Nützmann,
A. Pekdeger, The impact of variably saturated conditions on hydrogeochemical
changes during artificial recharge of groundwater, Appl. Geochem. 20 (2005)
1409–1426.
[7] A.F. Henzler, J. Greskowiak, G. Massmann, Modeling the fate of organic micro-
pollutants during river bank filtration (Berlin, Germany), J. Contam. Hydrol. 156
(2014) 78–92.
[8] S.K. Maeng, S.K. Sharma, C.D.T. Abel, A. Magic-Knezev, G.L. Amy, Role of biode-
gradation in the removal of pharmaceutically active compounds with different bulk
organic matter characteristics through managed aquifer recharge: batch and
column studies, Water Res. 45 (2011) 4722–4736.
[9] S.K. Maeng, S.K. Sharma, C.D.T. Abel, A. Magic-Knezev, K.-G. Song, G.L. Amy,
Effects of effluent organic matter characteristics on the removal of bulk organic
matter and selected pharmaceutically active compounds during managed aquifer
recharge: Column study, J. Contam. Hydrol. 140–141 (2012) 139–149.
[10] C. Bertelkamp, J. Reungoat, E.R. Cornelissen, N. Singhal, J. Reynisson, A.J. Cabo,
J.P. van der Hoek, A.R.D. Verliefde, Sorption and biodegradation of organic mi-
cropollutants during river bank filtration: a laboratory column study, Water Res. 52
(2014) 231–241.
[11] H.-C. Kim, W.M. Lee, S. Lee, J. Choi, S.K. Maeng, Characterization of organic pre-
cursors in DBP formation and AOC in urban surface water and their fate during
managed aquifer recharge, Water Res. 123 (2017) 75–85.
[12] K. Hellauer, D. Mergel, A. Ruhl, J. Filter, U. Hübner, M. Jekel, J. Drewes, Advancing
sequential managed aquifer recharge technology (SMART) using different inter-
mediate oxidation processes, Water 9 (2017) 221.
[13] J. Regnery, C.P. Gerba, E.R.V. Dickenson, J.E. Drewes, The importance of key at-
tenuation factors for microbial and chemical contaminants during managed aquifer
recharge: a review, Crit. Rev. Env. Sci. Tec. 47 (2017) 1409–1452.
[14] M. D’Alessio, T. Grischek, C. Ray, Water crisis: bank filtration and aquifer storage
recharge systems as possible alternatives, H. Hazard. Toxic Radioact. Waste 22
(2018) 04018028.
[15] J. Jährig, L. Vredenbregt, D. Wicke, U. Miehe, A. Sperlich, Capillary nanofiltration
under anoxic conditions as post-treatment after bank filtration, Water 10 (2018)
297
Journal of Hazardous Materials 377 (2019) 290–298
1599.
[16] K. Lekkerkerker-Teunissen, E. Chekol, S. Maeng, K. Ghebremichael, C.J. Houtman,
A.R. Verliefde, J. Verberk, G.L. Amy, J.C. Van Dijk, Pharmaceutical removal during
managed aquifer recharge with pretreatment by advanced oxidation, Water Sci.
Tech-W. Supply. 12 (2012) 755–767.
[17] S. Sudhakaran, S.K. Maeng, G. Amy, Hybridization of natural systems with ad-
vanced treatment processes for organic micropollutant removals: new concepts in
multi-barrier treatment, Chemosphere 92 (2013) 731–737.
[18] U. Hübner, U. Miehe, M. Jekel, Optimized removal of dissolved organic carbon and
trace organic contaminants during combined ozonation and artificial groundwater
recharge, Water Res. 46 (2012) 6059–6068.
[19] M.K. Yoon, J.E. Drewes, G.L. Amy, Fate of bulk and trace organics during a simu-
lated aquifer recharge and recovery (ARR)-ozone hybrid process, Chemosphere 93
(2013) 2055–2062.
[20] N. Nakada, H. Shinohara, A. Murata, K. Kiri, S. Managaki, N. Sato, H. Takada,
Removal of selected pharmaceuticals and personal care products (PPCPs) and en-
docrine-disrupting chemicals (EDCs) during sand filtration and ozonation at a
municipal sewage treatment plant, Water Res. 41 (2007) 4373–4382.
[21] U. Hübner, S. Kuhnt, M. Jekel, J. Drewes, Fate of bulk organic carbon and bromate
during indirect water reuse involving ozone and subsequent aquifer recharge, J.
Water Reuse Desal. 6 (2016) 413–420.
[22] Y. Yoon, Y. Hwang, M. Kwon, Y. Jung, T.-M. Hwang, J.-W. Kang, Application of O3
and O3/H2O2 as post-treatment processes for color removal in swine wastewater
from a membrane filtration system, J. Ind. Eng. Chem. 20 (2014) 2801–2805.
[23] I. Kim, H. Tanaka, Use of ozone-based processes for the removal of pharmaceuticals
detected in a wastewater treatment plant, Water Environ. Res. 82 (2010) 294–301.
[24] D.B. Miklos, C. Remy, M. Jekel, K.G. Linden, J.E. Drewes, U. Hübner, Evaluation of
advanced oxidation processes for water and wastewater treatment–A critical re-
view, Water Res. (2018).
[25] G.A. de Vera, J. Keller, W. Gernjak, H. Weinberg, M.J. Farré, Biodegradability of
DBP precursors after drinking water ozonation, Water Res. 106 (2016) 550–561.
[26] M.S. Elovitz, U. von Gunten, Hydroxyl Radical/Ozone ratios during ozonation
processes, I. The Rct Concept, Ozone-Sci. Eng. 21 (1999) 239–260.
[27] S.A. Huber, A. Balz, M. Abert, W. Pronk, Characterisation of aquatic humic and non-
humic matter with size-exclusion chromatography – organic carbon detection –
organic nitrogen detection (LC-OCD-OND), Water Res. 45 (2011) 879–885.
[28] J.B. Fellman, E. Hood, R.G.M. Spencer, Fluorescence spectroscopy opens new
windows into dissolved organic matter dynamics in freshwater ecosystems: a re-
view, Limnol. Oceanogr. 55 (2010) 2452–2462.
[29] J.H. Noh, S.-H. Lee, J.-W. Choi, S.K. Maeng, Dissolved organic matter character-
istics and bacteriological changes during phosphorus removal using ladle furnace
slag, Membr. Water Treat. 9 (3) (2018) 181–188.
[30] B. Aftab, J. Hur, Unraveling complex removal behavior of landfill leachate upon the
treatments of Fenton oxidation and MIEX® via two-dimensional correlation size
exclusion chromatography (2D-CoSEC), J. Hazard. Mater. 362 (2019) 36–44.
[31] R. Bro, PARAFAC. Tutorial and applications, Chemometrics Intellig, Lab. Syst. 38
(1997) 149–171.
[32] R. Bro, H.A. Kiers, A new efficient method for determining the number of compo-
nents in PARAFAC models, J. Chemom. 17 (2003) 274–286.
[33] A. Urtiaga, G. Pérez, R. Ibáñez, I. Ortiz, Removal of pharmaceuticals from a WWTP
secondary effluent by ultrafiltration/reverse osmosis followed by electrochemical
oxidation of the RO concentrate, Desalination 331 (2013) 26–34.
[34] M. Eddy, F. Burton, G. Tchobanoglous, R. Tsuchihashi, Wastewater Engineering:
Treatment and Resource Recovery, McGraw-Hill Education, New York, NY, USA,
2013.
[35] H.-C. Kim, J.H. Noh, S.-R. Chae, J. Choi, Y. Lee, S.K. Maeng, A multi-parametric
approach assessing microbial viability and organic matter characteristics during
managed aquifer recharge, Sci. Total Environ. 524–525 (2015) 290–299.
[36] F. Hammes, E. Salhi, O. Köster, H.-P. Kaiser, T. Egli, U. von Gunten, Mechanistic
and kinetic evaluation of organic disinfection by-product and assimilable organic
carbon (AOC) formation during the ozonation of drinking water, Water Res. 40
(2006) 2275–2286.
[37] M.M. Bazri, B. Barbeau, M. Mohseni, Impact of UV/H2O2 advanced oxidation
treatment on molecular weight distribution of NOM and biostability of water, Water
Res. 46 (2012) 5297–5304.
[38] Y. Choi, H. Park, M. Lee, G.-S. Lee, Y.-j. Choi, Seasonal variation of assimilable
organic carbon and its impact to the biostability of drinking water, Int. J. Civ.
Struct. Environ. Infrastruct. Eng. Res. Dev. 24 (2019) 501–512.
[39] P. Servais, G. Billen, M.-C. Hascoët, Determination of the biodegradable fraction of
dissolved organic matter in waters, Water Res. 21 (1987) 445–450.
[40] M. Berney, M. Vital, I. Hülshoff, H.-U. Weilenmann, T. Egli, F. Hammes, Rapid,
cultivation-independent assessment of microbial viability in drinking water, Water
Res. 42 (2008) 4010–4018.
[41] G. Liu, E. Van der Mark, J. Verberk, J. Van Dijk, Flow cytometry total cell counts: a
field study assessing microbiological water quality and growth in unchlorinated
drinking water distribution systems, Biomed Res. Int. 2013 (2013).
[42] C. Volk, P. Roche, C. Renner, H. Paillard, J. Joret, Effects of ozone-hydrogen per-
oxide combination on the formation of biodegradable dissolved organic carbon,
Ozone-Sci. Eng. 15 (1993) 405–418.
[43] M. Yan, D. Wang, B. Shi, M. Wang, Y. Yan, Effect of pre-ozonation on optimized
coagulation of a typical north-china source water, Chemosphere 69 (2007)
1695–1702.
[44] R. Treguer, R. Tatin, A. Couvert, D. Wolbert, A. Tazi-Pain, Ozonation effect on
natural organic matter adsorption and biodegradation – application to a membrane
bioreactor containing activated carbon for drinking water production, Water Res.
44 (2010) 781–788.
H.-C. Kim, et al.
[45] S.A. Huber, Significance, origin, and fate of natural organic matter (NOM) in boiler
feedwater preparation using surface water as raw water source, VGB
Kraftwerkstechnik 80 (2000) 96–101.
[46] P. Westerhoff, G. Aiken, G. Amy, J. Debroux, Relationships between the structure of
natural organic matter and its reactivity towards molecular ozone and hydroxyl
radicals, Water Res. 33 (1999) 2265–2276.
[47] C. Hoppe-Jones, G. Oldham, J.E. Drewes, Attenuation of total organic carbon and
unregulated trace organic chemicals in U.S. Riverbank filtration systems, Water Res.
44 (2010) 4643–4659.
[48] R.J. Diaz, R. Rosenberg, Marine benthic hypoxia: a review of its ecological effects
and the behavioural response of benthic macrofauna, Taylor & Francis (1995).
[49] T.H.Y. Tebbutt, 2 - characteristics of waters and wastewaters, in: T.H.Y. Tebbutt
(Ed.), Principles of Water Quality Control, fifth edition), Butterworth-Heinemann,
298
Journal of Hazardous Materials 377 (2019) 290–298
Oxford, 1998, pp. 12–27.
[50] T. Bouvier, P.A. Del Giorgio, J.M. Gasol, A comparative study of the cytometric
characteristics of High and Low nucleic-acid bacterioplankton cells from different
aquatic ecosystems, Environ. Microbiol. 9 (2007) 2050–2066.
[51] F. Wang, D. van Halem, G. Liu, K. Lekkerkerker-Teunissen, J.P. van der Hoek, Effect
of residual H2O2 from advanced oxidation processes on subsequent biological
water treatment: a laboratory batch study, Chemosphere 185 (2017) 637–646.
[52] R. Broséus, S. Vincent, K. Aboulfadl, A. Daneshvar, S. Sauvé, B. Barbeau, M. Prévost,
Ozone oxidation of pharmaceuticals, endocrine disruptors and pesticides during
drinking water treatment, Water Res. 43 (2009) 4707–4717.
[53] J. Meadows, R.C. Smith, Uric acid protection of nucleobases from ozone-induced
degradation, Arch. Biochem. Biophys. 246 (1986) 838–845.