Environ Sci Pollut Res (2015) 22:15812–15820

DOI 10.1007/s11356-015-4783-1

RESEARCH ARTICLE

Enhanced hydroxyl radical generation in the combined

ozonation and electrolysis process using carbon nanotubes
containing gas diffusion cathode

Donghai Wu 1 & Guanghua Lu 1 & Ran Zhang 2 &

Qiuhong Lin 1 & Zhenhua Yan 1 & Jianchao Liu 1 & Yi Li 1

Received: 13 March 2015 / Accepted: 25 May 2015 / Published online: 4 June 2015
# Springer-Verlag Berlin Heidelberg 2015

Abstract Combination of ozone together with electrolysis revealed that peroxone reaction was the main pathway
(ozone-electrolysis) is a promising wastewater treatment tech- of ·OH production in the present system, and cathodic
nology. This work investigated the potential use of carbon reduction of ozone could significantly promote ·OH
nanotube (CNT)-based gas diffusion cathode (GDC) for generation. These results suggested that application of
ozone-electrolysis process employing hydroxyl radicals (· CNT-based GDC offers considerable advantages in
OH) production as an indicator. Compared with conventional ozone-electrolysis of organic wastewater.
active carbon (AC)-polytetrafluoroethylene (PTFE) and car-
bon black (CB)-PTFE cathodes, the production of ·OH in the
coupled process was improved using CNTs-PTFE GDC. Keywords Ozone-electrolysis . Gas diffusion cathode .
Appropriate addition of acetylene black (AB) and pore- Carbon nanotubes . Hydroxyl radicals . Response surface
forming agent Na2SO4 could enhance the efficiency of methodology
CNTs-PTFE GDC. The optimum GDC composition was
obtained by response surface methodology (RSM) anal-
ysis and was determined as CNTs 31.2 wt%, PTFE
60.6 wt%, AB 3.5 wt%, and Na 2 SO 4 4.7 wt%. Nomenclature
Moreover, the optimized CNT-based GDC exhibited AB Acetylene black
much more effective than traditional Ti and graphite AC Active carbon
cathodes in Acid Orange 7 (AO7) mineralization and AOPs Advanced oxidation processes
possessed the desirable stability without performance AO7 Acid Orange 7
decay after ten times reaction. The comparison tests ANOVA Analysis of variances
BBD Box-Behnken design
Responsible editor: Bingcai Pan
CB Carbon black
Electronic supplementary material The online version of this article CNTs Carbon nanotubes
(doi:10.1007/s11356-015-4783-1) contains supplementary material, C.V. Coefficient of variation
which is available to authorized users.
DDI Deionized-distilled
* Guanghua Lu
GDC Gas diffusion cathode
ghlu@hhu.edu.cn O3 Ozone
·OH Hydroxyl radicals
1
pCBA p-Chlorobenzoic acid
Key Laboratory of Integrated Regulation and Resource Development
on Shallow Lakes of Ministry of Education, College of Environment,
PTFE Polytetrafluoroethylene
2
Hohai University, Nanjing 210098, China R Coefficient of determination
2
State Key Laboratory of Urban Water Resource and Environment,
RSM Response surface methodology
Harbin Institute of Technology, P.O. Box 2606, 202 Haihe Road, SEM Scanning electron microscopy
Harbin 150090, China TOC Total organic carbon
Environ Sci Pollut Res (2015) 22:15812–15820
Introduction
Water pollution by organic compounds is a worldwide prob-
lem at present. Advanced oxidation processes (AOPs) involve
the production of strongly oxidizing hydroxyl radicals (·OH)
and have been recommended as efficient technologies for wa-
ter purification, with many of them based on ozone (O3) and
electrolysis application (Gedam et al. 2014; Hermosilla et al.
2015). Ozone is a superior oxidant that can directly oxidize
organic compounds or can produce ·OH which are further
used to react with the target contaminants. However, some
pollutants are resistant to the ozonation alone, and addition
of high dose of ozone can produce toxic by-products (Chang
et al. 2014). Electrochemical oxidation is an economic tech-
nology and easy in operation (Oturan et al. 2013;
Vijayalakshmi et al. 2011). Nevertheless, long time required
limits its practical applications for water treatment (Garcia-
Morales et al. 2013).
Recently, combination of ozone and electrolysis (ozone-
electrolysis) is proposed to be a promising water treatment
technology with high efficiency and environmentally friendly
(Li et al. 2015). Evidence has shown that synergistic pollutant
degradation could be obtained using this coupled process (Qiu
et al. 2014), and gas diffusion cathode (GDC) was recom-
mended for improving the efficiency of organic pollutant min-
eralization (Yuan et al. 2013). The application of GDC in
ozone-electrolysis process brings about enhanced generation
of H2O2 from O2 reduction, owing to its three-phase structure.
The electrogenerated H2O2 can react with O3 (i.e., the so-
called peroxone oxidation) to produce ·OH (Eqs. (1) and (2))
(Bakheet et al. 2013), thereby accelerating pollutant degrada-
tion in the coupled process.
O2 þ 2Hþ þ 2e− →H2 O2 ð1Þ
H2 O2 þ O3 →•OH þ •O2 − þ Hþ þ O2 ð2Þ
The performance of GDC is mainly dependent on the struc-
ture and properties of the carbon support (Barros et al. 2013).
Binary carbon supports GDC can achieve a combination of
good morphology, high conductivity, and suitable hydropho-
bicity and exhibits better performance than single carbon sup-
ports GDC (Huang et al. 2005). Meanwhile, it was found that
addition of pore-forming agent could improve the effective-
ness of fabricated GDC (Xu et al. 2010). However, only car-
bon black (CB)-polytetrafluoroethylene (PTFE)-coated GDC
has been investigated hitherto for ozone-electrolysis of waste-
water, and few studies had examined the influence of gas
diffusion cathode composition on the effectiveness of this
coupled process. Carbon nanotubes (CNTs) have extraordi-
nary mechanical, chemical, and electrical properties.
Heterogeneous catalytic ozonation using CNTs can overcome
the drawbacks of ozonation alone, such as selective reactiva-
tion and by-product formation (Fan et al. 2014). As carbon
15813
materials, CNTs also exhibit excellent capability for
electrogeneration of H2O2 (Khataee et al. 2011; Vahid and
Khataee 2013). Thus, CNTs may be suitable cathode materials
for the ozone-electrolysis process.
Compared to large database on kinetics, the advanced ox-
idation mechanism of ozone-electrolysis process is relatively
not well established. Although production of excess ·OH has
been demonstrated as the main responsible for refractory or-
ganic compound degradation in ozone-electrolysis process,
the pathways of ·OH generation in literatures are ambiguous.
While some authors stated that cathodic reduction of O3 into ·
O3− was the main pathway of ·OH generation in ozone-
electrolysis system (Eqs. (3) and (4)) (Kishimoto et al.
2008), others suggested that peroxone reaction played a criti-
cal role in ·OH formation (Bakheet et al. 2013; Li et al. 2013).
O3 þ e− →•O3 − at cathodes ð3Þ
− −
•O3 þ H2 O→•OH þ O2 þ OH ð4Þ
The main objectives of this study were (1) to design and
optimize CNT-based GDC for ozone-electrolysis process
using response surface methodology (RSM); (2) to evaluate
the effectiveness of fabricated CNT-based GDC in ozone-
electrolysis of organic wastewater and elucidate the ·OH pro-
duction mechanism. Acid Orange 7 (AO7), a model refractory
organic contaminant in wastewater, was employed as the tar-
get pollutant. The effect of cathode composition on ·OH gen-
eration was examined, and the pathways of ·OH generation
was investigated.
Experimental
Chemicals and reagents
Multi-walled CNTs, with inner and outer diameters of 5–12
and 30–50 nm, respectively, were purchased from Chengdu
Organic Chemistry Co., Ltd, Chinese Academy of Sciences;
CB (Vulcan XC-72) was purchased from Cabot Corp., USA;
Active carbon powder was from Jiabao Carbon Co., China;
Acetylene black (AB) was from Timcal Co., Switzerland;
p-chlorobenzoic acid (pCBA) and indigo were from Sigma-
Aldrich. Other chemicals (e.g., AO7, PTFE, Na 2 SO 4 ,
Na2S2O3, and KI) were analytical grade and purchased from
Aladdin Industrial Co., China. All solutions were prepared
using deionized-distilled (DDI) water.
GDC preparation and optimization
The binary carbon support GDCs were prepared by the press-
ing method consisting of PTFE, CNTs, AB, pore-forming
agent Na 2 SO 4 , and current collector nickel mesh.
Fabrication of the GDCs followed published procedure
15814
(Wang et al. 2012), the obtained GDCs after pressing were
then boiled in DDI water for 1 h to remove Na2SO4. For
comparison, the CNTs-PTFE, CB-PTFE, and AC-PTFE
GDCs with carbon supports loading of 80 mg/cm2 and
PTFE loading of 160 mg/cm2 were also prepared.
RSM is a statistical method being useful for the experimen-
tal design and processes optimization (García-Gómez et al.
2014; Khataee et al. 2014). Box-Behnken design (BBD),
which is a widely used form of RSM, was employed for the
optimization of binary carbon support GDC. The CNT load-
ing was 80 mg/cm2, and the weight ratios of PTFE/CNTs, AB/
CNTs, and Na2SO4/CNTs (i.e., WPTFE:WCNTs, WAB:WCNTs,
and WNa2SO4:WCNTs) were selected as independent variables.
The amount of ·OH produced in the ozone-electrolysis system
was taken as the dependent variable. The BBD was applied
using Design-Expert 8.0.6 software. The levels and ranges of
the variables are summarized in Table 1. A quadratic model
was chosen in terms of independent variables, which is given
as follows:
3 3 2 3
Y ¼ β0 þ ∑ β i xi þ ∑ βii x2i þ ∑ ∑ βi j xi x j þ εði≠ jÞ
i¼1 i¼1 i¼1 j¼2
where Y is the response value; β0 is the intercept; βi, βii, βij are
the regression coefficients for linearity, square, and interac-
tion, respectively; Xi and Xj are the different independent var-
iables; ε is the random error.
To obtain the interaction between the variables and the
responses, the analysis of variances (ANOVA) was conducted.
The optimum parameters for GDC preparation were calculat-
ed using the fitted models and validated by the experiments.
Experimental setup and conditions
Experiments were carried out in an undivided poly reactor
(50 mm wide, 100 mm length, and 150 mm height) with
working volume of 500 mL (Fig. S1 in Supplementary
Material). A DC power supply (KPS-3030DA, Atten
Technology Co., China) was employed to provide constant
current between a 0.8 cm2 Pt anode and a 7.5 cm2 GDC.
The electrode gap was 30 mm, and the applied current was
300 mA, corresponding with current density of 40 mA/cm2 at
Table 1 Coded levels and independent variables for the experimental
design
Independent variables Codes Levels
−1 0 1
WAB:WCNTs (%) X1 0 7.5
WNa2SO4:WCNTs (%) X2 0 7.5
WPTFE:WCNTs (%) X3 50 150
Environ Sci Pollut Res (2015) 22:15812–15820
cathode. For ozonation and ozone-electrolysis tests, 0.3 L/min
ozonized gas generated from dry oxygen by a laboratory
ozone generator (WH-H-Y5, Wohuan Technology Co.,
China) was fed into the reactor through a bubble diffuser sit-
uated under the cathode, while equal oxygen was used in
experiments with electrolysis alone. When considering
peroxone oxidation only, ozonized gas was sparged into the
reactor through another bubble diffuser. The distance between
two bubble diffusers was 30 mm, and the ozone concentration
in ozonized gas was set at 9.8 mg/L.
Determination of ·OH and degradation of AO7 (100 mg/L)
were performed in the aqueous medium containing 0.05 M
Na2SO4 at ambient temperatures (25±2 °C). The ozone tests
with no electrode are termed Bozonation tests.^ The ozone
tests with CNT-based electrode are termed BCNTs catalytic
ozonation tests.^ The electrolysis tests with O2 flow through
GDC are termed Belectrolysis tests.^ The electrolysis tests
with ozonized gas flow through GDC are named Bozone-elec-
trolysis tests.^ The electrolysis test with both O2 (flow through
GDC) and ozonized gas (not flow through GDC) is the
ð5Þ Belectrolysis+peroxone oxidation test.^
Analytical methods
O3 concentration in gas was measured by the iodometric
method, and O3 residual in water was determined by the indi-
go method (APHA 1998). ·OH produced in the solution was
measured by the pCBA method (Cho et al. 2004), concentra-
tion of pCBA was analyzed with a HPLC analytical system
(Waters Co., 1525 Binary system, USA). The steady-state
concentrations of ·OH could be approximated from the known
second-order rate constant for the reaction of pCBA with hy-
droxyl radicals, i.e., kOH,pCBA (=5×109 M−1 s−1) (Neta and
Dorfman 1968) and the pseudo-first-order rate constant
of pCBA measured in various processes. The scanning
electron microscopy (SEM) images of GDCs were per-
formed on a Hitachi S-3400N II microscope (Japan).
The absorbance at 485 nm was analyzed by a UV-
Visible spectrophotometer (T6, Beijing Purkinje
General Instrument Co., Ltd., China) to evaluate the
decolorization efficiency. Total organic carbon (TOC)
was measured using a TOC analyzer (multiN/C 2100,
Analytik Jena AG, Germany) to evaluate the minerali-
zation of AO7 during the treatment.
Results and discussion
Comparison of AC-PTFE, CB-PTFE, and CNTs-PTFE
15 GDCs in ·OH generation
15
250 Three cathode materials were investigated for generation of ·
OH, and the results are presented in Fig. 1. The steady-state
Environ Sci Pollut Res (2015) 22:15812–15820 15815

0.0 Table 2 BBD matrix and response values for the yield of ·OH
-0.2 Exp. WAB:WCNTs WNa2SO4:WCNTs WPTFE:WCNTs Yield of ·OH
no. (%) (%) (%) (×10−13 mol/L)
-0.4
1 7.5 7.5 150 5.8
ln ([pCBA]t/[pCBA]0)

-0.6
2 15 15 150 6.0
-0.8 3 7.5 7.5 150 5.6
4 7.5 7.5 150 5.8
-1.0 AC-PTFE GDC
5 7.5 7.5 150 5.7
CB-PTFE GDC
-1.2 CNT-PTFE GDC 6 7.5 7.5 150 5.9
7 7.5 15 250 5.8
-1.4 8 15 7.5 250 5.5
0 2 4 6 8 10 9 15 7.5 50 3.8
Time (min) 10 0 7.5 250 5.2
11 0 7.5 50 2.7
Fig. 1 Decay of pCBA during ozone-electrolysis process using different
cathodes (current of 300 mA; ozonized gas flow rate of 0.3 L/min; inlet 12 0 15 150 4.6
O3 concentration of 9.8 mg/L) 13 15 0 150 5.2
14 7.5 0 250 5.0
15 7.5 15 50 4.2
concentrations of ·OH that calculated from the slope of the 16 7.5 0 50 3.5
curves in Fig. 1 were 4.7 × 10−13, 2.2 × 10−13, and 1.9 × 17 0 0 150 4.4
10−13 mol/L for CNTs-PTFE, CB-PTFE, and AC-PTFE
GDCs, respectively, suggesting that CNTs-PTFE is the most
response variable and the independent variables was derived
effective cathode for ·OH production in ozone-electrolysis
as follows:
process.
It is well known that ozone could react easily on the Y ¼ 5:76 þ 0:45X 1 þ 0:31X 2 þ 0:91X 3
surface of AC and CNTs due to their specific porous
structure. However, compared with CB and CNTs, AC þ 0:15X 1 X 2 − 0:20X 1 X 3
was an inefficient GDC material for O2 reduction to
þ 0:03X 2 X 3 − 0:52X 1 2 − 0:19X 2 2 − 0:94X 3 2 ð6Þ
generate H2O2 (Soltani et al. 2012), which can react
with sparged O3 to produce ·OH. Thus, the above phe- where Y was the yield of ·OH, and X1, X2, X3 were the
nomena can be explained by the fact that CNTs have a coded values of the independent variables (Table 1).
great number of mesoporous pores and large accessible The sufficiency of the model was evaluated through
surface area (Khataee et al. 2011), so that both O2 and ANOVA (Table S1 in Supplementary Material). The model
O3 can be reduced easily on the cathode surface to F-value was 36.43, and the model p value was less than
generate more ·OH. Since the oxidation power of 0.0500, implying that the model was statistically significant.
ozone-electrolysis system is mainly related to the pro- The determination coefficients (R2) of the models indicated
duced intermediate reactive oxygen species, the cathode that 97.91 % of the total variability could be explained by
materials which have high ·OH production ability are the model. The adjusted determination coefficient (adjusted
essential for the effective degradation of pollutants. R2 =0.9522) further testified the minute difference between
Therefore, CNTs are suitable cathode materials for the experimental and predicted values. The low coefficient
ozone-electrolysis process. variation value (C.V.=4.25 %) suggested the precision and
reliability of the experimental values.
CNT-based GDC optimization
Optimization of the variables
Model building and statistical analyses
A 3D representation of the polynomial (Eq. 6) obtained from
The full factorial BBD with three factors in three levels as well the experimental data was shown in Fig. 2. As seen in Fig. 2a,
as the results of ·OH generation performance of prepared the AB content in GDC exhibited a significant effect on the ·
GDC is presented in Table 2. The data of BBD experiment OH production in ozone-electrolysis system, whereas the per-
were analyzed by Design-Expert 8.0.6, and a quadratic poly- centage of pore-forming agent Na2SO4 exerted a weaker ef-
nomial equation reflecting the empirical relationship between fect. The generated ·OH concentration initially increased with
15816
Fig. 2 Response surface showing the effects of GDC composition on the
yield of ·OH in ozone-electrolysis process (current of 300 mA; ozonized
gas flow rate of 0.3 L/min; inlet O3 concentration of 9.8 mg/L)
increasing weight ratio of AB/CNTs; thereafter, it started
to decrease. When the weight ratio of Na2SO4/CNTs
increased, the yield of ·OH smoothly increased.
Appropriate addition of AB or pore-forming agent could
enhance the distribution of the gases in GDC (Maja
et al. 2000; Xu et al. 2010). However, a high level of
AB content would decrease the gas-liquid mass transfer
rate, and the presence of pore-forming agent would
block the transportation of electrons in electrode.
Environ Sci Pollut Res (2015) 22:15812–15820
Figure 2b shows the effects of PTFE and AB contents on ·
OH generation. An increase of PTFE/CNT weight ratio in-
creased the yield of ·OH, irrespectively of the AB content.
The roles of PTFE in GDC are binding the high surface carbon
particles into a cohesive layer and imparting some hydropho-
bic character to the layer (Giorgi et al. 1998). When high
content of PTFE are used, the yield of ·OH diminishes since
it may impact the hydrophilic porosity of fabricated GDC and
reduce the electrode conductivity.
Similar to Fig. 2a, b, the PTFE loading has an important
effect on the ·OH production while the content of pore-
forming agent Na2SO4 was insignificant to the ·OH genera-
tion, which is shown in Fig. 2c. Furthermore, SEM measure-
ments of prepared GDCs showed that the weight ratios of
PTFE/CNTs, AB/CNTs, and Na2SO4/CNTs have significant
influence on electrode morphology (Fig. 3). For GDC with
low PTFE content, pores were not shown distinctly on the
electrode surface (Fig. 3a). When the PTFE content increased,
porous morphology emerged (Fig. 3b), which could facilitate
O2 and O3 diffusion. On the other hand, the absence of AB or
pore-forming agent Na2SO4 induced the decrease in amount
of small pores, owing to the agglomeration of GDC active
layer (Fig. 3c, d). Thus, appropriate increasing of AB, pore-
forming agent Na2SO4, and PTFE contents could result in a
good porosity suitable for gas diffusion through the pores,
which was in agreement with the results of BBD experiment.
The optimized conditions of three independent variables
(i.e., WPTFE:WCNTs, WAB:WCNTs, and WNa2SO4:WCNTs) were
obtained by a response optimizer. The desired goal in terms
of hydroxyl radicals (·OH) production was defined as
Bmaximize^ to achieve highest ·OH generation performance,
while the independent variables were selected to be Bin the
range.^ Accordingly, it could be deduced that the optimal
GDC compositions are as follows: CNTs 31.2 wt%, PTFE
60.6 wt%, AB 3.5 wt%, and Na2SO4 4.7 wt%. With these
optimal GDC compositions, a maximum response of 6.2×
10−13 mol/L ·OH production was obtained by the mathemat-
ical prediction of the model.
Performance of optimized CNT-based GDC
in ozone-electrolysis treatment of AO7
To investigate the effectiveness of optimized CNT-based
GDC, the degradation of AO7 by the combined ozone-
electrolysis process was compared with that in each individual
unit process as illustrated in Fig. 4. Although CNTs have been
reported as superior adsorbents (Zhang and Xu 2014), only
about 3 % of AO7 was found to be adsorbed on the surface of
electrodes (Fig. 4a). This may be due to the fact that CNTs
used in this study were not suspended particulates, and the size
of cathode was small. The decolorization ratio for electrolysis
test was about 20 % at 10-min treatment. Contrarily, ozonation
removed approximately 95 % color after 10 min. No
Environ Sci Pollut Res (2015) 22:15812–15820
Fig. 3 SEM images of fabricated CNT-based GDCs (WAB:WCNTs =
7.5 %, WNa2SO4:WCNTs =15 %, WPTFE:WCNTs =50 % (a); WAB:WCNTs =
7.5 %, WNa2SO4:WCNTs =15 %, WPTFE:WCNTs =250 % (b); WAB:WCNTs =
significant improvement of color removal was observed in
ozone-electrolysis process, consisting with previous reports
(Bakheet et al. 2013; Hsu et al. 2003). This may be
due to the fact that electrolyte used in current work
was Na2SO4. Beltrán et al. suggested that null effect
of H2O2 was expected while the direct reactions be-
tween reactants and ozone are faster (Beltrán et al.
1996). However, ozone-electrolysis could present more
efficiency than ozonation for azo dye decolorization in
NaCl solution, owing to the production of hypochlorite
(Parsa et al. 2014).
During decolorization tests revealed in Fig. 4a, the miner-
alization of AO7 was also investigated (Fig. 4b). Electrolysis
or ozonation alone was inefficient for AO7 mineralization; the
detected TOC removal ratios were about 6.4 and 11.3 % at 10-
min treatments, respectively. Conversely, combined tests yield
synergistic mineralization of AO7, approximately 41.5 %
TOC removal was obtained. Even when taking the electrolysis
and O3 effects into consideration, enhancement in the ozone-
electrolysis process was notable. The mineralization of AO7
by ozone-electrolysis process using traditional Ti and graphite
cathodes was also presented in the Fig. 4b as references, and
only 16.9 and 20.1 % TOC removal were obtained within
10 min, respectively.
15817
0 %, WNa2SO4:WCNTs =7.5 %, WPTFE:WCNTs =250 % (c); WAB:WCNTs =
7.5 %, WNa2SO4:WCNTs =0 %, WPTFE:WCNTs =250 % (d))
Ten successive runs were performed to evaluate the
stability of CNT-based GDC in ozone-electrolysis sys-
tem. As shown in Fig. 5, no obvious change was ob-
served on the performance of CNT-based GDC after
the tenth run, and the mineralization efficiency of
AO7 in ozone-electrolysis remained at 39.8 % within
10 min, confirming that the stability of CNT-based
GDC can be maintained for a long time and it can
be reused.
Based on the experimental results, CNT-based GDC
is effective for ozone-assisted electrochemical degrada-
tion of organic compounds in wastewater. A possible
reason, as discussed in BComparison of AC-PTFE, CB-
PTFE, and CNTs-PTFE GDCs in ·OH generation,^ is
that the presence of CNTs in cathode contributed the
improvement of ·OH generation in ozone-electrolysis
system. This explanation is consistent with the results
in previous research, which showed that ·OH was the
main responsible for refractory organic compound deg-
radation in ozone-electrolysis process (Kishimoto et al.
2005). While ozone oxidizes organic compounds selec-
tively and partly, ·OH is very unselective and reacts at
rates that are several orders of magnitude higher than
O3 (Brillas et al. 2009). Consequently, generation of
15818 Environ Sci Pollut Res (2015) 22:15812–15820

0.0
(a) 100
ozone-electrolysis (CNTs-based GDC)
electrolysis (CNTs-based GDC)
80 CNTs-based GDC adsorption
ozonation -0.5

ln ([pCBA]t/[pCBA]0)
Color removal (%)

60

-1.0
40 electrolysis
ozonation
-1.5 CNTs catalytic ozonation
20 electrolysis+peroxone oxidation
ozone-electrolysis

0 -2.0
0 2 4 6 8 10 0 2 4 6 8 10
Time (min)
Time (min)
(b) 50 Fig. 6 Time-dependent decay of pCBA using different treatment
ozone-electrolysis (CNTs-based GDC) processes (current of 300 mA; O2 flow rate of 0.3 L/min; ozonized gas
ozone-electrolysis (graphite cathode)
flow rate of 0.3 L/min; inlet O3 concentration of 9.8 mg/L)
40 ozone-electrolysis (Ti cathode)
electrolysis (CNTs-based GDC)
CNTs-based GDC adsorption Mechanism of ·OH production
ozonation
30
TOC removal (%)

In Fig. 6, the generation of ·OH in various oxidation processes

20
10
0 Although some dissolved O3 would diffuse to the cathode, the
0 2
Fig. 4 Comparison of color (a) and TOC (b) removal using various
oxidation processes (current of 300 mA; ozonized gas flow rate of
0.3 L/min; inlet O3 concentration of 9.8 mg/L)
sufficient ·OH in ozone-electrolysis system destroyed
organics up to their mineralization.
is described. The calculated ·OH in electrolysis or ozonation
tests were about 0.3×10−13 mol/L, and CNT catalytic ozona-
tion only enhanced ·OH production slightly. In contrast, the
yield of ·OH in ozone-electrolysis process was 6.0 ×
10−13 mol/L, confirming that the model equation was suitable.
4 6 8 10
production of ·OH in the electrolysis+peroxone oxidation test
Time (min)
was only 3.8×10−13 mol/L.
The concentrations of dissolved O3 during various O3-
based treatments were also determined, and the results are
shown in Fig. 7. The dissolved O3 concentration increased
continuously during the ozonation treatment, suggesting that
O3 was transferred from the gas phase to the liquid phase more

3.0
40 ozonation
2.5 CNTs catalytic ozonation
Adueous ozone concentration (mg/L)

electrolysis+peroxone oxidation
30 ozone-electrolysis
2.0
TOC removal (%)

20
1.5

first run
fifth run 1.0
10 tenth run

0.5

0 0.0
0 2 4 6 8 10 0 2 4 6 8 10
Time (min) Time (min)
Fig. 5 Performance of CNT-based GDC in ten successive ozone- Fig. 7 Profiles of aqueous O3 concentrations in various ozone-based
electrolysis runs (current of 300 mA; ozonized gas flow rate of 0.3 L/ processes (current of 300 mA; O2 flow rate of 0.3 L/min; ozonized gas
min; inlet O3 concentration of 9.8 mg/L) flow rate of 0.3 L/min; inlet O3 concentration of 9.8 mg/L)
Environ Sci Pollut Res (2015) 22:15812–15820
e- e-
DC power supply
+ -
O2
H2O
+2e-
Acid Orange 7
Pt -e- H2O2
anode peroxone
•OHads •OHbulk reaction
CO2
O3
+ +e-
H2O
•O3-
: main pathway
: minor pathway
Fig. 8 Formation routes of ·OH in the ozone-electrolysis process
rapidly than it was consumed in the liquid phase. Similar to
the results in Fig. 6, CNT catalytic ozonation only improved
the decomposition of aqueous O3 slightly. In comparison, the
aqueous O3 concentrations were kept at much lower levels
within 10 min of electrolysis+peroxone oxidation process
(<1.5 mg/L) and ozone-electrolysis process (<1.0 mg/L), in-
dicating that the presence of electrolysis accelerated the dis-
solved O3 consumption. These phenomena agree with the
results in previous report, which implied that the consumption
of dissolved O3 by electrochemically driven reactions would
improve the utilization efficiency of O3 and enhance the pro-
duction of ·OH (Bakheet et al. 2014).
The above observations suggest that more than half of the ·
OH in ozone-electrolysis process was produced via peroxone
reaction. Additionally, the amount of ·OH produced in ozone-
electrolysis process was higher than the sum of that in elec-
trolysis+peroxon oxidation and CNT catalytic ozonation pro-
cesses, indicating that cathodic reduction of ozone may be
another important pathway for ·OH formation (Fig. 8).
It should be noted that operational parameters have signif-
icant influence on ·OH production in ozone-electrolysis pro-
cess (Li et al. 2014). The results present here should be
interpreted within its context and aim of this work, and further
investigation is required to better understand the mechanism
of ·OH generation at various operation conditions.
Conclusions
This paper revealed that CNTs are novel and efficient cathode
materials for ozone-electrolysis process. The electrode com-
position has a significant effect on the performance of fabri-
cated GDC. The optimized CNT-based GDC was obtained
using BBD based on RSM and achieved excellent perfor-
mance in ·OH production and AO7 mineralization in the
ozone-electrolysis system. Simultaneous peroxone oxidation
15819
and cathodic reduction of ozone resulted in excess ·OH gen-
eration in the present system, thereby enhancing the treatment
efficiency. It is concluded that CNT-based gas diffusion elec-
trode is promising to be potentially used as a new cathode for
ozone-electrolysis process.
CNT/PTFE
cathode Acknowledgments We are grateful for grants from Natural Science
Foundation of Jiangsu Province (No. BK20130835), China Postdoctoral
Science Foundation (No. 2013 M541600), Fundamental Research Funds
for the Central Universities of Hohai University (No. 2013B13020026),
and A Project Funded by the Priority Academic Program Development of
Jiangsu Higher Education Institutions.
O2/O3 References
APHA (1998) Standard Methods for the Examination of Water and
Wastewater, 20th edn. American Public Health Association
Washington DC, USA
Bakheet B, Yuan S, Li Z, Wang H, Zuo J, Komarneni S, Wang Y (2013)
Electro-peroxone treatment of Orange II dye wastewater. Water Res
47:6234–6243
Bakheet B, Qiu C, Yuan S, Wang Y, Yu G, Deng S, Huang J, Wang B
(2014) Inhibition of polymer formation in electrochemical degrada-
tion of p-nitrophenol by combining electrolysis with ozonation.
Chem Eng J 252:17–21
Barros WRP, Reis RM, Rocha RS, Lanza MRV (2013) Electrogeneration
of hydrogen peroxide in acidic medium using gas diffusion elec-
trodes modified with cobalt (II) phthalocyanine. Electrochim Acta
104:12–18
Beltrán FJ, Ovejero G, Rivas J (1996) Oxidation of polynuclear aromatic
hydrocarbons in water.4. Ozone combined with hydrogen peroxide.
Ind Eng Chem Res 35:891–898
Brillas E, Sires I, Oturan MA (2009) Electro-fenton process and related
electrochemical technologies based on fenton’s reaction chemistry.
Chem Rev 109:6570–6631
Chang J, Chen ZL, Wang Z, Shen JM, Chen Q, Kang J, Yang L, Liu XW,
Nie CX (2014) Ozonation degradation of microcystin-LR in aque-
ous solution: intermediates, byproducts and pathways. Water Res
63:52–61
Cho M, Chung H, Choi W, Yoon J (2004) Linear correlation between
inactivation of E. coli and OH radical concentration in TiO2 photo-
catalytic disinfection. Water Res 38:1069–1077
Fan XL, Restivo J, Orfao JJM, Pereira MFR, Lapkin AA (2014) The role
of multiwalled carbon nanotubes (MWCNTs) in the catalytic ozon-
ation of atrazine. Chem Eng J 241:66–76
García-Gómez C, Drogui P, Zaviska F, Seyhi B, Gortáres-Moroyoqui P,
Buelna G, Neira-Sáenz C, Estrada-alvarado M, Ulloa-Mercado RG
(2014) Experimental design methodology applied to electrochemi-
cal oxidation of carbamazepine using Ti/PbO2 and Ti/BDD elec-
trodes. J Electroanal Chem 732:1–10
Garcia-Morales MA, Roa-Morales G, Barrera-Diaz C, Bilyeu B, Rodrigo
MA (2013) Synergy of electrochemical oxidation using boron-
doped diamond (BDD) electrodes and ozone (O3) in industrial
wastewater treatment. Electrochem Commun 27:34–37
Gedam N, Neti NR, Kashyap SM (2014) Treatment of recalcitrant cap-
rolactam wastewater using electrooxidation and ozonation. Clean-
Soil Air Water 42:932–938
Giorgi L, Antolini E, Pozio A, Passalacqua E (1998) Influence of the
PTFE content in the diffusion layer of low-Pt loading electrodes
for polymer electrolyte fuel cells. Electrochim Acta 43:3675–3680
Hermosilla D, Merayo N, Gasco A, Blanco A (2015) The application of
advanced oxidation technologies to the treatment of effluents from
15820
the pulp and paper industry: a review. Environ Sci Pollut Res Int 22:
168–191
Hsu YC, Chen YF, Chen JH (2003) Peroxone process for RO-16 and RB-
19 dye solutions treatment. J Environ Sci Heal A 38:1361–1376
Huang H, Zhang W, Li M, Gan Y, Chen J, Kuang Y (2005) Carbon
nanotubes as a secondary support of a catalyst layer in a gas diffusion
electrode for metal air batteries. J Colloid Interface Sci 284:593–599
Khataee AR, Safarpour M, Zarei M, Aber S (2011) Electrochemical gen-
eration of H2O2 using immobilized carbon nanotubes on graphite
electrode fed with air: Investigation of operational parameters. J
Electroanal Chem 659:63–68
Khataee A, Safarpour M, Vahid B, Akbarpour A (2014) Degrading a
mixture of three textile dyes using photo-assisted electrochemical
process with BDD anode and O2-diffusion cathode. Environ Sci
Pollut Res 21:8543–8554
Kishimoto N, Morita Y, Tsuno H, Oomura T, Mizutani H (2005)
Advanced oxidation effect of ozonation combined with electrolysis.
Water Res 39:4661–4672
Kishimoto N, Nakagawa T, Asano M, Abe M, Yamada M, Ono Y (2008)
Ozonation combined with electrolysis of 1,4-dioxane using a two-
compartment electrolytic flow cell with solid electrolyte. Water Res
42:379–385
Li Z, Yuan S, Qiu C, Wang Y, Pan X, Wang J, Wang C, Zuo J (2013)
Effective degradation of refractory organic pollutants in landfill
leachate by electro-peroxone treatment. Electrochim Acta 102:
174–182
Li X, Wang Y, Yuan S, Li Z, Wang B, Huang J, Deng S, Yu G (2014)
Degradation of the anti-inflammatory drug ibuprofen by electro-
peroxone process. Water Res 63:81–93
Li Y, Shen W, Fu S, Yang H, Yu G, Wang Y (2015) Inhibition of bromate
formation during drinking water treatment by adapting ozonation to
electro-peroxone process. Chem Eng J 264:322–328
Maja M, Orecchia C, Strano M, Tosco P, Vanni M (2000) Effect of
structure of the electrical performance of gas diffusion electrodes
for metal air batteries. Electrochim Acta 46:423–432
Neta P, Dorfman LM (1968) Pulse radiolysis studies, XIII: rate constants
for the reaction of hydroxyl radicals with aromatic compounds in
aqueous solutions. Adv Chem Ser 81:222–230
The author has requested enhancement of the downloaded file. All in-text references underlined in blue are linked to publications on ResearchGate.
Environ Sci Pollut Res (2015) 22:15812–15820
Oturan N, Wu J, Zhang H, Sharma VK, Oturan MA (2013)
Electrocatalytic destruction of the antibiotic tetracycline in aqueous
medium by electrochemical advanced oxidation processes: Effect of
electrode materials. Appl Catal B-Environ 140:92–97
Parsa JB, Golmirzaei M, Abbasi M (2014) Degradation of azo dye C.I.
Acid Red 18 in aqueous solution by ozone-electrolysis process. J Ind
Eng Chem 20:689–694
Qiu C, Yuan S, Li X, Wang H, Bakheet B, Komarneni S, Wang
Y (2014) Investigation of the synergistic effects for p-
nitrophenol mineralization by a combined process of ozona-
tion and electrolysis using a boron-doped diamond anode. J
Hazard Mater 280:644–653
Soltani RDC, Rezaee A, Khataee AR, Godini H (2012) Electrochemical
generation of hydrogen peroxide using carbon black-, carbon nano-
tube-, and carbon black/carbon nanotube-coated gas-diffusion cath-
odes: effect of operational parameters and decolorization study. Res
Chem Intermed 39:4277–4286
Vahid B, Khataee A (2013) Photoassisted electrochemical recirculation
system with boron-doped diamond anode and carbon nanotubes
containing cathode for degradation of a model azo dye.
Electrochim Acta 88:614–620
Vijayalakshmi P, Raju GB, Gnanamani A (2011) Advanced Oxidation
and Electrooxidation As Tertiary Treatment Techniques to Improve
the Purity of Tannery Wastewater. Ind Eng Chem Res 50:10194–
10200
Wang Y, Li X, Zhen L, Zhang H, Zhang Y, Wang C (2012) Electro-
Fenton treatment of concentrates generated in nanofiltration of bio-
logically pretreated landfill leachate. J Hazard Mater 229–230:115–
121
Xu WY, Li P, Dong B (2010) Electrochemical disinfection using the gas
diffusion electrode system. J Environ Sci-China 22:204–210
Yuan S, Li ZX, Wang YJ (2013) Effective degradation of methylene blue
by a novel electrochemically driven process. Electrochem Commun
29:48–51
Zhang Z-y, Xu X-c (2014) Wrapping carbon nanotubes with poly
(sodium 4-styrenesulfonate) for enhanced adsorption of methylene
blue and its mechanism. Chem Eng J 256:85–92