Hydrogen and Hydrogen Peroxide formation in the

AC water-spray gliding arc reactor

R. Burlica, B. Hnatiuc, E. Hnatiuc
Technical University Gh. Asachi Iasi, Romania
E-mail: rburlica@ee.tuiasi.ro

Abstract-The goal of this paper is to study the formation
rates and energy yields of H
2
and H
2
O
2
formed from pure water
exposed to a non-thermal AC plasma-gliding arc reactor
equipped with a spray nozzle. The gliding arc reactor utilizes an
AC electrical discharge developed in a gas flowing between two
electrodes supplied by one high voltage transformer. Previous
works [1, 2, 3] showed that the efficiency of reactive and
molecular species formation in water treated by non-thermal
plasma depends on the gas-liquid interfacial contact area
between the water and the plasma zone. Spraying the liquid
through a special two-way nozzle directly into the plasma zone
is an effective method to enhance the efficiency of chemical
species formation in gliding arc reactors. In the present work,
the focus is on hydrogen formation from water spray exposed to
plasma generated by an AC gliding arc discharge using Ar as
carrier gas for different electrical powers of the discharge and
for different water flow rates. The formation rates of hydrogen
and hydrogen peroxide were also determined.

I. INTRODUCTION
Recently a great attention has been devoted to hydrogen
production technologies because hydrogen to energy
conversion, especially in fuel cells, does not lead to the
emission of hazardous byproducts and greenhouse gases. In
addition to water electrolysis, hydrogen is normally produced
from the reforming of natural gas, methanol, and gasoline via
catalytic conversion, all of which require a large amount of
energy. Recently, non-thermal plasma technologies have been
proven to be effective for the conversion of various organic
compounds into hydrogen [1]. Non-thermal plasma generates
highly energetic species (e.g., electrons, ions, excited atoms
and molecules) in the gas phase at ambient gas temperature
and atmospheric pressure. These energetic species interact
with water molecules leading to the formation of OH radicals,
H
2
O
2
, H
2
, O
2
and nitrates (nitrates in the case when air or
nitrogen is used as carrier gas). The goal of the work
reported in this paper is to determine the formation rates and
energy costs of molecular hydrogen and hydrogen peroxide
formation from liquid water spray processed through an AC
gliding arc reactor.
A gliding arc is an electrical discharge formed in a gas
flowing between two divergent electrodes which maintain the
non-thermal characteristics of the plasma [1-8].
It is known that the efficiency of production of chemically
active species in water treated by gliding arc depends, in part,
on the gas-liquid interfacial contact area between the liquid
and the plasma zone, therefore small water droplets have been
sprayed through a gas-liquid injection nozzle directly into the
plasma formed between the electrodes [3]. In order to
enhance the gliding arc efficiency through increasing of the
specific energy yield of hydrogen and hydrogen peroxide
formation an AC plasma gliding arc of relatively low power
(max 36 W) was utilized in a reactor with two divergent
electrodes placed in a narrow chamber meant to confine the
plasma and force the water droplets to contact the high energy
species [9]. The low power used here is such that the water
droplets do not vaporize within the reactor and it is
conjectured that these condensed water droplets enhance the
energy yield for H2 due to suppression of radical quenching
reactions.
In the present work, the focus is on hydrogen formation
from water spray exposed to plasma generated by a gliding
arc with a argon as carrier gas (which have been found to
have the highest efficiency for reactive species formation
[3]), for different electrical powers of the discharge and for
different water flow rates. Additional information on reactive
species formed in liquid phase with Air and Nitrogen as
carrier gases is also given. The formation rate and energy
efficiency yield of hydrogen peroxide was also determined.

II. EXPERIMENTAL
The gliding arc reactor utilizes two stainless steel divergent
electrodes attached to a ceramic support placed between two
rectangular glass sheets. Between the stainless steel
electrodes and the glass sheets two Teflon layers of the same
shape as the electrodes are placed in order to avoid the direct
contact between glass and steel. The reactor dimensions are
shown in Fig. 1.
The de-ionized water to be treated is sprayed directly into
the plasma formed between the electrodes through an
injection nozzle equipped with one port for the gas injection
and one port for the liquid, see Fig.2. The nozzle consists of
two concentric tubes attached to a Swagelok Tee (1/16 inch).
The two other ports of the Tee are used for gas and water
injection, Fig. 2. The water is carried with the gas flow
through the outlet port and it is atomized due the high
velocity of the carrier gas, which assures the disintegration of
the liquid. The very small diameter of the exit tube (=0.15
mm) led to an increase in relative exit velocity of the gas.
This ensured a better droplet density formation while
reducing the consumption of the carrier gas (Q
g
=2L/min) over
the system used in the previous study [8]. A denser spatial
1355 978-1-4244-7020-4/10/$26.00 '2010 IEEE
2010, 12th International Conference on Optimization of Electrical and Electronic Equipment, OPTIM 2010
distribution of the droplets leads to a better axial flow
direction through the plasma zone formed between the two
divergent electrodes and, therefore, an increasing of stability
of the discharge has been observed.


The gas cylinder was connected to the straight port of the
Swagelok Tee, providing gas flow through the inner tube of
the nozzle, and the water was injected by a pump through the
perpendicularly port of the tee. A high pressure pulse
injection pump was used to feed the liquid port of the nozzle.























The aerosolized water droplets formed in the flowing gas
provide a large contact surface area with the plasma, affecting
the chemical reactions between the active species formed in
the gas and water droplets. The input energy is not sufficient
to vaporize the water droplets and can lead to only maximum
10
o
C degrees rise in water temperature. The temperature of
the treated water was also measured at the outlet of the
reactor and was found to be 5-10
0
C higher than the initial
water temperature injected into the reactor, depending on the
power of the discharge. The electrical discharge forms at the
minimum distance between the electrodes and glides along
their edges in a mixture of gas (e.g., air, nitrogen, or argon)
containing the very fine droplets of water to be treated. In the
present study the gas flow rate, Q
g
, is approximately 2L/min
and the water flow rate, Qw, varies for different experimental
situations between 2 and 20 mL/min.
In the experimental setup, Fig. 3, the electrodes are
connected to an AC power supply (7 kV / 50 mA / 50 Hz),
which is an AC high voltage transformer type with magnetic
shunt, very suited to produce electrical discharges. The
power of the discharge was about 36 W for a current of
approximately 30 mA. Measurements of H
2
and H
2
O
2
were
conducted for different values of the electrical power injected
in the discharge and water flow rates. The liquid phase water
is collected in a glass vessel at the bottom of the reactor after
passing through the plasma and the gas is directed to a
sampling vessel. The H
2
O
2
formed in the liquid phase is
measured using an UV - Visible Shimadzu spectrophotometer
(UVmini 1240).



















The conductivity of the water collected at the bottom of the
reactor has been measured using a multi- parameter analyzer
Consort C533 type. Hydrogen was measured using a
Shimadzu gas chromatograph (GC2010) with thermal
conductivity detector. The gas samples were taken at the
outlet of the reactor after the water droplets were removed
with a liquid trap. The voltage and current waveforms of the
discharge were measured with a Tektronix TDS210
oscilloscope. The power was calculated for an average of 3
measurements of the voltage and the current. The real time
sampling rate of the oscilloscope was up to 1GS/s. The
voltage of the discharge was measured with a high-voltage

Fig. 3. Experimental set-up. E1, E2 -electrodes, P- pump, SH-shunt,
OSC- oscilloscope, TRAF- AC high voltage transformer, VAC-
variable AC transformer, HVD-high voltage probe


Fig. 2. The geometry of the nozzle.


Fig. 1. Glidarc reactor configuration.
1356
probe (HVP), 1/1000 connected to the electrodes. The current
was measured on a shunt to the ground of 100 in the
secondary of the transformer. The waveform for the current
and voltage are shown in Fig. 4.
Three repeated measurements for gaseous H
2
were taken
after 5 min. of reactor operation. It took 2 min to analyze
each sample and samples were obtained for 4, 6, 10, 15 and
20 mL/min water flow rates. The average of the three
repeated values of the hydrogen concentration in the gas
phase was reported. The hydrogen peroxide concentration in
water was measured after 10 min for the same values of the
water flow rates.


















The current waveform represented in Fig. 4 show a
periodic variation of the current and maximum value for
voltage in the moment of discharge break down at the
minimum distance of the electrodes. Since the discharge
glides along the divergent electrodes, the voltage values
follow a quasi-constant pattern until the discharge extinguish
at the maximum distance between the electrodes.


III. RESULTS AND DISCUSSIONS
In the confined space between the electrodes, due to the
water spray, the amount of water molecules in the plasma
zone is high and they are subjected to an intense electronic
and ionic bombardment leading to the formation of chemical
active species in water droplets [8-11]. The active species
generated in water play an important role in H
2
and H
2
O
2
production.
With either air or N
2
carrier gas, nitrogen oxides form in
the plasma. Nitrogen oxides can form nitrites which react
with hydrogen peroxide to form nitrate [7]. The nitrates affect
the conductivity and pH through the formation of acids and
ions in water [8-10] and reduce the formation of hydrogen in
gas phase and hydrogen peroxide in water droplets, reactions
(5-9). Ionization by electron/ion impact [8-9] may occur by
the reactions:
H
2
O + e
-
H
+
+ OH* +2e
-
(1)
2H
2
O + e
-
H
3
O
+
+OH
-
+ e
-
(2)
H
2
O
2
may also form through an overall reaction such as
2H
2
O H
2
O
2
+ H
2
(3)
H
2
and O
2
respectively, may also form through an overall
reaction such as
H
2
O H
2
+ O
2
(4)
In the case of air and nitrogen as carrier gas the formation
of NOx occur by reactions:
N
2
+ e
-
2N + e
-
(5)
O
2
+ e
-
2O + e
-
(6)
N + O NO (7)
NO + O NO
2
(8)
The nitrogen oxides affect the pH and the conductivity of
the water through the formation of ions and acids described
by reactions [8]:
NO
2
+ OH HNO
3
(9)
The conductivity increase is the largest with air as the
carrier, up to 800 S, gas due to the highest nitrate formation.
Argon has the lowest conductivity, about 50 S, and an
insignificant concentration of nitrates formed in water
(presumably due to traces of air in the reactor). The trends
reported here are similar to those in previous work [9].
The concentration of H
2
O
2
in the liquid phase for different
power values and water flow rates is shown in Fig. 5. The
highest concentration of hydrogen peroxide, about 10 mM,
was found for the Ar carrier gas and for 2 mL/min water flow
rate and for 36W power of the discharge [8, 9, 13]. Contrary
with the case of higher power alternating current gliding arc
discharges previously reported [8-9], a higher concentration
of H
2
O
2
was measured in solutions sprayed into the plasma.
This result is likely due to the fact that the low power AC
discharge requires less energy, therefore leading to lower gas
and liquid temperatures and rapid quenching of the plasma.
The rapid quenching may reduce degradation reactions of OH
radicals [15,16, 17].


















The hydrogen peroxide formation rate for different water
flow rates with Ar as carrier gas and for different values of


Fig. 4. Voltage and current waveforms of the discharge (voltage waveform
1V/div, current waveform 10 V/div, time base 10 ms).
0
2
4
6
8
10
12
0 2 4 6 8 10 12 14 16 18 20 22
Qw (mL/min)
H
2
O
2

(
m
M
)
7W
16W
24W
30W
36W

Fig. 5. Hydrogen peroxide concentrations for different power values
and water flow rates.
1357
the discharge power are shown in Fig. 6. Fig. 6 shows a
maximum for H
2
O
2
formation rate in water treated by plasma
for Qw = 10mL/min and 36 W power of the discharge. For
higher water flow rates the efficiency started to decrease due
perhaps to the thermal quenching of the plasma by water
droplets.



The energy yields for hydrogen peroxide formation for
different water flow rates and for different values of the
power are shown in Fig. 7.


In terms of energy efficiency of hydrogen peroxide
production the maximum value of 3.5 g/kWh have been
obtained for the case of the lowest power of the discharge
(7W) and for the same 10 mL/sec water flow rate. That can
be explained by the fact that injecting less energy into the
discharge the water temperature increase less than 5
o
C
therefore the most part of the energy is used for reactive
species generation in aerosol mixture and not for heating the
water droplets. The plasma is in this case is fast quenching
reducing the recombination process of OH radicals to H
2
O.
For the power P=7W, for which the energy efficiency is the
highest, the pH and conductivity of the water after treatment
in pulse glidarc reactor have been also measured.
In the case of the AC gliding arc discharge reactor the
solution pH decreased from an initial value of 5.8 to a final
value near 4 after 1 pass through the reactor, Fig.8.
0
1
2
3
4
5
6
7
0 2 4 6 8 10 12 14 16 18 20 22
Qw (mL/min)
p
H

Fig. 8. The evolution of the pH of the water treated in Glidarc reactor with
Ar as carrier gas.

The decreasing of the pH does not differ significantly from
the initial pH of the water for Ar as carrier gas, which is
consistent with previous work [8, 9]. The highest gradient of
the pH dropping was found to be for 2 mL/min water flow
rate followed by a pH stabilization around 4 after 4mL/min
water flow rate.
The increase of water acidity is accompanied by a
increase of water conductivity. The evolution of conductivity
for the AC gliding arc reactor is shown in Fig. 9.















Fig. 9. The evolution of water conductivity

Similar with the evolution of pH the conductivity increase
up to 24 S/cm for 2mL/min water flow rate. For water flow
rates higher than 4 mL/min the conductivity has the tendency
to level off around the value of 14 S/cm. The decreasing of
conductivity for higher flow rates of the water treated by
plasma is due to a process of dilution of ionic species present
in water.

0
0.1
0.2
0.3
0.4
0.5
0.6
0 2 4 6 8 10 12 14 16 18 20 22
Qw (mL/min)
H
2
O
2

(
u
M
/
s
e
c
)
7W 16W
24W 30W
36W
Fig. 6. The hydrogen peroxide production rate with Ar as carrier gas.
0
0.5
1
1.5
2
2.5
3
3.5
4
0 2 4 6 8 10 12 14 16 18 20 22
Qw (mL/min)
H
2
O
2

(
g
/
k
W
h
)
7W
16W
24W
30W
36W
Fig. 7. Energy yields for hydrogen peroxide formation.
0
2
4
6
8
10
12
14
16
18
20
22
24
26
0 2 4 6 8 10 12 14 16 18 20 22
Qw (mL/min)
C
o
n
d
.

(
u
S
/
c
m
)
1358
The glidarc reactor with spray produces less hydrogen
peroxide at the outlet of the reactor for 7W the power of the
discharge but the efficiency is almost 2.5 times higher. The
molecular hydrogen production rate measured at the outlet of
the reactor in the gas phase is given in Fig. 10.
Similar with hydrogen peroxide formation the highest
production rate of H
2
was found with the highest power of the
discharge 36 W and for the same 10 mL/min water flow rate.


Similar with the hydrogen peroxide formation in water
droplets the molecular hydrogen measured in the gas flow at
the outlet of the reactor emphasize an optimum point of
production rate for a power of 36W and 10mL/min water
flow rate. The water flow rates lower than 10mL/min lead to
a lower density of water particles in plasma zone therefore
less hydrogen and hydrogen peroxide have been measured. In
addition, in the case of molecular hydrogen, the pulsed
injection of the water into the reactor, at a low frequency of
the pomp pulses, could lead to a dilution of the hydrogen in
gas phase.
For the values of the water flow rate higher than 10
mL/min, the high density of water droplets injected into the
plasma zone of the reactor lead to a instability of the
discharge, the electrons energy decrease, the effect being a
reduced amount of molecular species formed in water and
gas phase.
The efficiency energy of molecular hydrogen production in
gas phase function of water flow rates and for different values
of the power of the discharge is presented in Fig.11.
Contrary with the hydrogen peroxide formation in water,
the molecular hydrogen found at the outlet of the reactor
emphasize a maximum point for the same water flow rate of
10 mL/min. but for 30W the power of the discharge.



The hydrogen efficiency energy increase with the water
particles density injected in plasma zone up to 0.9 g/kWh for
10 mL/min. Similar with the hydrogen peroxide production,
the energy efficiency started to decrease for water flow rates
higher than 10 mL/min due perhaps to the thermal quenching
of the plasma by water droplets.
The highest energy efficiency for hydrogen production
obtained for 30W power of the discharge could be explained
due to the fact that the molecular hydrogen is formed by the
interaction of high energy species of the plasma and the water
droplets and then released into the gas phase. Contrary, the
hydrogen peroxide is formed as a result of OH radicals
recombination and fast quenched in water droplets. For
hydrogen formation it seems that is necessary a higher energy
to be injected into the plasma in order to be release as
molecular hydrogen in gas phase. The thermodynamic limit
for the overall reaction (4) has been reported to be 325 g/kWh
and kinetic limits with absolute quenching and ideal
quenching are 65 g/kWh and 107 g/kWh, respectively [16].
For hydrogen peroxide the thermodynamic limit is 400
g/kWh and the kinetic limit (ideal quenching) is 180 g/kWh
[16, 17]. Improvements in the efficiency of hydrogen
formation of the water spray reactor may be possible through
improving the reactor/nozzle design in order to induce a
faster quenching of the plasma [3] and smaller water droplets
in order to avoid the OH radical recombination.
The hydrogen energy yields are given in Fig.11 based upon
the power measured in the discharge in the secondary side of
the high voltage transformer.
In order to emphasize the optimum point for hydrogen
efficiency production the energy yield for H
2
function of the
power of the discharge has been plotted in Fig.12.



0
0.2
0.4
0.6
0.8
1
1.2
0 2 4 6 8 10 12 14 16 18 20 22
Qw (mL/min)
H
2

e
n
e
r
g
y

e
f
f
i
c
i
e
n
c
y

(
g
/
k
W
h
)
7W
16W
24W
30W
36W
Fig. 11. Energy yields for hydrogen formation
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
0 2 4 6 8 10 12 14 16 18 20 22
Qw (mL/min)
H
2

(
u
M
/
s
e
c
)
7W
16W
24W
30W
36W
Fig. 10. Hydrogen production rate for different water flow rates and powers
of the discharge.
1359



Fig. 12 emphasize the presence of a point of optimum for
hydrogen efficiency production for 10 mL/min water flow
rate and 30W the power of the discharge. For the power of 30
W the hydrogen efficiency production do not differ
significantly for the water flow rates in the range of 6-20
mL/min.
All the measurements for hydrogen and hydrogen peroxide
have been performed after 1 pass of the water spray through
the reactor, which correspond to a time of residence of the
water droplets in plasma zone of 10 ms.

IV. CONCLUSIONS
Molecular Hydrogen and hydrogen peroxide production in
a AC gliding arc reactor with water spray significantly
depends on the water flow rate injected into the plasma and
on the power of the discharge. The hydrogen peroxide
formation rate in water emphasizes an optimum point for 10
mL/min water flow rate and a discharge power of 36W. The
glidarc reactor with spray produces less hydrogen peroxide at
the outlet of the reactor for 7W the power of the discharge but
the efficiency is almost 2.5 times higher. A scale up of the
reactor is feasible using more modules of 7W power reactors.
Similar with hydrogen peroxide production rate, the
amount of molecular hydrogen formed from water has a
optimum point, the highest quantity of H2 have been obtained
for 10 mL/min water flow rate and a discharge power of
36W. From the point of view of energy efficiency of
hydrogen production the optimum have been found to be for a
power of 30W.
Improvements in the efficiency of hydrogen formation of
the water spray reactor may be possible through improving
the reactor/nozzle design in order to induce a faster
quenching of the plasma [3] and smaller water droplets in
order to avoid the OH radical recombination (H
2
O
2

formation in water droplets).
The size of water particles formed in the gas flow along
with the geometry of the nozzle and the reactor (open/confine
reactors) also play important roles in the molecular species
produced in gas and liquid phase.

ACKNOWLEDGMENT
This work was financially supported by the Romanian
Grant PCE-Idei 331/2009.

REFERENCES
[1] A. Czernichowski, Gliding arc applications to engineering and
environment control, Pure Appl. Chem., vol. 66, no. 6, pp. 13011310,
1994.
[2] B. Benstaali, D. Moussa, A. Addou, and J. L. Brisset, Plasma treatment
of aqueous solutes: Some chemical properties of a gliding arc in humid
air, Eur. Phys. J. Appl. Phys, vol. 4, no. 2, pp. 171179, 1998.
[3] A. Fridman and L. A. Kennedy, Plasma Physics and Engineering.
NewYork: Taylor & Francis, 2004.
[4] S. Mededovic and B. R. Locke, Primary chemical reactions in pulsed
electrical discharge channels in water, J. Phys. D, Appl. Phys., vol. 40,
no. 24, pp. 77347746, Dec. 2007.
[5] F. Moras, D. Sguin, B. Benstaali, A. Addou, B. G. Chron, E. Hnatiuc,
J-L. Brisset, "Interaction between a Non-Thermal Oxygenated Plasma
and Aqueous Solutions", Proceedings of International Conference
HAKONE, 2000
[6] Ji-Jun Zou, Yue-Ping Zhang , Chang-Jun Liu, Hydrogen production from
dimethyl ether using corona discharge plasma, Journal of Power
Oxygenated Plasma
[7] A. Czernichowski, Gliding arc applications to engineering and
environment control, Pure Appl. Chem., vol. 66, no. 6, pp. 13011310,
(1994).
[8] R. Burlica, M. Kirkpatrick, and B. R. Locke, The formation of reactive
species in gliding arc discharges with liquid water, J. Electrostat., vol.
64,no. 1, (2006).
[9] R. Burlica and B. R. Locke, Pulsed plasma gliding-arc discharges with
water spray, IEEE Trans. Ind. Appl., vol. 44, no. 2,(2008).
[10] M. Kirkpatrick and B. R. Locke, Hydrogen, oxygen, and hydrogen
peroxide formation in electrohydraulic discharge, Ind. Eng. Chem. Res.,
347 vol. 44, no. 3, (2005).
[11] T. J. Dolan, Electron and ion collisions with water vapor, J. Phys.D,
Appl. Phys., vol. 26, no. 1, (1993).
[12] M. Anbar and H. Taube, Interaction of nitrous acid with hydrogen
peroxide and with water, J. Amer. Chem. Soc., vol. 76, no. 24, (1954).
[13] D. Porter, M. D. Poplin, F. Holzer, W. C. Finney, and B. R. Locke,
Formation of Hydrogen Peroxide, Hydrogen, and Oxygen in Gliding Arc
Electrical Discharge Reactors With Water Spray, IEEE Transaction on
Industry Applications, Vol. 45, No. 2. (2009).
[14] S. Futamura, A. Zhang, T. Yamamoto, The dependence of non-thermal
plasma behavior of VOCs on their chemical structures, J. Electrostatics,
vol. 42, (1997).
[15] B. R. Locke, S. Mededovic Thagard, Analysis of Chemical Reactions in
Gliding-Arc Reactors With Water Spray Into Flowing Oxygen, IEEE
Transactions on Plasma Science, Vol 37-1, (2009).
[16] A. Fridman, Plasma Chemistry, Cambridge University Press, (2008).
[17] H. Kabashima, H. Einaga, and S. Futamura, Hydrogen generation from
water, methane, and methanol with nonthermal plasma, IEEE
Transactions on Industry Applications, vol.39, no 2, 2003


0
0.2
0.4
0.6
0.8
1
1.2
0 5 10 15 20 25 30 35 40
P (W)
e
n
e
r
g
y

e
f
f
i
c
i
e
n
c
y

H
2

(
g
/
k
W
h
)
Qw=2 mL/min
Qw=6 mL/min
Qw=10 mL/min
Qw=15 mL/min
Qw=20 mL/min
Fig. 12. The energy efficiency yield for H2 formation function of the
power for different water flow rates.
1360