Neutral H2O2 synthesis by electrolysis of water and O2

Article abstract of DOI:10.1002/anie.200704431

(Chemical Equation Presented) Direct, safe, and continuous synthesis of neutral H₂O₂ solutions with concentrations of up to 8 wt% from O₂ and water has been accomplished by an improved electrolysis method involving a solid-polymer electrolyte (SPE). The formation and accumulation of neutral H₂O₂ were strongly enhanced by exposing the cathode to a stream of O₂.

Full text of DOI:10.1002/anie.200704431

Communications
DOI: 10.1002/anie.200704431
Hydrogen Peroxide Production
Neutral H2O2 Synthesis by Electrolysis of Water and O2
Ichiro Yamanaka* and Toru Murayama
Growth in demand for hydrogen peroxide is expected due to
its wide utilization as a disinfectant and as an oxidant for
chemical processes, because it generates only water as
byproduct. However, the cost of H2O2 production by multi-
step anthraquinone-based processes and transport restric-
tions[1] are major factors in meeting this demand.[2] Therefore,
an on-site production method for H2O2, for example, direct
synthesis of H2O2 from H2 and O2 with Pd and Au/Pd catalysts
in acid or methanol solutions have been studied.[3–8] However,
a gaseous mixture of H2 and O2 has the possibility of
exploding, so a safer procedure is essential.
We previously reported an H2/O2 fuel-cell system for
direct formation of H2O2.[9,10] The fuel cell can be safely
Figure 1. Diagram of a) standard SPE method and b) exposure SPE
operated for H2O2 production because H2 and O2 are
separated by the electrolytic membrane. Palladium mem-
brane catalysts can be also used for safe synthesis of H2O2, but
the formation rate and concentration of H2O2 need to be
improved.[11,12] We recently reported an improved fuel-cell
system and new electrocatalysts for H2O2 synthesis. In the
case of 2n NaOH as electrolyte, 7 wt% H2O2 was synthesized
at the mixed-carbon cathode with 93% current efficiency
(CE).[13] With 1.2n H2SO4 as electrolyte, 3.5 wt% H2O2 was
synthesized at a cathode derivatized with Mn porphyrin with
45% CE.[14] The H2O2/NaOH solution is useful for pulp
bleaching, and the H2O2/H2SO4 solution can be used for
oxidation in organic synthesis. However, neutral aqueous
H2O2 solution without salts is the most useful and flexible
form.
If we use a solid polymer electrolyte (SPE), except for
soluble supporting electrolytes, electrolyte-free product sol-
utions can be obtained. This SPE electrolysis method has
been used for several kinds of electrochemical syntheses[15]
and for H2 generation (water decomposition). If we can find a
suitable electrocatalyst (cathode) and reaction conditions,
formation of neutral H2O2 can be expected.
(Exp-SPE) method. WE=working electrode, CE=counterelectrode,
RE=reference electrode.
Other possible reactions are formation of H2O and H2 at the
cathode. In addition, water is decomposed at the anode.
We selected a mixed-carbon cathode (2 cm2) prepared
from activated carbon (AC), vapor-grown carbon fiber
(VGCF), and teflon powder by the hot-press method.[6] This
[AC + VGCF] cathode is effective for formation of H2O2 by
the fuel-cell method with an H2SO4 electrolyte.[10] The anode
was prepared from 45 wt% Pt supported on carbon black (Pt/
CB), VGCF, and teflon powder. The electrodes were attached
on each side of a nafion-117 membrane (DuPont) under
5 MPa at 413 K. Yields of H2O2 were determined by chemical
titration with aqueous KMnO4/H2SO4 solutions. The CE of
H2O2 formation was calculated as a two-electron reaction
against the quantity of charge passed.
First, we filled the cathode and anode compartments
(30 mL) with deionized water and applied a cathode potential
of ꢀ0.5 V (vs. Ag/AgCl) for 2 h. A very low H2O2 yield of
0.8 mmol was detected with a low CE of 1.4%. The majority of
the electrolysis current was consumed in H2 evolution. The
electrochemical reduction rate of O2 was much lower than
that of H+ in H2 formation. The H2O2 yield was considerably
lower than that of the fuel-cell method using H2SO4 solu-
tions.[10]
The acidity of nafion-H is as high as that of H2SO4
solution; therefore, its pH conditions on the cathode may be
similar and should not be a major reason for formation of less
H2O2 in the SPE electrolysis method (Figure 1a). In the fuel-
cell system, the cathode was exposed to gaseous O2.[10,13,14]
The concentration of gaseous O2 (1 atm) of 41 mm at 298 K is
much higher than that in an aqueous electrolyte (ca. 1 mm).
On the basis of the above considerations, we exposed half
of the [AC + VGCF] cathode to an O2 stream by decreasing
the amount of deionized water to 15 mL, to give what we call
the exposure SPE (Exp-SPE) method (Figure 1b). We found
a dramatic improvement in H2O2 formation by applying the
An SPE electrolysis cell unit was prepared from a
cathode, an anode, and nafion-H membrane. This cell unit
was fixed in a two-compartment glass cell, as shown in
Figure 1. Deionized water was infused into both compart-
ments, and O2 and Ar were introduced into the cathode and
anode compartments, respectively. Given a suitable cathode,
reduction of O2 to H2O2 (O2 + 2H+ + 2eꢀ!H2O2) and
accumulation of H2O2 in the deionized water are expected.
[*] I. Yamanaka, T. Murayama
Department of Applied Chemistry
Graduate School of Science and Engineering
Tokyo Institute of Technology
Ookayama, Meguro-ku, Tokyo 152-8552 (Japan)
Fax : (+81)3-5734-2144
E-mail: yamanaka@apc.titech.ac.jp
1900
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 1900 –1902

Angewandte
Chemie
Table 1: Effect of initial volume of deionized water in the cathode
compartment on the neutral H2O2 synthesis by the Exp-SPE method.[a]
Run
1
2
3
4
initial V [mL]
final V [mL]
15
15
21.1
125
0.06
15.9
3.0
3.1
18.5
103
0.23
14.9
0.6
0.0
0.76
20.0
132
1.30
17.8
0.17
16.5
74.2
2.90
12.1
Id [mAcmꢀ2
]
r(H2O2) [mmolhꢀ1
C(H2O2) [wt%]
CE [%]
]
[a] T=281 K, O2/waterjAC+VGCF cathodejnafion-117jPt/CB anodej
Ar, cathode potential ꢀ0.3 V(Ag/AgCl), reaction time 2 h. Experimental
errors: Vꢁ0.1 mL and Cꢁ0.01 wt%.
run 3) after electrolysis. We determined that the increase in
volume corresponded to the amount of charge passed. In
detail, three molecules of water of hydration per proton
moved from the anode to the cathode. This suggests that
initial presence of water at the cathode is not essential for the
formation of neutral H2O2 solutions, because water of
hydration accumulates in the cathode bath during electrolysis.
Therefore, we performed Exp-SPE electrolysis without initial
water in the cathode bath, that is, with complete exposure of
the cathode (complete Exp-SPE method, Table 1, run 4). A
higher H2O2 concentration could be obtained (0.85m (2.9
wt%), 0.17 mL), but the formation rate and CE were slightly
lower than in runs 1–3.
The electrochemical reduction of O2 to H2O2 should
proceed at the three-phase boundary of nafion-H electrolyte
(liquid phase), AC surface (solid phase), and O2 (gas phase).
To increase the formation rate of H2O2, the area of the three-
phase boundary should be expanded. Therefore, we
attempted to modify the [AC + VGCF] cathode by coating
treatment with nafion solutions to increase the three-phase
boundary. Nafion solutions (5 wt% in 2-propanol and water,
Aldrich Co.) were painted on one side of the [AC + VGCF]
cathode, which dried during decompressing. The loading of
nafion was determined by the difference in weight before and
after coating. This cathode was denoted [AC + VGCF/
nafion]. Neutral H2O2 synthesis was conducted at the [AC +
VGCF/nafion] cathode by the complete Exp-SPE method
(Table 2, run 8). The current density and the formation rate of
H2O2 were significantly increased compared to the results
Figure 2. Effect of cathode potential on a) the rate of formation of
neutral H2O2, r(H2O2), and current density, Id, and b) current efficiency
CE in the standard SPE [Figure 1a: cathode soaking in water (O2)] and
Exp-SPE (Figure 1b: cathode exposed to O2/water) setups. Experimen-
tal conditions: jAC+VGCF cathodejnafion-117I Pt/CB anodejwa-
ter(Ar), T=281 K, reaction time 2 h.
Exp-SPE method. Figure 2 shows the effects of cathode
potential on the formation of H2O2. Current density and
formation rate of H2O2 increased with increasingly more
negative potential, and the latter showed a maximum rate at
ꢀ0.3 V. The H2O2 solutions in the cathode were found to be
neutral. The CE also showed its maximum of 18% at ꢀ0.3 V.
The formation rates of H2 were 54 mmolhꢀ1 (CE 6.8%) at
ꢀ0.3 Vand 254 mmolhꢀ1 (23.4%) at ꢀ0.4 V. In addition, lower
current density and yield of H2O2 were observed at 0 to
ꢀ0.3 V when the whole cathode was soaked in water. The
concentration of O2 in the gas phase (43 mm) is much higher
than that in water (1.8 mm) at 1 atm and 281 K. In addition,
the diffusion coefficient of gas (on the order of 100 m2 sꢀ1) is
much larger than that of liquid (on the order of 10ꢀ9 m2 sꢀ1).
These differences in physical properties suggest that the
steady-state concentration of O2 on the cathode in the Exp-
SPE method should be dramatically larger than that with the
conventional method and accelerate the formation rate.
However, the concentration of 15-mL H2O2 solutions of was
only 0.016m (0.056 wt%).
An attractive feature of the cathode is the accumulation of
more highly concentrated H2O2 solutions above 1 wt%
(0.29m), because the reduction rate of H2O2 to water is
usually very fast. We reduced the volume of deionized water
by decreasing the thickness of the cathode bath (Table 1,
runs 1–3). We note that current density and the formation rate
of H2O2 were almost constant among runs 1 (15 mL), 2
(3 mL), and 3 (0.6 mL), with no dependence on the initial
volume of deionized water. Therefore, concentrations of
H2O2 dramatically increased to 0.23 wt% (0.065m) and
1.30 wt% (0.38m) due to the reductions in volume. These
results suggest that successive reduction of H2O2 to water is
very slow at the [AC + VGCF] cathode.
Table 2: Effect of oxidative treatment of AC and nafion coating on neutral
H2O2 synthesis by the complete Exp-SPE method.[a]
Run
5
6
7
8
9
10
electrocatalyst
nafion [mg]
AC AC(air) AC(HNO3) AC AC(air) AC(HNO3)
–
–
–
5.5 5.1
24.6 46.8
191 352
20.9 20.2
217 343
5.99 6.98
6.5
Id [mAcmꢀ2
]
16.5 22.5
74.2 169
12.1 20.2
174 218
2.90 5.27
24.1
147
16.6
250
4.07
59.4
577
26.5
489
8.06
r(H2O2) [mmolhꢀ1
CE [%]
]
final V [mL]
C(H2O2) [wt%]
[a] AC(air): AC oxidized by air for 1 h at 537 K, AC(HNO3): AC oxidized by
3.8n HNO3 for 1 h under reflux. T=281 K, O2 jelectrocatalyst+VGCF
cathodejnafion-117jPt/CB anodejwater(Ar), cathode potential ꢀ0.3 V
(vs. Ag/AgCl), reaction time 2 h. Experimental errors: Vꢁ1 mL and
Cꢁ0.01 wt%.
Peculiarly, the solution volume increased from 3.00 to
3.14 mL (Table 1, run 2) and from 0.60 to 0.76 mL (Table 1,
Angew. Chem. Int. Ed. 2008, 47, 1900 –1902
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1901

Communications
with the [AC + VGCF] cathode (Table 2, run 5). The concen-
tration of H2O2 increased to 1.76m (5.99 wt%) with 20.9%
CE over 2 h.
without any toxic byproducts by using the complete Exp-SPE
method with [AC(HNO3) + VGCF/nafion] cathode. The
neutral H2O2 solution can be used as a disinfectant for
bathrooms, air conditioners, and running water, and as an
abstergent for semiconductors and an oxidant for the syn-
thesis of fine chemicals.
We have proposed a primitive reaction scheme in which
=
functional groups on AC surfaces (CO2H, OH, O, etc.)
catalyze the electrochemical reduction of O2, and VGCF was
used as a current collector (lead wire) for the [AC + VGCF]
cathode in the fuel-cell system.[10] In fact, the electrocatalytic
activity of the cathode for H2O2 formation disappeared when
functional groups on the AC surface were removed by
thermal treatment in He at 1073 K. We attempted to increase
electrocatalytic activity of the AC surface by oxidative
treatment.
The surface of AC was oxidized with 1.0–13.8m HNO3
solutions under reflux conditions for 1 h, and the effect of
oxidative treatment of AC on H2O2 formation at ꢀ0.3 V (Ag/
AgCl) in the complete Exp-SPE setup was studied. The AC
oxidized with 3.8m HNO3 (AC(HNO3)) was suitable for H2O2
synthesis. The formation rate, CE, and concentration of H2O2
significantly increased at the [AC(HNO3) + VGCF] cathode
(Table 2, run 7). Oxidation of the surface of AC in air at 473–
673 K was studied. The AC oxidized in air at 523 K (AC(air))
was suitable for formation of H2O2. The electrocatalytic
activity of the [AC(air) + VGCF] cathode was as good as that
of the [AC(HNO3) + VGCF] cathode (Table 2, run 7). Nafion
coating for neutral H2O2 formation was also effective on both
[AC(air) + VGCF/nafion] and [AC(HNO3) + VGCF/nafion]
cathodes (Table 2, runs 9 and 10, respectively). In particular,
the [AC(HNO3) + VGCF/nafion] cathode performed best
with respect to H2O2 formation: C(H2O2) = 2.4m (8.06 wt%)
with 26.5% CE over 2 h.
Received: September 26, 2007
Revised: December 6, 2007
Published online: January 25, 2008
Keywords: electrochemistry · green chemistry ·
.
hydrogen peroxide · reduction
[1] M. C. Campos-Martin, G. Blanco-Brieva, J. L. G. Fierro, Angew.
Chem. 2006, 118, 7116 – 7139; Angew. Chem. Int. Ed. 2006, 45,
6962 – 6984.
[2] H. Reidl, G. Pfleirender, US 2215883, 1940.
[3] P. C. Foller, R. T. Bombard, J. Appl. Electrochem. 1995, 25, 613 –
627.
[4] Y. Izumi, US 4009252, 1978.
[5] L. W. Gosset, US 4681751, 1988.
[6] G. Blanco-Brieva, E. Cano-Serrano, M. C. Campos-Martin,
J. L. G. Fierro, Chem. Commun. 2004, 1184 – 1185.
[7] P. Landon, P. J. Collier, A. J. Papworth, C. J. Kiely, G. J. Hutch-
ings, Chem. Commun. 2002, 2058 – 2059.
[8] T. Ishihara, Y. Ohura, S. Yoshida, Y. Hata, H. Nishiguchi, Y.
Takita, Appl. Catal. A 2005, 291, 215 – 221.
[9] K. Otsuka, I. Yamanaka, Electrochim. Acta 1990, 35, 319 – 322.
[10] I. Yamanaka, T. Hashimoto, K. Otsuka, Chem. Lett. 2002, 852 –
853.
[11] V. R. Choudhary, A. G. Gaiward, S. D. Sansare, Angew. Chem.
2001, 113, 1826 – 1829; Angew. Chem. Int. Ed. 2001, 40, 1776 –
1779.
[12] S. Melada, F. Pinna, G. Strukul, S. Perathoner, G. Centi, J. Catal.
2006, 237, 213 – 219.
[13] I. Yamanaka, T. Onizawa, S. Takenaka, K. Otsuka, Angew.
Chem. 2003, 115, 3781 – 3783; Angew. Chem. Int. Ed. 2003, 42,
3653 – 3655.
[14] I. Yamanaka, T. Onizawa, H. Suzuki, N. Hanaizumi, Chem. Lett.
2006, 35, 1330 – 1331.
[15] Z. Ogumi, K. Nishio, S. Yoshizawa, Electrochim. Acta 1981, 26,
1779 – 1782.
[16] K. Otsuka, I. Yamanaka, Catal. Today 1998, 41, 311 – 325.
Time trials of neutral H2O2 synthesis were carried out at
the [AC(HNO3) + VGCF/nafion] cathode in the complete
Exp-SPE setup at ꢀ0.3 V for 6 h. The current densities were
almost constant over 6 h. The total yield of H2O2 linearly
increased with a constant CE of 25%. However, the concen-
tration of H2O2 remained constant at 8 wt% because the
volume of the solutions linearly increased with reaction time.
When H2O2 synthesis was performed in air (P(O2) = 0.21 atm)
at ꢀ0.3 V, neutral 3 wt% H2O2 solutions were obtained with
11% CE.
In conclusion, neutral aqueous H2O2 solutions (max
8 wt%) can be continuously produced from O2 and water
1902 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 1900 –1902