Safe direct synthesis of high purity H2O2 through a H2/O2 plasma reaction

Article abstract of DOI:10.1002/anie.201304134

Plasma makes it safer: A gaseous H₂/O₂ plasma reaction has been developed for the safe direct synthesis of H₂O ₂. Low electron density favors the generation of H₂O ₂ by a chain termination path. This plasma method is promising for the direct synthesis of neutral, high concentration (ca. 60 wt %), and high purity (electronic grade) H₂O₂. Copyright

Full text of DOI:10.1002/anie.201304134

.
Angewandte
Communications
DOI: 10.1002/anie.201304134
Hydrogen Peroxide Synthesis
Safe Direct Synthesis of High Purity H O through a H /O Plasma
2
2
2
2
Reaction**
Yanhui Yi, Juncheng Zhou, Hongchen Guo,* Jianli Zhao, Ji Su, Li Wang, Xiangsheng Wang, and
Weimin Gong
[
1]
Hydrogen peroxide (H O ), the most desired green oxidant,
Herein, we report an experimental realization of control-
lable H /O combustion processes by an optimized plasma
reactor. High purity (Grade 1 electronic grade H O accord-
2 2
2
2
is almost exclusively produced by an anthraquinone (AQ)
2
2
[
2]
process. Direct oxidation of H2 with O2 has long been
[3]
considered an ideal alternative for H O2 production.
ing to the SEMI standard) and highly concentrated H O
2
2
2
Extensive studies have been done on direct H O synthesis
solution (ca. 60 wt%) can be directly produced from a H /O
2
2
2
2
from a H /O mixture. To achieve high efficiency, direct H O
2
mixture without explosion. These results suggest a different
mechanism from conventional H /O combustion processes in
2
2
2
synthesis is generally performed in acidified solvent over
supported noble-metal catalysts (Au, Pd, Au–Pd, and Pd–
2
2
the H /O2 plasma reaction. As shown in Scheme 1, the
2
[
4–11]
Pt).
However, the direct synthesis of H O from a H /O
electron activation of H2 into H is responsible for H O
2
2
2
2
2
2
mixture catalyzed by metals is quite hazardous, and it is very
difficult to directly obtain high-purity and high-concentration
H O .
2
2
[12,13]
Research
published in the 1960s has demonstrated
that H O can be generated in H /O non-equilibrium plasma
2
2
2
2
through free-radical reactions in the absence of any catalyst or
chemical. However, this plasma method has not yet drawn
much attention, owing to low H O yield (less than ca. 5%)
2
2
and safety concerns about the discharge-triggered H /O
2
2
[
12,14]
reaction.
The content of O must be strictly controlled
2
Scheme 1. Main reaction network for formation of H O and H O in
2
2
2
[15]
below 4 mol% in order to prevent explosion and ignition.
non-equilibrium H /O plasma.
2
2
[
16]
Our previous research showed that the structure of the
plasma reactor played an important role in the direct
synthesis of H O . A H /O mixture containing 3 mol% of
formation. However, the electron activation of O into O
2
2
2
2
2
O reaches 100% O conversion, but the H O selectivity is
and O * (vibrational and electronic excited states) leads to
2
2
2
2
2
only 3.5% (based on O ) in a single dielectric barrier
H O formation and explosion of the H /O mixture. More-
2
2
2
2
discharge (SDBD) plasma reactor with a naked metal high-
voltage (HV) electrode and an aqueous grounding electrode.
On the other hand, 57.8% O conversion and 56.3% H O
over, low-electron-density H /O plasma leads to a low degree
2 2
of H and O activation, which plays quite a significant role in
2
2
producing H O and controlling H /O combustion.
2
2
2
2
2
2
2
selectivity (based on O ) can be obtained by using a double
The plasma reactor used in our experiments is a double
aqueous electrode DDBD reactor (Supporting Information,
Figure S1), which has an aqueous grounding electrode and an
aqueous HV electrode. When a mixture of H /O is trans-
2
dielectric barrier discharge (DDBD) plasma reactor with
a pyrex-covered metal HV electrode (the pyrex cover acts as
an additional dielectric barrier) and an aqueous grounding
electrode. Although the selectivity has been greatly
improved, the safety concerns and low efficiency, owing to
2
2
formed into H /O plasma, H O and H O are formed and
2
2
2
2
2
condense on the reactor wall, then flow into the collector.
As shown in Table 1, using the DDBD reactor with double
aqueous electrode, the selectivity and concentration of H O
low O content, are still big challenges.
2
2
2
are almost more than 60%. The content of O in the feed can
2
be increased by decreasing the flow rate of H ; by this
2
[
+]
[+]
[
*] Y. H. Yi, J. C. Zhou, Prof. H. Guo, J. L. Zhao, J. Su, L. Wang,
method, the allowable O content reaches 30 mol% while
2
X. S. Wang, W. M. Gong
ensuring the safety of the reaction. With a H /O mixture
2
2
State Key Laboratory of Fine Chemicals, Department of Catalytic
Chemistry and Engineering Dalian University of Technology
Dalian 116012 (P.R. China)
ꢀ1
containing 2.0 mol% O at a total flow rate of 500 mLmin ,
2
5
7% O2 conversion and 72% H O2 selectivity can be
2
obtained (based on O ). The concentration of H O is
E-mail: hongchenguo@163.com
2
2
2
+
67 wt% when the space-time yield of H O
is
2
2
[
] These authors contributed equally to this work.
ꢀ
1
ꢀ1
2
6 gH2O2 L h . As the content of O in the feed was increased
2
[
**] We acknowledge the financial support from the Natural Science
Foundation of China (20233050). We acknowledge some equipment
help from Prof. Ma Xuehu and Prof. Zhang Jialiang.
to 14.3 mol%, the conversion of O reached 91%, and the
2
selectivity and concentration of H O were 64% and 62 wt%,
2
2
respectively. In addition, the space-time yield of H O
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201304134.
2
2
ꢀ1
ꢀ1
increased to 37 gH2O2
L h .
8446
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 8446 –8449

Angewandte
Chemie
Table 1: H O synthesis with varying O content in the double aqueous
2
2
2
[
a]
electrode DDBD reactor.
O con- O2 H O
CH2O2 Space-time
Energy con-
sumption
[kWhKgH2O2 ]
2
2
2
tent
conv. selectivity [wt%] yield
ꢀ1
ꢀ1 [b]
ꢀ1
[V%]
[%]
[%]
[gH2O2
L
h
]
2
4
6
9
4.3
0.0
5.0
0.0
.0
.0
.3
.1
57
70
79
85
91
96
98
99
72
70
68
66
64
59
57
50
67
66
64
63
62
57
54
48
26
32
35
36
37
37
36
32
27
23
20
20
19
19
20
22
1
2
2
3
Figure 1. Optical images of H /O DBD plasma. a) Optical image in
2
2
the SDBD reactor. b) Optical image in the double aqueous electrode
ꢀ1
ꢀ1
ꢀ
1
DDBD reactor. (H 152 mLmin , O 8 mLmin , aqueous grounding
2 2
[a] Aqueous grounding electrode at 58C, 1 atm, 10 mLmin O flow,
2
electrode at 58C, input power 10 W).
input power=10 W, 14.1 mL reactor. [b] The space-time yield of H O is
2
2
counted based on 100 wt%. Experimental error: Conversion ꢁ1%,
ꢀ
1
ꢀ1
Selectivity ꢁ1%, C
ꢁ1 wt%, Space-time yield ꢁ1 g
L h ,
H2O2
H2O2
ꢀ
1
ꢀ1
Energy consumption ꢁ1 g
kW
h
.
DDBD reactor conducts a weak and homogeneous discharge,
which has lower electron density.
H2O2
Second, when the reactors were used for O discharge,
2
Significantly, without any concentration or purification,
the H O solution is obtained in a concentrated (ca. 60 wt%)
optical emission spectra (OES) show that the plasma
generated by the SDBD reactor have more intensive
2
2
form, which complies with the requirements for Grade 1
electronic-grade H O , according to the SEMI standard
O lines (Figure 2a) and singlet O spectra (Figure 2b) than
2
the double aqueous electrode DDBD reactor. Similarly, when
the reactors were used to synthesize H O through a H /O
2
2
[
17]
(
Table S1). The energy consumption for H O production
2 2
2
2
2
2
ꢀ
1
is 19.0 kWhKgH2O2 at 14.3% O (Table 1), which is about
mixture discharge (Figure 2c), the plasma generated by the
SDBD reactor shows much stronger H lines, O lines and
2
a factor of five times higher than for the AQ process.
However, the highly energy-consuming concentration and
purification processes are avoided, and the equipment invest-
ment costs would be dramatically reduced. Therefore, this
simple plasma method is attractive for the direct production
of concentrated, neutral, and high purity H O , although the
H emission bands than those in the double aqueous electrode
2
DDBD reactor. The intensities of the O lines (777.5 nm and
844.7 nm) increases with increasing O content in both
2
reactors (Figure 2d; see also Figures S5A and B), whereas
the intensity of the H line shows the reverse trend (Fig-
ure S5C). Surprisingly, when an explosion took place in the
SDBD reactor and the double aqueous electrode DDBD
reactor, the intensities of the O lines (777.5 and 844.7 nm)
2
2
selectivity of H O is lower than in AQ process.
2
2
Safety is a challenge in the direct synthesis of H O by
2
2
a H /O mixture because of the broad composition range that
2
2
is explosive (6–96 mol% O ). In the double aqueous elec-
were near 260 and 150, respectively; the O content was about
2
2
trode DDBD reactor, the H /O plasma reaction is safe when
10 mol% for the SDBD reactor and 30 mol% for the double
aqueous electrode DDBD reactor. The intensities of the
O lines are an indicator of the densities of active oxygen
2
2
the O2 content is up to 30 mol%. This is in contrast to
a SDBD plasma reactor (Figure S2), where explosions would
take place immediately after the start of the discharge when
[
18]
(Sections S7 and S8), thus the explosion of H /O mixture
2
2
the O content was above 10 mol%. To understand why the
under non-equilibrium plasma conditions will take place
when the density of active oxygen is beyond the critical value.
2
H /O plasma reactions can be safely conducted with such
2
2
a high O content in the double aqueous electrode DDBD
Moreover, the active oxygen also leads to H O formation.
2
2
reactor, several comparative in situ diagnostic studies were
carried out in the SDBD and double aqueous electrode
DDBD reactors.
It seems puzzling that the yields of H O remains at high
values for the double aqueous electrode DDBD reactor when
2
2
the content of O increases from 6.3% to 30% (Table 1), as
2
First, the discharge behavior of the double aqueous
electrode DDBD reactor is quite different from that of the
SDBD reactor. Briefly, the discharge of the SDBD reactor is
from spark filaments (local and highly ionized narrow current
pathways; Figure 1a); it is accompanied by strong discharge
current pulses (Figure S3) and continuous temperature pul-
sating (25–508C; Figure S4). However, the discharge of the
double aqueous electrode DDBD reactor is diffusive and
uninterrupted throughout the discharge zone (Figure 1b); it
has weak discharge current pulses (Figure S3) and almost
unchanged space temperature (ca. 228C; Figure S4C). The
higher the discharge current, the higher the electron density.
These facts suggest that the reaction temperatures in the two
reactors are at a low level and the double aqueous electrode
the amounts of O and excited O increase with increasing O
content (Figure 2). However, in terms of the relative inten-
2
2
sities of the O lines and the degree of activation of O , these
2
values remain constant over the whole range of O content
2
(Figure S7A and B). Furthermore, the degree of activation of
O2 in the double aqueous electrode DDBD reactor is
approximately one third of that in the SDBD reactor.
Hence, it is reasonable that the degree of activation of O2
mainly determines the formation efficiency of H O when the
2
2
degree of activation of H (Figure S7C and D) is fixed and
2
sufficient. A low degree of activation of O enhances the
2
formation efficiency of H O , whereas a high degree of
2
2
activation of O favors the formation of H O. The densities of
2
2
Angew. Chem. Int. Ed. 2013, 52, 8446 –8449
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 8447

.
Angewandte
Communications
Figure 3. Isotopic product distribution of H O synthesized by H /O
2
2
2
2
ꢀ
1
ꢀ1 16
ꢀ1
plasma. a) 165 mLmin H flow, 4 mLmin
O flow, 4 mLmin
2
2
1
8
ꢀ1
ꢀ1 16
O flow. b) 165 mLmin H flow, 8 mLmin
O flow.
2
2
2
O and excited state O determine whether the explosion of
2
H /O takes place.
2
2
Third, to understand the mechanism of this H O synthesis
2
2
method, isotope tracing experiments were conducted.
Figure 3 shows that the isotopic product distribution of the
plasma method is remarkably different from that of the noble-
[
19]
metal-based direct synthesis. It can be deduced that the
recombination of HO is the primary path of H O formation
2
2
2
(
HO + HO !H O + O ; Table S2) and HO radical is a key
2
2
2
2
2
2
intermediate. This conclusion is consistent with the exper-
imental results. On the one hand, OES analysis (Figure 2c)
shows that hydrogen atoms are the main active atom species,
which means that HO can be formed by H + O !HO . On
2
2
2
the other hand, H O is the principal product in the double
2
2
aqueous electrode DDBD reactor. However, the isotope
tracing experiments provide little insight into the formation of
H O.
2
Generally, the chain-termination reaction (H + O !
2
HO ) leads to H O formation, whereas the chain-branching
2
2
2
reactions (H + O !OH + O and O + H !OH + H) result in
2
2
[
20,21]
H O formation.
At ambient temperature, the rate
2
coefficient of the chain termination reaction, H + O !HO ,
2
2
is about eight orders of magnitude larger than that of the
chain branching reaction, H + O !OH + O, owing to the
2
[
22,23]
energy barrier of the latter reaction.
However, the rate
coefficient of H + O !OH + O will dramatically increase
2
when the oxygen molecule is in an excited state, because the
[
24,25]
barriers of the reaction decrease for excited-state O2.
Hence, the rate ratio of H + O !OH + O to H + O !HO
2
2
2
increases with an increase of the ratio of excited O molecules
2
to ground-state O molecules, which results in an increase in
2
the OH/HO ratio. Furthermore, OH can also be produced by
Figure 2. OES of O plasma and H /O plasma in the SDBD and
2
2
2
2
double aqueous electrode DDBD reactors (aqueous grounding elec-
O + H !OH + H, and the O atoms are mainly produced by
2
ꢀ
1
trode at 58C, 1 atm, input power 10 W). a) 40 mLmin O flow,
the electron impact dissociation of O and H + O *!OH + O.
2
2
2
ꢀ
1
ꢀ1
3
1
8
00 Gmm grating, 0.5 s exposure time. b) 40 mLmin O flow,
2
The rate coefficient of O + H !OH + H is small, but it can be
2
ꢀ
1
ꢀ1
800 Gmm grating, 120 s exposure time. c) 152 mLmin H flow,
2
remarkably enhanced by vibrational excited electronic
ꢀ
1
ꢀ1
mLmin O flow, 300 Gmm grating, 0.5 s exposure time. d) Inten-
2
[26]
ground-state H2. The high concentration of vibrational H2
sity of O lines of H /O plasma in the SDBD and double aqueous
2
2
ꢀ1
in the H /O2 plasma makes this path important for OH
electrode DDBD reactors with different O content (160 mLmin
H +O flow, 300 Gmm grating, 0.5 s exposure time). DDBD-O
2
2
ꢀ1
formation (Scheme 1).
2
2
7
77.5 nm (*), DDBD-O 844.7 nm (~), SDBD-O 777.5 nm (*), SDBD-
O 844.7 nm (~).
It is easy for an explosion to take place in the SDBD
reactor. This is because the high electron density results in
a high density of H2 (v = n) and active oxygen species
(
(
Figure 2), which favors chain-branching reactions
Scheme 1). In the double aqueous electrode DDBD reactor,
8448 www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 8446 –8449

Angewandte
Chemie
the discharge (Figure 1b) is very similar to the Townsend
discharges characterized by low current density and low
[3] a) V. R. Choudhary, A. G. Gaikwad, S. D. Sansare, Angew.
Chem. 2001, 113, 1826 – 1829; Angew. Chem. Int. Ed. 2001, 40,
1776 – 1779; b) P. Landon, P. J. Collier, A. J. Papworth, C. J.
Kiely, G. J. Hutchings, Chem. Commun. 2002, 2058 – 2059.
4] G. Blanco-Brieva, E. Cano-Serrano, J. M. Campos-Martin,
J. L. G. Fierro, Chem. Commun. 2004, 1184 – 1185.
[27]
electron density,
which favors the chain termination
reaction and thus results in a low degree of activation of
oxygen (Section S9, Figure S7B and D), enhanced safety of
[
t
h
e
H
/
O
d
i
s
c
h
a
r
g
e
,
a
n
d
t
h
e
h
i
g
h
e
r
s
e
l
e
c
t
i
v
i
t
y
f
o
r
H
O
2
2
2
2
[5] a) V. R. Choudhary, C. Samanta, A. G. Gaikwad, Chem.
Commun. 2004, 2054 – 2055; b) V. R. Choudhary, C. Samanta,
P. Jana, Chem. Commun. 2005, 5399 – 5401.
(
Scheme 1).
In summary, gaseous H /O plasma reactions have been
2
2
[
6] F. Menegazzo et al., J. Catal. 2008, 257, 369 – 381.
demonstrated to be controllable for the direct synthesis of
H O when using a double aqueous electrode DDBD reactor.
[
7] a) Q. S. Liu, J. C. Bauer, R. E. Schaak, J. H. Lunsford, Angew.
Chem. 2008, 120, 6317 – 6320; Angew. Chem. Int. Ed. 2008, 47,
2
2
In this method, low electron density favors the generation of
H O by a chain termination path. This plasma method is
6221 – 6224; b) J. K. Edwards, E. Ntainjua N, A. F. Carley, A. A.
2
2
Herzing, C. J. Kiely, G. J. Hutchings, Angew. Chem. 2009, 121,
8664 – 8667; Angew. Chem. Int. Ed. 2009, 48, 8512 – 8515.
promising for the direct synthesis of neutral, high concen-
tration (ca. 60 wt%) and high purity (electronic grade) H O .
[8] J. K. Edwards, B. Solsona, E. Ntainjua N, A. F. Carley, A. A.
Herzing, C. J. Kiely, G. J. Hutchings, Science 2009, 323, 1037 –
1041.
2
2
It has also been successfully integrated with a titanium silicate
[28]
molecular sieve catalyst for propene epoxidation.
On the other hand, the selectivity of H O should be
[
9] E. N. Ntainjua N, M. Piccinini, S. J. Freakley, J. C. Pritchard, J. K.
2
2
Edwards, A. F. Carley, G. J. Hutchings, Green Chem. 2012, 14,
improved further, because it is still much lower than that of
the AQ process (above 95%); moreover, more energy
170 – 181.
[
10] a) S. Parasher, M. A. Rueter, B. Zhou, U.S. Patent
20120020872A1, 2012; b) J. K. Edwards, J. Pritchard, M. Picci-
nini, G. Shaw, Q. He, A. F. Carley, C. J. Kiely, G. J. Hutchings, J.
Catal. 2012, 292, 227 – 238.
[11] J. Xu, L. Ouyang, G. J. Da, Q. Q. Song, X. J. Yang, Y. F. Han, J.
Catal. 2012, 285, 74 – 82.
12] K. Morinaga, Bull. Chem. Soc. Jpn. 1962, 35, 345 – 348.
13] M. Venugopalan, R. A. Jones, Chem. Rev. 1966, 66, 133 – 160.
14] K. Morinaga, Bull. Chem. Soc. Jpn. 1962, 35, 627 – 630.
15] S. Endoh, K. Namba, S. Yagi, K. Maeda, U.S. Patent 5378436,
1995.
16] J. C. Zhou, H. C. Guo, X. S. Wang, M. X. Guo, J. L. Zhao, L. X.
Chen, W. M. Gong, Chem. Commun. 2005, 1631 – 1633.
17] R. Abejꢀn, A. Garea, A. Irabien, AICHE J. 2012, 58, 3718 –
efficient means of triggering the H /O2 plasma reaction
2
should be studied in the future. Significantly, the mechanism
[
29]
of the control process sheds new light on plasma chemistry,
[
30,31]
[32]
radical chemistry,
and material treatment by plasma.
This control process could be helpful for other chemical
reactions, especially for certain selective redox reactions.
[
[
[
[
Experimental Section
[
The flow velocities of H2 and O were monitored by a mass flow
2
controller and mixed homogeneously to pass through the plasma
reactor. The temperature of circulating water was maintained at ca.
[
3
730.
[18] a) R. E. Walkup, K. L. Saenger, G. S. Selwyn, J. Chem. Phys.
986, 84, 2668 – 2674; b) M. A. Worsley, S. F. Bent, N. C. M.
5
8C by a refrigeration unit. After about 10 min (remove O and N to
2 2
ensure safety), the voltage of the HV electrode was adjusted by
controlling the plasma power (high performance computerized
plasma and corona discharge experiment generators CTP-2000K) to
initiate the DBD discharge. The H /O composition of the feed and
1
Fuller, T. Dalton, J. Appl. Phys. 2006, 100, 083301.
[
[
[
19] D. P. Dissanayake, J. H. Lunsford, J. Catal. 2003, 214, 113 – 120.
20] E. P. Dougherty, H. Rabitz, J. Chem. Phys. 1980, 72, 6571 – 6586.
21] J. Li, Z. W. Zhao, A. Kazakov, F. L. Dryer, Int. J. Chem. Kinet.
2
2
effluent was analyzed by an on-line gas chromatograph. The H O
2
2
concentration of the collected product solution was determined by
iodometry, then the H O selectivity was calculated. In the process,
2
004, 36, 566 – 575.
22] C. Xu, D. Xie, D. H. Zhang, S. Y. Lin, H. Guo, J. Chem. Phys.
005, 122, 244305.
23] S. R. Sellevꢁg, Y. Georgievskii, J. A. Miller, J. Phys. Chem. A
008, 112, 5085 – 5095.
2
2
[
[
[
[
[
the discharge voltage, discharge current, and power were measured
on site by a digital oscilloscope (Tektronix DPO 3012, HV probe
Tektronix P6015A, current probe Pearson 6585). The discharge
images were taken by a camera (Nikon D50).
2
2
24] J. Y. Ma, H. Guo, C. J. Xie, A. Y. Li, D. Q. Xie, Phys. Chem.
Chem. Phys. 2011, 13, 8407 – 8413.
25] G. Quꢂmꢂner, B. K. Kendrick, N. Balakrishnan, J. Chem. Phys.
Received: May 14, 2013
Published online: June 26, 2013
2010, 132, 014302.
26] R. A. Sultanov, N. Balakrishnan, Astrophys. J. 2005, 629, 305 –
310.
27] U. Kogelschatz, IEEE. T. Plasma. Sci. 2002, 30, 1400 – 1408.
28] J. L. Zhao, J. C. Zhou, J. Su, H. C. Guo, X. S. Wang, W. M. Gong,
AIChE. J. 2007, 53, 3204 – 3209.
[29] W. Sun, M. Uddi, S. H. Won, T. Ombrello, C. Carter, Y. Ju,
Combust. Flame 2012, 159, 221 – 229.
[30] J. Mao et al., Atmos. Chem. Phys. 2010, 10, 5823 – 5838.
[31] P. S. Monks, Chem. Soc. Rev. 2005, 34, 376 – 395.
[32] F. S. Denes, S. Manolache, Prog. Polym. Sci. 2004, 29, 815 – 885.
Keywords: green chemistry · hydrogen peroxide ·
.
plasma chemistry · radical chemistry · synthetic methods
[
[
[
1] H. T. Hess in Kirk-Othmer Encyclopedia of Chemical Engineer-
ing (Eds.: I. Kroschwitz, M. Howe-Grant), Wiley, New York,
1995, chap. 13, pp. 961.
[
2] J. M. Campos-Martin, G. Blanco-Brieva, J. L. G. Fierro, Angew.
Chem. 2006, 118, 7116 – 7139; Angew. Chem. Int. Ed. 2006, 45,
6962 – 6984.
Angew. Chem. Int. Ed. 2013, 52, 8446 –8449
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 8449