DMFC with hydrogen peroxide cogeneration

Article abstract of DOI:10.1149/1.3461161

The process of H₂O₂ generation in the direct methanol fuel cell (DMFC) composed of the Black Pearls 2000 gas-diffusion cathode, washed by sodium chloride brine or sea water for O₂ electroreduction and the half-membrane-electrode assembly with the Nafion 115 cation-exchange membrane, hot-pressed on one side with the carbonaceous gas-diffusion anode impregnated by Pt/Ru catalyst for methanol oxidation, has been developed. The effect of NaCl concentration in the catholyte, methanol and the KOH concentration in the anolyte, electrolyte temperature, and circulation flow rate on the generated cell voltage, current, and H₂O₂ current efficiency have been studied in 2 h trials. An 87-95% current efficiency for H₂O² generation with a 5.286-9.44 mg h ^-1 cm^-2 rate and at 15 mA cm^-2 current density was achieved in the DMFC with 1 M CH₃OH+7 M KOH in the anolyte and 30 g L^-1 NaCl in the catholyte.

Full text of DOI:10.1149/1.3461161

E140
Journal of The Electrochemical Society, 157 ͑9͒ E140-E147 ͑2010͒
0013-4651/2010/157͑9͒/E140/8/$28.00 © The Electrochemical Society
DMFC with Hydrogen Peroxide Cogeneration
Giorgi Agladze,a,b,z Paata Nikoleishvili,a Gigla Tsurtsumia,b Valentina Kveselava,b
Georgi Gorelishvili,a and Roman Latsuzbaiab
aDepartment of Electrochemical Engineering, Georgian Technical University, 0175 Tbilisi, Georgia
bLaboratory of Electrochemistry and Electrometallurgy, R. Agladze Institute of Inorganic Chemistry and
Electrochemistry, 0186 Tbilisi, Georgia
The process of H2O2 generation in the direct methanol fuel cell ͑DMFC͒ composed of the “Black Pearls 2000” gas-diffusion
cathode, washed by sodium chloride brine or sea water for O2 electroreduction and the half-membrane-electrode assembly with the
Nafion 115 cation-exchange membrane, hot-pressed on one side with the carbonaceous gas-diffusion anode impregnated by Pt/Ru
catalyst for methanol oxidation, has been developed. The effect of NaCl concentration in the catholyte, methanol and the KOH
concentration in the anolyte, electrolyte temperature, and circulation flow rate on the generated cell voltage, current, and H2O2
current efficiency have been studied in
2 h trials. An 87–95% current efficiency for H2O2 generation with a
5.286–9.44 mg h−1 cm−2 rate and at 15 mA cm−2 current density was achieved in the DMFC with 1 M CH3OH + 7 M KOH in
the anolyte and 30 g L−1 NaCl in the catholyte.
© 2010 The Electrochemical Society. ͓DOI: 10.1149/1.3461161͔ All rights reserved.
Manuscript submitted March 17, 2010; revised manuscript received June 16, 2010. Published July 26, 2010. This was Paper 317
presented at the Vienna, Austria, Meeting of the Society, October 4–9, 2009.
H2 → 2H + 2e Eoa = 0.0 V
͓5͔
͓6͔
+
Hydrogen peroxide is widely used in medicine, pulp and paper
industry, detergents, and water treatment as the safest, most effec-
tive, and environmentally friendly sanitizer, disinfectant, and oxidiz-
ing and bleaching reagent.
The industrial chemical method of hydrogen peroxide produc-
tion, based on the cyclic reduction of oxygen by hydrogen using
anthraquinone as a catalyst, needs feedback from huge centralized
hydrogen production plants that are located only at several sites.
That results in costly and dangerous transportation of concentrated
hydrogen peroxide solutions to remote sites where diluted ͑0.1–3%
H2O2͒ solutions are usually utilized for the above-mentioned pur-
poses.
The alternative is based on a two-electron electrochemical reduc-
tion of oxygen ͑air͒ at carbonaceous cathodes.1,2 Due to the poor
solubility of oxygen in aqueous solutions, this reaction proceeds at
low current densities on plate graphite cathodes. A variety of elec-
trochemical reactors with expanded surface area ͑gas diffusion,
packed bed, and carbon felt͒ electrodes fed from the outer source of
a dc was used in comprehensive studies of H2O2 electrosynthesis.1-9
The cost of electric power is the main component of running
expenses for these processes. This can be avoided using fuel cell
͑FC͒ systems.
¯
−
¯
H2 + 2 OH → 2H2O + 2e ͑alkaline solutions͒
Eoa = − 0.828 V
or
−
¯
CH3OH + 6 OH → CO2 + 5H2O + 6e ͑alkaline solutions͒
Eoa = − 0.81 V
͓7͔
CH3OH + H2O → CO2 + 6H+ + 6e−
͑neutral or acid solutions͒
Eoa = − 0.081 V
͓8͔
͓9͔
Total
For the four-electron O2 reduction reaction
2H2 + O2 → 2H2O E = Eoa − Eco = 1.229 V
2CH3OH + 3 O2 → 2CO2 + 4H2O E = Eoa − Eco = 1.211 V
͓10͔
The thermodynamic potential gradient, formed by oxygen reduc-
tion and H2 or CH3OH oxidation reactions, is sufficient for the
operation of the FC system with free electron transfer in the external
circuit and for the formation of H2O or HO−2 ions in the cathode
chamber.1-8,10
for the two-electron O2 reduction
H2 + O2 + OH− → HO2− + H2O E = Eao − Eoc = 0.763 V
͓11͔
CH3OH + 3 O2 + 3 OH− → CO2 + 3HO2− + 2H2O
Cathode:
−
−
E = Eoa − Eoc = 0.745 V ͓12͔
¯
O2 + H2O + 2e → OH + HO2 ͑alkaline solutions͒
Eoc = − 0.065 V
Conventional proton exchange membrane FCs are based on the
four-electron reduction of O2 to H2O at the cathode ͑Reactions 3, 4,
9, and 10͒. Though alternative two-electron reduction combined
with H2 ͑Reaction 11͒ or methanol oxidation usually generates a
relatively lower potential gradient, the possibility of autonomous
work with cogeneration of hydrogen peroxide is the most important
advantage of the latter system especially for remote and specific
consumers of oxidizing agents or for local wastewater treatment and
bleaching operational systems.1,2,10,11
͓1͔
or
O2 + 2H + 2e → H2O2 ͑acid solutions͒ Eoc = 0.69 V ͓2͔
+
¯
O2 + 4H + 4e → 2H2O ͑acid solutions͒ Eoc = 1.21 V ͓3͔
+
¯
or
O2 + 2H2O + 4e → 4 OH ͑alkaline solutions͒ Eoc = 0.401 V
͓4͔
−
Hydrogen was used as a fuel for cogeneration of OH−2 in a small-
scale undivided alkaline fuel cell ͑AFC͒, consisting of a H2 diffusion
anode, an O2 diffusion cathode, and 1–6 mol dm−3 KOH solution
as a flowing electrolyte.12 Close to 100% current efficiency ͑CE͒
was achieved when a fresh KOH solution was injected into the AFC
in continuous flow-through operation mode, but the final OH−2 con-
centration in the flowing electrolyte solution was very low. The CE
¯
Anode:
z E-mail: agladzegogi@hotmail.com
Downloaded on 2015-12-23 to IP 131.111.164.128 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 157 ͑9͒ E140-E147 ͑2010͒
E141
Figure 1. Principal scheme of the DMFC
for hydrogen peroxide cogeneration.
using the standard permanganate titration method. OH− ion concen-
tration was determined by titration with a standardized 1 M HCl
solution.
dropped significantly in batch recirculation operation due to OH2−
decomposition at the anode and in the solution bulk.
The FC, divided by the membrane electrode assembly ͑MEA͒
consisting of a Nafion 117 cation-exchange membrane ͑CEM͒ with
Pt deposited on one side, where H2 is oxidized, and graphite or Au
mesh of 1.3 cm2 surface area, adhered on the reverse face of the
CEM, where O2 was reduced to H2O2, was described in Ref. 11.
The product was accumulated with 40.1% CE under short-circuit
conditions in the aqueous solution of HCI or H2SO4 at
12.7 mA cm−2 current density.
The difficulties caused by the generation and storage of hydro-
gen, which is explosive in oxygen atmosphere, do not occur with
direct alcohol FCs. H2O2 cogeneration in the direct methanol fuel
cell ͑DMFC͒ with 2–6 mole dm−3 KOH solution, used as a
catholyte, has been developed in our previous work.13,14 Taking into
account that such FCs are mainly intended for the treatment of con-
taminated natural or wastewaters in the cathode chamber, the utili-
zation of concentrated alkaline catholyte constricts their application
area.
Figure 1 illustrates the main components of the used FC, which
consists of ͑1͒ gas-diffusion electrode ͑GDE͒ and ͑2͒ CEM with ͑3͒
anode, hot-pressed at one side of the membrane. Each anode and
cathode had 9 cm2 apparent surface areas. The ͑4͒ 10 ϫ 10 cm
cathode frame had the ͑5͒ dished 4 ϫ 4 ϫ 0.2 cm size gas cham-
ber. Air was supplied as an oxygen source to the reverse side of the
O2-diffusion cathode into the gas chamber and ͑6͒ steel mesh, which
served as a current feeder. The ͑7͒ spacer, pressed between ͑8͒ rub-
ber gaskets, was used for uniform distribution of electrolyte at the
cathode surface. The anolyte, supplied through the inlet hole to ͑9͒
anode frame, washed the gas-diffusion anode, which was connected
to the ͑10͒ meshed titanium current feeder, coated with mixed
TiO2–RuO2 oxide layers.
The experimental setup is shown in Fig. 2. Anolyte and catholyte
This paper reports a study of the influence of anolyte and
catholyte composition and concentration, circulation rate and tem-
perature on the cell voltage, current density, and HO−2 generation CE
in the DMFC with relatively dilute chloride solutions and sea water
used as the catholyte, which can be considered as a low cost and
acceptable supporting electrolyte for many environmental and in-
dustrial applications.15
Experimental
Test solutions were prepared using distilled water and chemical-
grade reagents: methanol, potassium, and sodium hydroxide. The
used “sea salt for medical care” was supplied by Aversi Ltd. The
“A-12CUST, Custom Specialty ELAT for H2O2 generation” gas-
diffusion cathode with “Black Pearls 2000” carbon black active sur-
face and an improved hydrophobic back side on “Best Effort” basis,
used as a cathode, and “HES12D-W Custom HALF-E-TEK series
12D-W MEA ͑three-layer͒ custom half MEA with type 4 ͑Pt:Ru͒
anode, gas-diffusion layer, based on woven web,” used as an anode,
were supplied by BASF Fuel Cell, Inc. The anode was hot-pressed
on one side of the Nafion 115 CEM. Measurements of current, volt-
age, and pH were carried out using an M2015 ammeter, an M106
high impedance voltmeter, and an HI 8424 microcomputer pH meter
accordingly. H2O2 concentration in the electrolyte was determined
Figure 2. Experimental setup.
Downloaded on 2015-12-23 to IP 131.111.164.128 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

E142
Journal of The Electrochemical Society, 157 ͑9͒ E140-E147 ͑2010͒
Figure 4. Variation in generated current with time in trials with different
concentrations of NaCl in the catholyte ͑᭡: 1 g L−1, ᮀ: 10 g L−1, and ᭝:
30 g L−1͒. Anolyte: Aqueous solution with 2 M KOH + 4 M CH3OH.
The effect of NaCl concentration variation in the 1–30 g L−1
range on HO−2 CE is shown in Fig. 6. Although at the initial stage of
operation, 30–60 min decay of CE was more significantly revealed
in tests with 1 g L−1 NaCl solution, after 120 min this effect was
not clearly recognized. In general, we observed a stable performance
Figure 3. Variation in generated voltage and power in trials with different
concentrations of NaCl ͑᭡: 1 g L−1, ᮀ: 10 g L−1, and ᭝: 30 g L−1͒ in the
catholyte. Anolyte: 2 M KOH + 4 M CH3OH solution.
solutions were circulated through the anode and cathode compart-
ments of ͑1͒ DMFC to corresponding ͑2 and 3͒ tanks by means of a
͑4͒ pump. Pressure of air, fed from the ͑7͒ air pump to the gas
chamber was controlled by a ͑5͒ manometer. The external load was
changed using a box of resistors R and the FC parameters were
controlled by an ammeter and a high impedance voltmeter.
Results and Discussion
Figure 5. NaOH concentration growth in the catholyte ͑᭿͒ and decay in the
anolyte ͑ᮀ͒ with time in trials with initial concentration of NaCl in the
catholyte 30 g L−1. Anolyte: 2 M KOH + 4 M CH3OH solution.
Effect of sodium chloride concentration in the catholyte on HO2Ϫ
electrogeneration rate and on voltammetric characteristics of the
DMFC.— The results of the tests, carried out with batch recircula-
tion of the catholyte at 390 mL min−1 flow rate with different ͑1,
10, and 30 g L−1͒ contents of NaCl at 20°C and of the anolyte with
2 M KOH + 4 M CH3OH solution at 500 mL min−1, are shown in
Fig. 3a and b and 4. Each catholyte and anolyte volume was 100
mL. Air was supplied to the cathode chamber at 355 Pa.
Increased cell voltage and generated current were achieved
through the growth of NaCl concentration admittedly due to the rise
of conductivity of the catholyte solution. Accumulation of alkali in
the catholyte due to Reaction 1 also leads to growth of conductivity,
which
is
significant
in
relatively
diluted
solutions
͑1–10 g L−1 NaCl͒, whereas at higher concentrations this effect is
not sizable ͑Fig. 4͒.
Bonding of K+ or Na+ ions, penetrating through the CEM from
the anolyte to the catholyte, with OH− and OH2− anions formed at the
cathode,1 leads to a growth of alkalinity and, accordingly, of internal
conductivity during the FC operation ͑Fig. 5͒; alkalinity in the
catholyte was calculated on NaOH. In 30 g L−1 NaCl, the effect of
additional alkalinity is not significant as the primordial conductivity
of catholyte solution is high enough.
Figure 6. Variation in H2O2 CE with time in trials with different concentra-
tions of NaCl in the catholyte ͑᭡: 1 g L−1, ᮀ: 10 g L−1, and ᭝:
30 g L−1͒, recirculated at 390 mL min−1 flow rate. Anolyte: 2 M KOH
+ 4 M CH3OH solution, recirculated at 500 mL min−1 flow rate.
Downloaded on 2015-12-23 to IP 131.111.164.128 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 157 ͑9͒ E140-E147 ͑2010͒
E143
Figure 7. Variation in generated current with time in trials with different
concentrations of CH3OH aqueous solution with 5 M KOH in the anolyte.
Catholyte: 30 g L−1 NaCl solution.
of the DMFC with 100–93% H2O2 CE using brines with concentra-
tions of chlorides and conductivity close to those in lakes and sea
water.
Effect of methanol concentration.— The effect of variation in
methanol concentration on DMFC performance in tests where a 5 M
KOH solution was used as a supporting electrolyte in the anolyte is
shown in Fig. 7 and 8. The increase in methanol concentration from
0.5 to 0.1 M was reflected in the rise of the generated current from
7.5 to 10.8 mA cm−2, but the next growth of the methanol concen-
tration in the 1–4 M CH3OH range was not followed by a significant
change in the generated current. Moreover, judging from the
current–power ͑i-P͒ curves, even higher operational parameters
͑11 mA cm−2 short-circuit current and 13 mW cm−2 peak power͒
were achieved in 1 M methanol solution used as an anolyte ͑Fig. 8͒.
Other drawbacks of utilization of higher concentrations of
methanol could be the higher probability of CH3OH penetration into
the CEM to the catholyte, where it could be consumed in the H2O2
reduction reaction. The rate of methanol oxidation could be limited
by desorption of the CO2 from the catalytically active sites of the
anode, which is independent from methanol concentration in the 1–4
M CH3OH range.7
Figure 9. ͑Color online͒ Dependence of ͑a͒ cell voltage and ͑b͒ power gen-
erated by the FC at 20°C on variation in KOH concentration in the anolyte
with 4 M CH3OH aqueous solution. Catholyte: 30 g L−1 NaCl solution.
was 500 mL min−1, catholyte flow rate was 390 mL min−1, NaCl
concentration in the catholyte was 30 g L−1, temperature was
20°C, and air pressure in the gas chamber was 355 Pa.
The positive influence of increased KOH concentrations agrees
with the known effect of accelerated activity of catalysts of metha-
nol oxidation ͑Pt, Ru͒ at increased alkalinity.15,16 The highest values
of voltage and power generated by the DMFC had been achieved
when 7 M KOH was used.
Influence of anolyte and catholyte circulation rate on the FC
operation.— The following trials were carried out with variation in
the circulation rate of the anolyte in the 700–1470 mL min−1 range
and the catholyte in the 390–500 mL min−1 range. The temperature
of the 5 M KOH + 4 M CH3OH anolyte solution was 20°C,
whereas that of a 30 g L−1 NaCl catholyte solution was 10°C. Air
pressure in the gas chamber was monitored at 355 Pa.
Effect of KOH concentration in the anolyte.— Results of tests,
performed using a 4 M CH3OH anolyte with variation in KOH
concentration in the 0.1–8 M range, are shown in Fig. 9a and b.
Each anolyte and catholyte volume was 100 mL, anolyte flow rate
Improvement of the DMFC parameters was achieved with accel-
erating the anolyte circulation rate from 700 to 1000 mL min−1
.
The next rise of the anolyte circulation rate to 1470 mL min−1 had
a slight influence on the shape of the initial i-P curve ͑Fig. 10a and
b͒, which shows that maximal generated power has been achieved at
0.25 V cell voltage and 50 mA current ͑5.5 mA cm−2 current den-
sity͒, whereas the maximal generated current increased with time
significantly at a 1470 mL min−1 flow rate ͑Fig. 11͒.
The Reynolds number values calculated for the above-mentioned
circulation rates are 408, 583, and 857, correspondingly. This indi-
cates the laminar flow regime of the anolyte.
Generated power increased with the growth of the catholyte flow
rate from 390 to 450 and 500 mL min−1 ͑Fig. 12a and b and 13,
respectively͒, whereas at higher flow rates ͑700 mL min−1͒, the
flooding of the GDE and penetration of the solution to the gas cham-
ber has been observed.
Improvement of FC operation parameters at increased electrolyte
flow rates could be related to accelerated desorption of intermediate
products from the surface of the electrode.
Figure 8. Variation in generated power in trials with different concentrations
of CH3OH aqueous solution with 5 M KOH in the anolyte. Catholyte:
30 g L−1 NaCl solution.
Generation of HOϪ2 ions in sea water
eration in sea water were performed at 20°C in the cell with 150 mL
.— Tests with H2O2 gen-
Downloaded on 2015-12-23 to IP 131.111.164.128 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

E144
Journal of The Electrochemical Society, 157 ͑9͒ E140-E147 ͑2010͒
Figure 10. Dependence of ͑a͒ cell voltage and ͑b͒ power generated by the
FC at 0.1 ⍀ external resistance on current at different anolyte flow rates ͑ᮀ:
700 mL min−1, ᭝: 1000 mL min−1, and ᭡: 1470 mL min−1͒ and constant
catholyte flow rate ͑500 mL min−1͒.
Figure 12. Dependence of ͑a͒ cell voltage and ͑b͒ power generated by the
FC at 20°C on the current at different catholyte flow rates ͑᭡:
390 mL min−1, ᮀ: 450 mL min−1, and ᭝: 500 mL min−1͒ and constant
anolyte flow rate ͑500 mL min−1͒.
of 5 M KOH + 4 M CH3OH solution used as an anolyte and 150
mL of sea water as a catholyte. The catholyte flow rate was
500 mL min−1 and the anolyte flow rate was 1000 mL min−1. Air
was supplied to the gas chamber at 35 mm H2O pressure. Voltage
has decreased from 1.1 V at open-circuit conditions ͓open-circuit
voltage ͑OCV͔͒ to less than 0.2 V at 0.1 ⍀ outer resistance ͑Fig.
14͒.
The formation of the white scale of magnesium and calcium
hydroxides has been observed at the GDE surface due to the rise of
alkalinity during the H2O2 generation in sea water. Despite the pores
of the GDE being partially blocked by hydroxides that caused de-
creased H2O2 CE ͑Fig. 15͒, the FC continuously generated a stable
current ͑Fig. 16͒.
Influence of electrolyte temperature on DMFC parameters and
HOϪ2 CE.— The effect of temperature on DMFC performance,
H2O2 stability, and CE was studied in tests, carried out using 5 M
KOH + 4 M CH3OH solution as an anolyte and 30 g L−1 NaCl as
a catholyte. The anolyte temperature had been fixed by means of a
heater at 20, 40, 60, and 80°C levels in different trials, which influ-
Figure 11. Variation in generated current with time at different anolyte flow
rates ͑ᮀ: 700 mL min−1, ᭝: 1000 mL min−1, and ᭡: 1470 mL min−1͒ and
constant catholyte flow rate ͑500 mL min−1͒.
Figure 13. Variation in generated current with time in trials with different
catholyte flow rates ͑᭡: 390 mL min−1, ᮀ: 450 mL min−1, and ᭝:
500 mL min−1͒ and constant anolyte flow rate ͑500 mL min−1͒.
Downloaded on 2015-12-23 to IP 131.111.164.128 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 157 ͑9͒ E140-E147 ͑2010͒
E145
Figure 14. Dependence of cell voltage on current generated in the FC with
sea water catholyte.
enced accordingly 20, 28, 32, and 50°C temperature in the
catholyte. The anolyte flow rate was 1000 mL min−1, whereas the
catholyte flow rate was 500 mL min−1
.
Elevating of the anolyte temperature up to 60°C decreased the
internal resistance of the cell and increased reagent mass-transfer
rates that resulted in higher current density and generated power, but
HO−2 CE decreased over time from 85 to 75% during 2 h of opera-
tion at 50°C, whereas at 20°C, the corresponding values were 100
and 92°C ͑Fig. 17a and b and Fig. 18͒.
Figure 17. Variation in ͑a͒ generated voltage and ͑b͒ power in trials with
different temperatures of anolyte. Catholyte: 30 g L−1 NaCl and anolyte: 2
M KOH + 4 M CH3OH solution.
Application of CEM prevents access of HO−2 ions to the anode,
which means that the destruction of peroxide ions with the FC op-
eration time can be determined by electrochemical reactions at the
cathode or by chemical reactions in the solution bulk. The effect of
chemical destruction was studied separately in the next tests where
both anolyte and catholyte were heated to the same temperature ͑25
and 50°C͒. The outer circuit of the DMFC was disconnected in
these tests after H2O2 electrogeneration and concentration of H2O2
was monitored in the solution, which circulated in the cell without
current generation ͑Fig. 19͒.
The decomposition degree of H2O2, calculated as
Figure 15. Variation in H2O2 CE with time in the FC with sea water
catholyte.
Figure 16. Variation in generated current with time in the FC with sea water
catholyte.
Figure 18. Variation in H2O2 CE over time in trials with different tempera-
tures. Catholyte: 30 g L−1 NaCl and anolyte: 2 M KOH + 4 M CH3OH.
Downloaded on 2015-12-23 to IP 131.111.164.128 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

E146
Journal of The Electrochemical Society, 157 ͑9͒ E140-E147 ͑2010͒
O3 + H2O2 → 2 OHء + 3/2 O2
H2O2 + Fe2+ → Fe3+ + OHء + OH−
͓15͔
͓16͔
As to Fenton, electro-Fenton, photoelectro-Fenton, and solar
photoelectron-Fenton processes, based on the reaction of hydrogen
peroxide with ferrous salts ͑Reaction 16͒, which proceeds with
maximal rate at pH 3, the application of the proposed DMFC in
these processes can be effective in the so-called ex-cell operation
mode, maintaining subacid conditions for chemical Fenton reaction
in a separate vessel.2
Conclusions
Electrochemical characteristics of the DMFC with H2O2 cogen-
eration has been studied in trials with different ͑1–30 g L−1͒ con-
centrations of NaCl used as the catholyte and 0.1–8.0 M KOH
+ ͑0.5–4.0͒ M CH3OH as the anolyte. H2O2 CE decreased from
100% at the initial stage of operation to 95% after 120 min. Current
density increased from 7.5 to 10.8 mA cm−2 with a rise of CH3OH
concentration in the anolyte from 0.5 to 1.0 M but the further growth
of methanol concentration was not followed with significant changes
in the generated current. Moreover, it stimulated penetration of
methanol through the membrane into the catholyte and partial losses
of the generated hydrogen peroxide on oxidation of CH3OH. Varia-
tion in KOH concentration in the anolyte in the 0.5–8 M range
showed that the highest values of generated current were achieved in
tests with 7 M KOH.
Figure 19. Accumulation of HO−2 over time in the catholyte at different
temperatures of 25 ͑᭿͒ and 50°C ͑ᮀ͒ with and without FC operation.
minitial − m
␶
␤ =
ϫ 100%
minitial
where minitial is the initial weight of generated H2O2͑g͒ and m is the
␶
weight of H2O2͑g͒ in time ␶ ͑min͒. The decomposition degree in-
creased almost 50% with elevation of the solution temperature from
25 up to 50°C at ͑Fig. 20͒.
The rise of anolyte and catholyte circulation rates from 700 to
1000 mL min−1 and from 390 to 500 mL min−1, respectively, had
improved the voltammetric curve of the FC.
The influence of anolyte temperature variation has been studied
in the 20–80°C range, which influenced the maintenance of the
catholyte temperature in line with 20, 28, 32, and 50°C. Although
the catholyte temperature growth up to 50°C increased the current
density and the generated power ͑due to decreased resistance of the
solution͒, the hydrogen peroxide CE decreased by 25% in 2 h, in
contrast with only 8% decay at 20°C.
The OCV of the DMFC was 1.1 V in tests with the sea water
used as the catholyte and 5 M KOH + 4 M CH3OH solution as the
anolyte. The generated current was stable during the tests but H2O2
CE decreased from the initial 100 to 65% after 3 h admittedly due to
coatings of the cathode surface with magnesium and calcium hy-
droxides, which could be removed by acid treatment.
Adding the amount of decomposed H2O2 to the data of hydrogen
peroxide determined by analysis during the electrogeneration pro-
cess results in values corresponding to 100% CE. So the chemical
Reaction 13 in the solution bulk could be considered as the reason of
H2O2 CE decay in time.
H2O2 = H2O + 1/2 O2
͓13͔
The selection of the optimal temperature for operation with hydro-
gen peroxide cogeneration depends on the application area of the
FC. Ambient temperatures are preferable when the target is the ac-
cumulation of hydrogen peroxide, whereas elevated temperatures
could provide higher values of generated power and more effective
wastewater treatment. H2O2/UV and H2O2/O3 advanced oxidation
processes are used for the destruction of biorefractory organic sub-
stances. These processes are based on immediate catalytic conver-
sion of H2O2 generated in the catholyte to an even more effective
oxidant ͑E = 2.8 V͒, hydroxyl radical OHء by means of ozone or
16-22
Acknowledgments
UV irradiation due to Reactions 15 and 16
The designate project has been fulfilled by the support of Geor-
gia National Science Foundation ͑grant no. GNSF/ST08/7-483͒.
Any idea in this publication is possessed by the authors and may not
represent the opinion of “Georgia National Science Foundation.”
h␯
H2O2→2 OHء
͓14͔
Rafael Technical University of Georgia assisted in meeting the publica-
tion costs of this article.
References
1. G. R. Agladze, G. S. Tsurtsumia, B.-I. Jung, J.-S. Kim, and G. G. Gorelishvili, J.
Appl. Electrochem., 37, 375 ͑2007͒.
2. G. R. Agladze, G. S. Tsurtsumia, B.-I. Jung, J.-S. Kim, and G. G. Gorelishvili, J.
Appl. Electrochem., 37, 385 ͑2007͒.
3. G. R. Agladze, G. S. Tsurtsumia, B.-I. Jung, J.-S. Kim, and G. G. Gorelishvili, J.
Appl. Electrochem., 37, 985 ͑2007͒.
4. E. Yeager, P. Krouse, and K. V. Rao, Electrochim. Acta, 9, 1057 ͑1964͒.
5. I. Morcos and E. Yeager, Electrochim. Acta, 15, 953 ͑1970͒.
6. R. J. Taylor and A. A. Humffray, J. Electroanal. Chem., 64, 63 ͑1975͒.
7. H.-H. Yang and R. L. McCreery, J. Electrochem. Soc., 147, 3420 ͑2000͒.
8. M. Sudoh, T. Kodera, K. Sakai, K. Zhang, and K. Koide, J. Chem. Eng. Jpn., 19,
513 ͑1986͒.
9. F. Alcaide, E. Brillas, P. L. Cabot, and J. Casado, J. Electrochem. Soc., 145, 3444
͑1998͒.
10. A. Bard, R. Parsons, and J. Jordan, Standard Potentials in Aqueous Solutions,
Chaps. 3 and 4, Marcel Dekker, New York ͑1985͒.
11. K. Otsuka and I. Yamanaka, Electrochim. Acta, 35, 319 ͑1990͒.
12. E. Brillas, F. Alcaide, and P. L. Cabot, Electrochim. Acta, 48, 331 ͑2002͒.
13. G. Agladze and P. Nikoleishvili, Patent of Georgia no. 10349/01 ͑2007͒.
14. P. Nikoleishvili, G. Agladze, G. Tsurtsumia, and V. Kveselava, Proceedings of the
Figure 20. Variation in the H2O2 decomposition degree with time in the
catholyte at 25 ͑᭿͒ and 50°C ͑ᮀ͒ after disconnection of the outer circuit.
Downloaded on 2015-12-23 to IP 131.111.164.128 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 157 ͑9͒ E140-E147 ͑2010͒
E147
Georgian National Academy of Sciences, Chem. Series, 35, 285 ͑2009͒.
19. E. Brillas and J. Casado, Chemosphere, 47, 241 ͑2002͒.
15. G. Agladze and P. Nikoleishvili, PG AP no. 11282/01 ͑2009͒.
16. R. Parsons and T. VanderNoot, J. Electroanal. Chem., 257, 9 ͑1988͒.
17. A. V. Tripković, K. D. Popović, B. N. Grgur, B. Blizanac, P. N. Ross, and N. M.
Marcović, Electrochim. Acta, 47, 3707 ͑2002͒.
20. E. Brillas, R. Sauleda, and J. Casado, J. Electrochem. Soc., 145, 759 ͑1998͒.
21. I. Sirés, N. Oturan, M. A. Oturan, R. M. Rodriguez, J. A. Garrido, and E. Brillas,
Electrochim. Acta, 52, 5493 ͑2007͒.
22. C. Flox, P. L. Cabot, F. Centellas, J. A. Carrido, R. M. Rodriguez, C. Arias, and E.
18. Y. Sun and J. J. Pignatello, Environ. Sci. Technol., 27, 304 ͑1993͒.
Brillas, Appl. Catal., B, 75, 17 ͑2007͒.
Downloaded on 2015-12-23 to IP 131.111.164.128 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).