Accepted Manuscript

Discharge characteristics of Plasma induced by water drop and its potential for
water treatment

Xiaoping Wang, Dengmei Zhao, Xuemei Tan, Yixia Chen, Zihan Chen, He Xiao

PII: S1385-8947(17)31229-9
DOI: http://dx.doi.org/10.1016/j.cej.2017.07.082
Reference: CEJ 17348

To appear in: Chemical Engineering Journal

Received Date: 13 February 2017

Revised Date: 1 July 2017
Accepted Date: 14 July 2017

Please cite this article as: X. Wang, D. Zhao, X. Tan, Y. Chen, Z. Chen, H. Xiao, Discharge characteristics of Plasma
induced by water drop and its potential for water treatment, Chemical Engineering Journal (2017), doi: http://
dx.doi.org/10.1016/j.cej.2017.07.082

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Discharge characteristics of Plasma induced by water drop
and its potential for water treatment

Xiaoping Wang, Dengmei Zhao, Xuemei Tan, Yixia Chen, Zihan Chen, He Xiao

College of Environment and Resources, Chongqing Key Laboratory of Catalysis and

Environmental New Material, Chongqing Technology and Business University, Chongqing
400067, China

Email: xpwang@ctbu.edu.cn

Abstract
Water surface discharge is an effective water treatment method. However, the industrial
application of this technology is limited by some problems including the difficulty to get a
large area of uniform discharge plasma and the insufficient durability of high voltage
electrodes. To solve these problems, the water to be treated was dispersed as droplets and
sprayed to the high voltage electrodes to make a water drop induced discharge system. The
discharge photographs indicated that when dripping down in the discharge gap, water drops
can induce the happening of discharge. The discharge channels concentrated near the water
drops and the discharge intensity changes with water flow rate. This feature could reduce the
requirement for the uniformity of electrodes as the uniform discharge could be obtained by
making a uniform water distribution. In addition, both the discharge channels and active
species were concentrated on or very close to the water drops to be treated. The transfer
distance of active species was shortened, and then both the chemical reaction efficiency and
the energy efficiency could be increased. Ibuprofen degradation results showed that the
ibuprofen can be removed exceeding 99% after 1 hour treatment. The maximum energy yield
exceeded 10 g/kWh. Factors including discharge voltage, water flow rate, water conductivity
and electrode gap distance influenced the ibuprofen removal.

Keywords: Water drop induced discharge; Plasma; Water treatment; Ibuprofen removal
1. Introduction
Atmospheric air discharge above water surface (water surface discharge) is an effective
technology on water treatment [1-3]. Malik compared the energy efficiency of various plasma
water treatment reactors in his review [4], and found that the energy efficiency of water
surface discharge is much higher than that of in-water discharge. Especially, the energy
efficiency of the pulsed discharge over a thin water film and the pulsed discharge with water
spray are almost 2 orders of magnitude higher. Stratton el. al. tested several kinds of water
surface discharge reactors featured the point-plate electrode configuration and proposed a
general design principle of water surface discharge reactor, which has some guiding
significance for the design and application of this kind of reactor [5]. However, some
problems affecting the application of such discharge reactor are still unsolved including the
uniformity of discharge and the durability of high voltage electrodes.
The prominent configuration feature of a water surface discharge reactor is suspending the
high voltage electrode over the water surface and putting the ground electrode in water [6].
The electrical discharge occurs in the air and on the air-water interface. Active species (e.g.,
O, OH, O3 and so on) generated by gaseous electrical discharge dissolve in water through
the air-water interface, and react with pollutants or microorganisms in water. Based on this,
several typical forms of water surface discharge reactors were used widely including the
needle (or multi-needles)-plate discharge reactor[7-9], nonmetallic electrode discharge
reactor[10,11], wire-plate discharge reactor[12], coaxial cylindrical wetted-wall discharge
reactor[13-15], water surface DBD reactor[16, 17], water spray discharge reactor [18-21] and
so on. These discharge reactors are difficult to make a large scale uniform discharge electrode
system which needs to satisfy two basic requirements: firstly, all high voltage electrodes have
the same geometrical morphology; secondly, all high voltage electrodes have the same
distance to the ground electrode or to the water surface. In practice, if these two requirements
are not satisfied, the discharge will propagate along the easiest pathway but other places rarely.
This problem harms the water treatment efficiency seriously especially for a large flux water
treatment system.
Another problem is the insufficient durability of high voltage electrodes. In traditional
water surface discharge reactors, the ablation of high voltage electrodes is serious because of
the much smaller thermal absorption and conduction of air compared with that of water.
Especially, because of the short lifetime of active species generated in discharge, the distance
between the high voltage electrode and the ground electrode is usually reduced to increase the
concentration and shorten the transfer pathway of active species. However, the smaller
discharge gap distance results in the lower spark breakdown voltage [22-24]. Even though a
spark discharge can lead to more active species, the ablation of the high voltage electrode is
also enhanced comparing to streamer or corona discharge.
To improve the behavior of water surface discharge, the water injection way is changed in
the present paper. The water to be treated was dispersed as droplets and sprayed upon the high
voltage electrodes. After that, the water drops dripped down from the high voltage electrode
to the ground electrode. Several benefits might be brought in this kind of discharge reactor.
Firstly, the high voltage electrodes could be protected against ablation by the water. Secondly,
a uniform discharge could be gotten by making a uniform water distribution. And the
requirement for the uniformity of electrodes would be reduced. Thirdly, the absorption of
active species could be enhanced when the water drops passing along the direction of
discharge channels. In addition, influenced by the water drops between the electrodes, the
electric field distribution and the discharge would be different from typical water surface
discharges.
2. Experimental
2.1 Experimental setup
Figure 1 shows the diagram of the water drop induced discharge system. This discharge
system used some existing designs for reference [25-27]. The water to be treated was
dispersed into droplets and sprayed to the high voltage electrode, and then flowed around the
tubular high voltage electrode. After leaving the high voltage electrode, the water drops
dripped down in the discharge gap. Based on this, three different discharge reactors were used
in the experiment.
Reactor a (Ra in figure 1) is a single water spout discharge reactor. The water spout was a
plastic pipe with an inner diameter of 1.5 mm. The high voltage electrode was a stainless steel
pipe with an outer diameter of 3 mm and a length of 7 cm. The ground electrode was a
graphite plate with a length of 9 cm and a width of 6 cm. Several small holes were made
through the ground electrode to reduce the residual water on the ground electrode. When
changing the water spraying rate, the gap distance between the high voltage and the ground
electrode could be maintained as a constant (1 cm in Ra).
Reactor b (Rb in figure 1) is a double water spout discharge reactor. The only difference
between Ra and Rb was the number of water spouts. The distance between two water spouts
was 2.5 cm.
Reactor c (Rc in figure 1) is a multiple point discharge reactor. The water was dispersed by
a shower. The high voltage electrode was 15 stainless steel pipes with an outer diameter of 3
mm and a length of 8 cm. The distance between two adjacent steel tubes was 2 mm. The
ground electrode was a stainless steel plate with a length of 9 cm and a width of 8 cm. The
distance between the high voltage electrode and the ground electrode could be changed from
0.5 cm to 3.0 cm.
Figure 1. Sketch diagram of the water drop induced discharge system. Ra represents
reactor a, Rb represents reactor b, Rc represents reactor c.
The high voltage power supply was a commercial product provided by Mianyang Shangwu
Electronic Technology Co., Ltd., which could output positive pulsed voltage with a maximum
peak voltage of 70 kV and a maximum pulse repetition frequency of 200 pulses per second
(pps). The time duration of a single pulse was about 500 ns and the voltage rising edge is
about 60 ns (the pulse duration and the rising edge might vary with the load and the discharge
type). In this paper, the pulse repetition frequency was kept as 100 pps and the peak voltage
was adjusted as needed. The typical voltage and current waveforms were shown in figure2.

(a)
(b)
Figure 2. Typical voltage and current waveforms. (a) streamer discharge; (b) spark
discharge.
2.2 Experimental procedure
When studying the discharge characteristics including discharge photographs, voltage
waveforms and current waveforms, the tap water with a conductivity of 290 S/cm was used.
When studying the water treatment capacity, the ibuprofen containing water with a
concentration of 10 mg/L was used. The solution was prepared by dissolving ibuprofen in
50mg/L sodium hydroxide solution.
2.3 Analytical procedures
The pulsed voltage and current supplied to the discharge reactors were monitored by a
digital oscilloscope (Tektronix TDS 2014C, USA) with a 1:1000 high voltage probe
(Tektronix P6015A, USA) and a current probe (Tektronix A622, USA). The discharge
photographs were taken by a digital single lens reflex (Canon EOS 70D, Japan) and a zoom
lens (Canon EF-S 18-135mm f/3.5-5.6 IS STM, Japan).
The concentration of ibuprofen was tested by a high performance liquid chromatography
(Techcomp LC 2050, China) with a C18 column (Agilent Extend-C18 5m 4.6250 mm,
USA). The mobile phase was acetonitrile, 0.1 mol/L monopotassium phosphate solution and
phosphoric acid (600:400:0.5). The flow rate of mobile phase was 1.0 ml/min. The total
organic carbon (TOC) was measured by a TOC analyzer (Shimadzu TOC-Vcph, Japan).
3. Results and Discussion
3.1 Discharge characteristics
3.1.1 Discharge photographs
 Discharge photographs were shown in figure 3, figure 4 and figure 5.
Figure 3 compared the discharge photographs with a single water spout at different flow
rates and applied voltages. The photographs were arrayed horizontally with the increase of
water flow rate and vertically with the increase of applied voltage. The water flow rate of 0
ml/min and 60 ml/min were not listed in figure 3, since no obvious discharge occurred.
As shown in figure 3, the discharge intensity increased with water flow rate. Taking as an
example the peak voltage of 16.6 kV, the discharge region and the discharge luminance
increased when increasing the water flow rate from 5 ml/min to 40 ml/min gradually, which
could be regarded as the evidence of the increase of discharge intensity. The similar trend
appeared when changing the applied peak voltage to 13.8 kV and 18.4 kV.
Figure 3 also indicated that the peak voltage influenced the discharge. Taking as an
example the water flow rate of 20ml/min, both the discharge region and the discharge
luminance increased when increasing the applied peak voltage.
It should be note that not only the discharge intensity but also the discharge form converted
with water flow rate and applied voltage. At low water flow rates and applied peak voltages,
streamer discharge was likely to be ignited. When increasing the water flow rate and the
applied voltage, the discharge converted from streamer discharge to hybrid streamer-spark
discharge, and pure spark discharge at last.
It should also be pointed out that no obvious discharge occurred in two cases in the
experimental condition. The first was no water drop in the discharge gap (0 ml/min), and the
second was so large a water flow rate that a continuous water column connected the high
voltage electrode and the ground electrode (60 ml/min). For the former, the discharge was not
ignited as 18.4 kV was much lower than the gaseous discharge inception voltage of Ra (about
30 kV/cm). For the later, the discharge electrodes were communicated by the continuous
water column. The discharge system became like a liquid phase discharge, and the discharge
inception voltage was even higher.
In summary, the water drops in the discharge gap played the role of an inducer of discharge,
which made the discharge easier. This result provided a convenient method for improving and
controlling the water treatment system by changing the water flow rate. In addition, the water
drop induced discharge could bring another benefit for the water treatment. The discharge
channels existed only in the places with water drops. As a result, the active species generated
in electrical discharge were concentrated on or very close to the water surface. The transfer
distance of the active species could be reduced. Correspondingly, the chemical reaction
efficiency and the energy efficiency would be increased.
Figure 3. Discharge photographs with a single water spout at different water flow rates
and applied voltages (aperture value: 6.3, exposure time: 5 s, shutter speed: 1600).
Figure 4 showed the discharge photographs with two water spouts to verify the possibility
of simultaneous discharge in multiple positions. The discharge intensities of these two
discharge positions were not exactly the same because of the difficulty to ensure the same
water flow rate of each water spout. The discharges could occur in both two positions
synchronously when the water drops being sprayed from these two spouts. So it is reasonable
to suppose that the discharges could occur at all water dripping places when adding spouts.
This result provided the basis to make a water drop induced discharge reactor with multiple
discharge positions, which is important to design a water treatment reactor using this kind of
discharge.

Figure 4. Discharge photographs with two water spouts (aperture value: 6.3,
exposure time: 10 s, shutter speed: 1600).
In order to further verify the feasibility of such discharge for water treatment, a multiple
position discharge reactor with a shower dispersor was designed (Rc in figure 1). The
discharge photographs at different applied voltages were shown in figure 5. As can be seen,
multiple water spray points could trigger multiple discharge channels. The discharge density
increased with applied voltage. The change of discharge with water flow rate was similar to
that of single and double spouts discharge, which was not displayed on a photograph
repeatedly.
Figure 5. Discharge photographs with a shower dispersor (aperture value 6.3,
shutter speed 1600): (a) exposure time 5 s; (b) exposure time 5 s; (c) exposure time 5 s; (d)
exposure time 0.1 s.

3.1.2 Voltage-current characteristic

The voltage-current relationships of the single spout discharge reactor at different water
flow rates were shown in figure 6.
It can be seen from figure 6 that the peak current increased when the water flow rate
increasing from 0 ml/min to 40 ml/min gradually. However, if the water flow rate increased to
60 ml/min, the peak current became a little larger than that of 0 ml/min but smaller than 10
ml/min. This result was consistent with the analysis of the discharge photographs. When
increasing the water flow rate in the range of 0 ml/min to 40 ml/min, the discharge intensity
of the same applied voltage increased. But for the water flow rate of 60 ml/min, no water drop
but a continuous water column existed in the discharge gap. As a result, the electrodes were
connected by the water column and the current recorded was the leak current if the applied
voltage below the discharge inception voltage. Because of the little change of water
conductivity in the experiment, the peak current increased with peak voltage almost linearly.
Another conclusion educed from the voltage-current characteristic was that the
streamer-spark transition voltage decreased with the increase of water flow rate. As can be
seen from figure 6, the peak voltage at which the peak current increased rapidly decreased
when increasing the flow rate. This result indicated that the more water drops in the discharge
pathway, the more likely the streamer discharge converts to spark discharge.
Figure 6. The voltage-current relationship of single spout discharge.
The voltage-current relationships of the single spout discharge, double spout discharge and
shower dispersor discharge were compared in figure 7. It could be found that the peak current
increased when increasing the water spray points. In other words, the more water spray points,
the higher energy needed to sustain the discharge. It should be noticed that the current value
was not linearly related to the amount of water spraying point. The partial currents of each
spray points were not linearly superposed because of the possible phase differences.
Figure 7 Comparison of the voltage-current relationships of the single spout discharge
(flow rate 10 ml/min), double spout discharge (flow rate 10 ml/min), shower dispersor
discharge (flow rate 950 ml/min).

3.2 Discharge mechanism of water drop induced discharge

To research the mechanism of water drop induced electrical discharge, the background light
was enhanced when taking photographs. Under this condition, the water drops could be
identified in the photographs as shown in figure 8.
The propagation of the discharge channels could be seen from the discharge photographs.
At the peak voltage of 16.6 kV, only streamer discharge could be ignited. The streamer
channels propagated from the water drop to the ground electrode. No discharge channel
propagated between the high voltage electrode and the water drops.
However, when increasing the peak voltage to 17.6 kV, both the streamer discharge and the
spark discharge were ignited. The spark channels initiated from the high voltage electrode and
propagated along the surface of the water drops and finally reached to the ground electrode.

Figure 8 Discharge photographs under stronger background light (aperture value: 6.3,
exposure time: 5 s, shutter speed: 1600).
Based on the discharge channel propagation shown in figure 8, the mechanism of water
drop induced discharge was speculated as the electric field distortion caused by water drops as
shown in figure 9 (a) and (b). When no water drops existing in the discharge gap, the electric
field distributed uniformly between the electrodes if ignoring the edge effect of the electrodes.
In this case, the discharge could be ignited only if the electric field exceeding the inception
electric field of the gaseous discharge.
Once spraying the water drop into the gap between the electrodes, the electric field was
bundled by the water drop and the free charges in the water drop were separated to its ends.
As the water drop dripping down, the distance between the water drop and the ground
electrode decreased gradually, and then the electric field between the water drop and the
ground electrode increased correspondingly. When the water drop dripping to a certain height,
the electric field between the water drop and the ground electrode became high enough to
ignite the discharge, and then the streamer discharge occurred.
When increasing the applied voltage, the water drop height at which the streamer discharge
could be ignited also increased. When the applied voltage increasing high enough, the
discharge occurred before the water drop leaving the high voltage electrode and then the spark
voltage would be ignited. In this case, the water drop became a part of the high voltage
electrode, and the discharge gap distance was shortened. As a result, the spark discharge could
be ignited at a relatively lower voltage than that of no water drops.
It must be pointed out that if the discharge was not ignited before the water drop leaving
the high voltage electrode, no spark discharge would occur in the later process. The energy
carried by the water drop would be preferentially released in the form of streamer discharge
because of the lower inception electric field. In this case, the water drop played the role of an
energy carrier which transported the energy from the high voltage electrode to an appropriate
position to ignite the discharge.

Figure 9 Speculated mechanism of the water drop induced discharge: (a) electric field
distribution without water drop, (a) electric field distribution with water drop, (c)-(e)
discharge as the applied voltage increasing. (Lines in figure (a) and (b) represent the
electric field distribution)

3.3 Ibuprofen removal

3.3.1 Ibuprofen removal in the single spout discharge
(a)

(b)
(c)
Figure 10 Ibuprofen removal in the single spout discharge reactor: (a) effect of applied
voltage; (b) effect of water spraying rate; (c) effect of water conductivity.
The ibuprofen removal in the single spout discharge reactor (Ra) was shown in figure 10.
The ibuprofen solution volume treated was 150 ml. The single spout discharge system had a
less removal rate of ibuprofen because of the small discharge scale.
The effect of peak voltage on ibuprofen removal was shown in figure 10(a). The water flow
rate was 25 ml/min, and the water conductivity was 0.26 mS/cm. The removal rate increased
when increasing the peak voltage. For 14 kV, the removal rate was about 5% after 100 min
treatment. For 15 kV, the final removal rate was about 8%. For 15 kV, the final removal rate
was about 13%. For 17 kV, the final removal rate was about 20%.
The effect of water flow rate on ibuprofen removal was shown in figure 10(b). The peak
voltage was 17 kV, and the water conductivity was 0.26 mS/cm. As can be seen, the removal
rate increased when increasing the water flow rate (no continuous water column formed). The
final removal rates of 8.5 ml/min, 17 ml/min, 25 ml/min, 33.5 ml/min and 42 ml/min were
about 14%, 17%, 20%, 27% and 33% respectively.
The influence of water flow rate on ibuprofen removal could be derived from two reasons.
Firstly, the higher water flow rate resulted in the stronger discharge as shown in figure 3. As a
result, more active species were generated in the reactor. Naturally, more ibuprofen could be
removed. Secondly, a higher water flow rate resulted in a longer average residence time in the
discharge region, and then the absorption efficiency of active species could be improved. This
also improved the ibuprofen removal rate.
The influence of water conductivity on ibuprofen removal was shown in figure 10(c). The
peak voltage was 17 kV, and the water flow rate was 25 ml/min. The ibuprofen removal rate
increased gradually when the water conductivity increasing from 0.26 mS/cm to 2 mS/cm.
After that, the removal rate decreased with the increase of water conductivity. The discharge
intensity increased with water conductivity, which could enhance the generation of active
species. However, the increased inorganic salt would increase the consumption of active
species. As a result, the ibuprofen removal rate increased first and then decreased with the
increase of water conductivity.
3.3.2 Ibuprofen removal in the shower dispersor discharge

(a)
(b)

(c)
(d)
Figure 11 Ibuprofen removal in the shower dispersor discharge reactor: (a) effect of
applied voltage; (b) effect of water spraying rate; (c) effect of water conductivity; (d)
effect of discharge gap distance.

The ibuprofen removal in the shower dispersor discharge reactor (Rc) was shown in figure
11. The distance between the high voltage electrode and the ground electrode was 2 cm when
researching the effect of the peak voltage, the water flow rate and the initial water
conductivity. The ibuprofen solution volume treated was 1000 ml. As can be seen, the
ibuprofen removal rate was much higher than that of the single spout discharge reactor even
treated a larger volume of solution. This result indicated that the water drop induced discharge
could be scaled up to make a large flux water treatment reactor.
The effect of peak voltage on ibuprofen removal was shown in figure 11(a). The water flow
rate was 0.95 L/min, and the water conductivity was 0.26 mS/cm. For 19 kV, the removal rate
was about 35% after 100 min treatment. For 22.5 kV, the final removal rate was about 63%.
For 27 kV, the final removal rate was 100%. For 31 kV, the removal rate reached up to 100%
after 80 min treatment.
The effect of water flow rate on ibuprofen removal was shown in figure 11(b). The peak
voltage was 30 kV, and the initial water conductivity was 0.26 mS/cm. The removal rate
increased with water flow rate (no continuous water column formed). The removal rate of
0.95 L/min reached to 100% after 100 min treatment. For 1.7 L/min, 2 L/min, 2.65 L/min and
2.8 L/min, all ibuprofen were removed after 80 min treatment.
The influence of water conductivity on ibuprofen removal was shown in figure 11(c). The
peak voltage was 30 kV, and the water flow rate was 1.4 L/min. The ibuprofen removal rate
increased a little when increasing the water conductivity from 0.26 mS/cm to 2 mS/cm. After
that, the removal rate decreased with the increase of water conductivity.
The influence of discharge gap distance on ibuprofen removal was shown in figure 11(d).
The experiment was conducted at a peak voltage of 30 kV, a water flow rate of 1.4 L/min and
a water conductivity of 0.26 mS/cm. As can be seen from figure 11(d), the ibuprofen removal
rate increased with the decrease of discharge gap distance. The discharge intensity would
increase when decreasing the gap distance, which could generate more active species. As a
result, the ibuprofen removal would be enhanced.
3.4 Mineralization of ibuprofen
The TOC removal rate was compared with the ibuprofen removal rate in figure 12. As can
be seen, the TOC removal was much slower than ibuprofen removal. The ibuprofen removal
rate reached up to 100% after 80 minute treatment. However, the removal rate of TOC was
only 46% after 100 minute treatment. This result demonstrated that some organic by-products
of IBP remained in the solution. The similar results have been reported for IBP decomposition
by other AOPs.
The detailed pathway of ibuprofen decomposition was studied in our previous work[15]
and other researchers such as Magureanu[28]. So the removal mechanism of ibuprofen was
not involved in this paper repeatedly.

Figure 12 TOC removal in the shower dispersor reactor. (peak voltage 30 kV, water flow
rate 1.4 L/min, water conductivity 0.26 mS/cm, gap distance 2 cm)
3.5 Energy efficiency
The energy yield of ibuprofen removal (removed amount per kWh electric energy, g/kWh)
was shown in figure 13. The energy yield was calculated by the following formula,
V (C0 Ct )
G (1000 3600)
T UIdt ft
0
In which V was the solution treated, C0 and Ct were the ibuprofen concentration before and
after treated respectively, U was the voltage, I was the current, T was the pulse duration time,
f was the pulse repetition frequency (pulse per second), t was the water treatment time.
Though the removal rate increased with peak voltage in the preceding discussion, the
energy yield decreased with the increase of voltage as shown in figure 13. The energy yield
obtained in this paper was not very different from that obtained in a falling film discharge
reactor[15].

Figure 13 The energy efficiency of ibuprofen removal. (water flow rate 0.95 L/min,
water conductivity 0.26 mS/cm, gap distance 2 cm)
Conclusion
The water drop existed in the discharge gap could promote the discharge. The discharge
channels were concentrated in the positions with water drops. This feature made this kind of
electrical discharge has potential advantages in water treatment as follows.
(1) Because of the inducing effect of the water drop, the discharge inception voltage was
decreased. As a result, the required output capability of the power supply could be reduced
which would reduce the equipment cost and the running cost.
(2) Active species generated in electrical discharge were concentrated on or very close to
the water surface. Therefore, the transfer distance of active species was shortened and then
both the chemical reaction efficiency and the energy efficiency could be increased.
 (3) Discharges only occurred in the positions with water drops. Therefore, the discharge
uniformity could be improved by improving the uniformity of water distribution, which could
reduce the requirement for the uniformity of electrodes. In addition, the high voltage electrode
could be protected against ablation by the water drops.
(4) In a water drop induced discharge reactor, the pollutant removal rate was influenced by
the electrical parameters, the discharge gap distance, the water flow rate, the water
conductivity so on. The discharge reactor could be scaled up by making a large area of
uniform water distribution.

Acknowledgement
We appreciated the financial support to this study by National Natural Science Foundation
of China (No. 21406022), Chongqing Science and Technology Commission (No.
cstc2015jcyjA20017), Chongqing Education Commission (NO. Kj1500610) Key Laboratory
of Natural Medicine Research of Chongqing (NO. 1556034).

Reference
[1] Sato M., Tokutake T., Ohshima T., et al. Aqueous phenol decomposition by pulsed
discharges on the water surface [J]. IEEE Trans. Ind. Applicat., 2008, 44(5): 1397-1402.
[2] Krause H., Schweiger B., Schuhmacher J., et al. Degradation of the endocrine disrupting
chemicals (EDCs) carbamazepine, clofibric acid, and iopromide by corona discharge
over water [J]. Chemosphere, 2009, 75(2): 163-168.
[3] Kadowaki K., Sone T., Kamikozawa T., et al. Effect of water-surface discharge on the
inactivation of Bacillus subtilis due to protein lysis and DNA damage [J]. Biosci.
Biotechnol. Biochem., 2009, 73(9): 1978-1983.
[4] Malik M. A. Water purification by plasma: which reactors are most energy efficient ? [J].
Plasma Chem Plasma Process., 2010 30: 21-33.
[5] Stratton G. R., Bellona C. L., Dai F., et al. Plasma-based water treatment: Conception and
application of a new general principle for reactor design [J]. Chem. Eng. J., 2015, 273:
543-550.
[6] Grymonpre, D. R., Finney, W. C., Clark, R.J., et al. Hybrid gas-liquid electrical discharge
reactors for organic compound degradation [J]. Ind. Eng. Chem. Res., 2004, 43(9):
1975-1989.
[7] Bruggeman P., Guns P., Leys C., et al. Influence of the water surface on the glow-to-spark
transition in metal-pin-to-water electrode system [J]. Plasma Source Sci. Technol., 2008,
17(4):969-977.
[8] Sun B., Aye N. N., Gao Z., et al. Characteristics of gas-liquid pulsed discharge plasma
reactor and dye decoloration efficiency [J]. J. Environ. Sci., 2012, 24(5): 840-845.
[9] Shiota H., Itabashi H., Satoh K., et al. Phenol decomposition by pulsed-discharge plasma
above a water surface in oxygen and argon atmosphere [J]. Electr. Eng. JPN., 2013,
184(1): 297-304.
[10] Wang X. P., Lan T., Li Z. J., et al. Sulfite oxidation in seawater flue gas desulfurization
by plate falling film corona-streamer discharge [J]. Chem. Eng. J., 2013, 225: 16-24.
[11] Wang X. P., Zhang X. W., Lei L. C. High conductivity water treatment using water
surface discharge with nonmetallic electrodes [J]. Plasma Sci. Technol., 2013, 15(6):
528-534.
[12] Grabowski L. R., van Veldhuizen E. M., Pemen A. J. M., et al. Corona above water
reactor for systematic study of aqueous phenol degradation [J]. Plasma Chem. Plasma
Process, 2006, 26(1): 3-17.
[13] Faungnawakij K., Sano N., Charinpanitkul T. Modeling of experimental treatment of
acetaldehyde-laden air and phenol-containing water using corona discharge technology
[J]. Environ. Sci. Technol., 2006, 40(5): 1662-1628.
[14] Wang X. P., Li Z. J., Lei L. C. Model study of sulfite oxidation in seawater flue gas
desulfurization by cylindrical wetted-wall corona-streamer discharge [J], Chem. Eng.
Sci., 2013, 97: 7-15.
[15] Zeng J. H., Yang B., Li Z. J., et al. Degradation of pharmaceutical contaminant ibuprofen
in aqueous solution by cylindrical wetted-wall corona discharge [J], Chem. Eng.
J., 2015, 267: 282-288.
[16] Hijosa-Valsero M., Molina R., Schikora H., et al. Removal of cyanide from water by
means of plasma discharge technology [J], Water Res., 2013, 47(4):1701-1707.
[17] Hijosa-Valsero M., Molina R., Schikora H., et al. Removal of priority pollutants from
water by means of dielectric barrier discharge atmospheric plasma [J], J. Hazard.
Mater., 2013, 262(8): 664-673.
[18] Minamitani Y., Shoji S., Ohba Y., et al. Decomposition of Dye in Water Solution by
Pulsed Power Discharge in a Water Droplet Spray[J]. IEEE Transactions on Plasma
Science, 2008, 36(5): 2586-2591.
[19] Porter D., Poplin M. D., Holzer F., et al. Formation of Hydrogen Peroxide, Hydrogen,
and Oxygen in Gliding Arc Electrical Discharge Reactors With Water Spray[J]. IEEE
Transactions on Industry Applications, 2009, 45(2):.623-629.
[20] Handa T., Minamitani Y. The Effect of a Water-Droplet Spray and Gas Discharge in
Water Treatment by Pulsed Power[J]. IEEE Transactions on Plasma Science, 2009,
37(1):.179-183.
[21] Kobayashi T., Sugai T., Handa T., et al. The Effect of Spraying of Water Droplets and
Location of Water Droplets on the Water Treatment by Pulsed Discharge in Air[J]. IEEE
Transactions on Plasma Science, 2010, 38(10):.2675-2680.
[22] Bruggeman P., Slychen J.V., Degroote J., et al. Characteristics of atmospheric pressure
air discharges with a liquid cathode and a metal anode [J]. Plasma Source Sci. Technol.,
2008, 17(2): 431-438.
[23] Bozhko I. B., Kondratenko I. P., Serdyuk Y. V. Corona Discharge to Water Surface and
Its Transition to a Spark [J]. IEEE Trans. Plasma Sci., 2011, 39(5): 1228-1233.
[24] Bruggeman P., Graham L., Degroote J., et al. Water surface deformation in strong
electrical fields and its influence on electrical breakdown in a metal pinwater electrode
system [J]. J. Phys. D: Appl. Phys., 2007, 40(16): 4779-4786.
[25] Machala Z., Tarabova B., Hensel K., et al. Formation of ROS and RNS in Water Electro
Sprayed through Transient Spark Discharge in Air and their Bactericidal Effects[J].
Plasma Processes & Polymers, 2013, 10(7): 649-659.
[26] Jiang S., Wen Y., Liu K. Investigation of pulsed dielectric barrier discharge system on
water treatment by liquid droplets in air[J]. IEEE Transactions on Dielectrics &
Electrical Insulation, 2015, 22(4): 1866-1871.
[27] Kornev J., Yavorovsky N., Preis., et al. Generation of Active Oxidant Species by Pulsed
Dielectric Barrier Discharge in Water-Air Mixtures[J]. Ozone Science & Engineering,
2006, 28(4): 207-215.
[28] Magureanu M., Mandache N. B., Parvulescu V. I. Degradation of pharmaceutical
compounds in water by non-thermal plasma treatment [J]. Water Research, 2010, 44(11):
3445-3453.
Highlights

Improved the behavior of water surface discharge by changing the water injection way in
reactor.
The inducing effect and the mechanism of water drops on discharge were discussed.
Effects of peak voltage, water flow rate, water conductivity and discharge gap distance
on pollutant removal rate were considered.
The reactor was scaled up step by step to get a large flux treatment reactor.