Journal of The Electrochemical Society, 156 ͑4͒ E75-E80 ͑2009͒
E75
0013-4651/2009/156͑4͒/E75/6/$23.00 © The Electrochemical Society
Electrochemical Generation of Ozone in a Membrane Electrode
Assembly Cell with Convective Flow
a
Kwong-Yu Chan,
and Xiao-Yan Lib
aDepartment of Chemistry, and bDepartment of Civil Engineering, The University of Hong Kong,
Hong Kong SAR, China
Highly efficient electrochemical generation of ozone on doped tin dioxide anodes was reported recently. Here, we report the scale
up of such ozone generation on a membrane electrode assembly ͑MEA͒ made with 8 ϫ 13 cm–doped tin dioxide anode. The
effects of water flow rate, operating voltage, and current on dissolved ozone concentration, ozone production, current efficiency,
and energy efficiency are reported. Ozone production and current efficiency increased with water flow. Operating with a single
MEA, 218 mg/h of dissolved ozone was produced at an applied current of 6 A ͑current density = 57.6 mA/cm2͒. With four MEAs
operated in a stack, the dissolved ozone production increased to 1.1 g/h at a total current of 20.6 A ͑current density
= 49.5 mA/cm2͒ and individual cell voltage of 3.3 V. For the multiple MEA operation, the highest current efficiency was 21.7%
based alone on dissolved ozone generation. The lowest energy consumption achieved was 42 kWh/kg ͑O3͒ at 643 mg ozone per
hour at current of 10.1 A ͑current density = 24.3 mA/cm2͒ and water flow of 5.4 L/min ͑linear velocity = 7.03 cm/s͒.
© 2009 The Electrochemical Society. ͓DOI: 10.1149/1.3072686͔ All rights reserved.
Manuscript submitted August 8, 2008; revised manuscript received December 7, 2008. Published January 30, 2009.
materials with an effective anode area of 1.5 ϫ 5 cm.13 High con-
centration of ozone and high current efficiency were also achieved
with doped tin dioxide electrode recently.14,15 Acid electrolytes are
usually adopted to investigate performance of electrochemical ozone
production. For corrosion and safety consideration, it will be desir-
able to have an acid-free process in practical ozone generation, par-
ticularly for domestic household applications. Furthermore, hydro-
gen generated at the cathode during electrolysis creates a safety
concern. These problems can be circumvented by the use of a water-
filled polymer electrolyte membrane ͑PEM͒ cell fitted with an air
cathode, as reported recently.16 The ozone concentrations reported
with doped tin oxide anodes are much higher than the CCD process.
Our previous work on a membrane electrode assembly ͑MEA͒
using pure water was on a 4 ϫ 6 cm–doped tin oxide anode without
convection.16 In studies of electrochemical generation of ozone on
PbO2 anode, efficiency increased with flow of electrolyte using ei-
gate convection effects and the scale-up effect in a design using
large 8 ϫ 13 cm MEAs. An ozone production rate at 1.1 g/h was
achieved with a four-MEA stack running with water, demonstrating
the practicality of the doped oxide anode technology for small ozone
generation units. In the following sections, the experimental setup,
effects of flow rate and current ͑voltage͒ on ozone concentration,
ozone production, and current efficiency are discussed.
Ozone has broad applications in disinfection, bleaching, water
treatment, wastewater treatment, environmental cleanup, and chemi-
cal synthesis. The oxidation potential of ozone is 2.07 V, higher
than that of most common oxidizing chemicals, such as chlorine.1
As a powerful oxidant, ozone can rapidly remove a large variety of
organic pollutants and persistent pathogens. More importantly, there
is no harmful by-product or no residual effects as ozone decomposes
into oxygen with time. However, this property makes ozone difficult
to be stored. Ozone must be generated and applied in situ. Ozone is
conventionally synthesized through high voltage cold corona dis-
charge ͑CCD͒, producing a 2–3% concentration of ozone2 from air.
The overall energy consumption of the CCD process is high because
cooling and drying of air are needed in addition to high voltage. The
CCD process can produce nitrogen oxide ͑NOx͒ as a by-product
unless nitrogen is removed by additional energy. The energy re-
quired to purify air to oxygen is at least two times that of making
ozone from oxygen.3 An alternative generation method is via UV
radiation at 185 nm. However, the generated ozone concentration is
lower than that of CCD production and more electrical energy is
consumed. UV lamps also need periodic replacement with additional
expenses. Electrochemical ozone generation in liquid phase is an
attractive alternative. It does not require high voltage to ionize gas-
eous oxygen molecules and a few volts of dc source are sufficient
for ozone generation in an electrolyte. Water electrolysis in a low-
voltage condition eliminates any possibility of NOx formation. If
dissolved ozone is the desired product, then the water electrolysis
route is more direct and effective. The decay and loss during transfer
from gaseous ozone to dissolved ozone are avoided.
Experimental
Preparation of electrode and MEA.— The components and op-
eration of the MEA are shown schematically in Fig. 1a. A mesh
anode coated with doped SnO2 was hot pressed to a Nafion PEM
together with an air diffusion cathode.
The electrochemical synthesis of ozone was investigated in the
early 1980s and was reviewed recently.4,5 The effectiveness of elec-
trolytic ozone generation depends critically on electrode materials
and the reactor design. A number of anode materials have been
examined previously for the electrochemical ozone generation, but
dioxide has been the most investigated anode material with consid-
erable effectiveness, but is environmentally prohibitive for a general
usage. A very high efficiency of 47% on electrochemical ozone pro-
duction has been reported by using boron-doped diamond electrode
The anode was prepared by the repeated procedure of coating
titanium mesh ͑Dexmet Corporation, 5 Ti 5-031͒, 0.127 mm thick
with a nominal aperture of 0.787 mm, wire strand width of
0.127 mm was used as the substrate. The mesh was then rinsed with
acetone, treated in 10% boiling oxalic acid for 1 h, washed with
deionized water, and dried. Different from the previous
substrate in the first firing, electrodeposition of the antimony-tin
layer was applied first. For electrodeposition, the pretreated Ti mesh
was placed as the cathode in 500 mL of alcohol solution containing
43.75 g SnCl4·5H2O ͑98%, ABCR͒, 1.395 g SbCl3 ͑BDH, 99.5%͒.
The deposition procedure was 1 A ͑current density = 4.8 mA/cm2͒
for 1 min followed by 0.5 A ͑current density = 2.4 mA/cm2͒ for
20 min in a three-chamber cell with two counter electrodes made of
*
**
Electrochemical Society Student Member.
Electrochemical Society Active Member.
c Present address: School of Environmental Science and Engineering, Huazhong
University of Science and Technology, Wuhan 430074, China.
d Present address: Department of Environmental Science and Engineering, Faculty
of Energy and Power, Xi’an Jiaotong University, Xi’an, 710049, China.
z E-mail: hrsccky@hku.hk