Dihydrogen trioxide (HOOOH) is generated during the thermal reaction between hydrogen peroxide and ozone

Article abstract of DOI:10.1002/anie.200460457

Quantifiable amounts of HOOOH can be detected by ¹H NMR spectroscopic analysis; of the thermal reaction of O₃ and HOOH (see scheme). This first experimental report of a link between these three oxidants suggests that HOOOH may be involved in oxidation reactions that span biological, atmospheric, and environmental systems.

Full text of DOI:10.1002/anie.200460457

Communications
Oxidation
Dihydrogen Trioxide (HOOOH) Is Generated
during the Thermal Reaction between Hydrogen
Peroxide and Ozone**
Paul T. Nyffeler, Nicholas A. Boyle, Laxman Eltepu,
Chi-Huey Wong, Albert Eschenmoser,
Richard A. Lerner, and Paul Wentworth, Jr.*
In memory of Bernard (Bernie) M. Babior
The reaction between HOOH and O has fascinated scientists
3
[
1–3]
for almost a century.
The reactivity of O is markedly
3
increased by the presence of HOOH, although a clear
mechanistic understandingof this phenomenon has thus far
[
4–8]
eluded researchers.
In modern times, the reaction between
HOOH and O has received industrial utility and has been
3
named both the peroxone process and an advanced oxidation
[
9,10]
reaction.
It is one of the most potent antibacterial
processes known and is a powerful chemical remediation
tool utilized to treat soil, groundwater, and wastewater
contaminated with poly-
cyclic aromatic hydrocarbons, polychlorinated biphenyls
(PCBs), methyl tert-butyl ether (MTBE), benzene, toluene,
ethylbenzene and xylene (BTEX), trinitrotoluene (TNT), and
[
10]
other persistent organic pollutants.
Recently, the reaction between HOOH and O3 has
become of biological interest. It has been discovered that all
antibodies have the ability to catalyze the oxidation of water
1
by singlet oxygen ( D ), with the postulated intermediacy of
g
HOOOH, to generate HOOH and detectable amounts of an
[
11–13]
oxidant with the chemical signature of O3.
The proposi-
tion of HOOOH as a key intermediate in the antibody-
catalyzed reaction has led to an interest in the study of
thermally accessible routes to HOOOH that may utilize
biologically relevant components.
[
*] P. T. Nyffeler, Dr. N. A. Boyle, Dr. L. Eltepu, Prof. C.-H. Wong,
Prof. R. A. Lerner, Prof. P. Wentworth, Jr.
Department of Chemistry &
The Skaggs Institute for Chemical Biology
The Scripps Research Institute
10550 N. Torrey Pines Rd., La Jolla, CA92037 (USA)
Fax: (+1)858-784-2590
E-mail: paulw@scripps.edu
Prof. P. Wentworth, Jr.
Oxford Glycobiology Institute
Department of Biochemistry, University of Oxford
South Parks Rd., Oxford, OX1 3QU (UK)
Prof. A. Eschenmoser
Laboratorium für organische Chemie
Eidgenössische Technische Hochschule (ETH)
Hönggerberg HCl-H309
Universitaetstrasse 16 8093 Zürich (Switzerland)
[
**] This work was supported by the Skaggs Institute for Chemical
Biology. P.W. thanks the members of The Scripps Research Institute
NMR facility, especially Dr. D. Huang and Dr. L. Pasternack, for
expert assistance during this study.
4
656
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/anie.200460457
Angew. Chem. Int. Ed. 2004, 43, 4656 –4659

Angewandte
Chemie
A rigorous kinetic examination of the reaction between
O3 and HOOH was first performed by Rothmund and
[
1]
[14]
Burgstaller in 1917. Weiss first proposed a free-radical
[
3]
mechanism for this process, and subsequently Bray con-
cluded that either free radicals, COH and hydroperoxyl (HO C),
2
or HOOOH were necessary to explain the observed kinetics
[
2]
of the process. Taube and Bray in 1940 then established the
role of COH and HO C as chain-initiatingintermediates in a
2
complex radical cascade (Table 1, entries 1–4). Subsequent
ꢀ
detection of COH, the superoxide anion (O C ), and the
2
ꢀ
ozonide radical anion (O C ) as intermediates in this process
3
supported the free-radical mechanism proposed by
[
15–17]
Weiss.
Thus, the contemporary view is that a common
free-radical mechanism may account for the decomposition of
O3 by HOOH under all conditions, with the potential
involvement of HOOOH being largely ignored. Recently
however, HOOOH has been thrust into center stage as a
plausible intermediate in the photochemical peroxone pro-
1
Figure 1. H NMR spectra (600 MHz recorded at ꢀ608C) of the perox-
one reaction between d=13.50 and 13.70 ppm downfield from phenyl-
[
18]
trimethylsilane (PhTMS) in [D ]methanol (99%, external reference).
a) H O +O /O (20 s), b) H O +O /O (20 s)+H O (authentic
sample). The authentic HOOOH used as an additive in (b) was gener-
ated by the ozonation of a resin-supported 1,2-diphenylhydrazine
4
cess. Engdahl and Nelander observed that mixingHOOH
2
2
3
2
2
2
3
2
2
3
and O in an argon matrix results in a complex that, when
3
photolyzed at 266 nm, produces HOOOH. However, there is
no report of HOOOH beingdetected from the strictly
thermal peroxone reaction.
Herein we demonstrate the direct observation of
HOOOH formed duringthe thermal reaction between
[19]
(
hydrazobenzene) derivative. An aliquot of this HOOOH solution
(100 mL) was added with a glass syringe and needle to the teflon-
valved NMR tube at ꢀ608C. Note that no signal at about 13.60 ppm
was present in the following control NMR spectra: 1) After O /O was
bubbled through [D ]acetone (99.9%) for 20 s and 2) HOOH (28m) in
D ]acetone (99.9%).
3
2
6
HOOH and O . Passage of a stream of ozone/oxygen gas
3
[
6
through a solution of aqueous hydrogen peroxide (ca. 96%
to HOOOH, an authentic solution
Table 1: Selected equations from the reaction between ozone and hydrogen peroxide (in aqueous media
unless otherwise stated).
of HOOOH in [D ]acetone was
6
required as a control. Neither of
the two main chemical routes for
the preparation of HOOOH
Entry
Equation
Comment
[
[
[
[
[
[
[
a]
b]
c]
d]
e]
f]
ꢀ2 ꢀ1 ꢀ1
1
2
3
4
5
6
7
O +H O !COH+HO C+O
k =6.5”10
m
s
3
2
2
2
2
1
8
ꢀ1 ꢀ1
O +COH!HO C+O
k2=1.1”10 m
s
3
2
2
(
pulsed radiolysis of air-saturated
ꢀ
+
9
ꢀ1 ꢀ1
O + O C+H !COH+2O
k =1.6”10 m
s
3
2
2
3
[24,25]
7
ꢀ1 ꢀ1
perchloric acid solutions
or
H O +COH!HO C+H O
k =2.7”10 m
s
2
2
2
2
4
3
ꢀ1
ꢀ1
reduction of ozone by reducing
agents such as 2-ethylanthrahydro-
O +H O !H O + O
DH =ꢀ32.3 kcalmol
3
2
2
ꢀ
2
3
2
r
ꢀ
6
ꢀ1 ꢀ1
O +HO !HO +O2
k =2.8”10 m
s
3
2
3
6
g]
1
[20]
H O ! O +H O
DH =ꢀ16.2 kcalmol
,
quinone and 1,2-diphenylhydra-
2
3
2
2
r
ꢀ
ꢀ
1
1[h]
DH (0H O)=45.1 kcalmol
[20,21,23]
a
2
zine
) were suitable for this
1
DH (1H O)=12.7 kcalmol
a
2
control sample. The lifetime of
HOOOH in acid solutions
ꢀ
DH (2H O)=0.2 kcalmol
a
2
[
[
a] Ref. [2,33]. [b] Ref. [34]. [c] Ref. [35]. [d] Ref. [36,37]. [e] Gas-phase calculation: Ref. [31]. [f] Ref. [32].
g] Ref. [21,38]. [h] The number of water molecules (in parentheses) signifies the molar equivalents of
(ca. 200 ms in perchlorate pH 2)
precludes the pulsed radiolysis
method as beingviable for NMR
analysis (a single pulse sequence in
water participating as a catalyst in the reaction.
1
w/w), generated either by vacuum distillation of commercially
available HOOH (containingstabilizers) or pure HOOH
generated by reduction of O2, in [D ]acetone or [D ]tetra-
a typical H NMR experiment shown in Figure 1 is > 25 s).
[
26]
Plesnicar et al.
have shown that HOOOH is always
[
19]
contaminated with a number of oxidation products of 1,2-
diphenylhydrazine and HOOH when the low-temperature
ozonation of 1,2-diphenylhydrazine in an organic solvent is
employed. Therefore a new methodology was required for the
generation of a pure solution of HOOOH.
6
10
hydrofuran at ꢀ788C produced an intermediate that could be
1
characterized by H NMR analysis (ꢀ608C) as havinga
typical OOOH resonance at d = 13.6 ꢁ 0.2 ppm relative to
tetramethylsilane (Figure 1b).
This signal disappeared upon warming the [D ]acetone
The development of polymer-supported methodologies to
facilitate solution-phase chemistry has exploded over the past
decade in response to the stringent requirements of high-
6
and [D ]tetrahydrofuran solutions to room temperature and
1
0
1
did not reappear upon recooling. The H NMR chemical shift
and thermal instability of this intermediate are both in line
with the wealth of physical data available for HOOOH.
throughput solution-phase combinatorial and parallel syn-
[
18,20–23]
[27–30]
thesis.
We rationalized that HOOOH could be prepared
1
To add support to the notion that this H NMR resonance
free of the contamination products arisingfrom the oxidation
of 1,2-diphenylhydrazine (1,2-DPH) if the 1,2-DPH was on a
at d = 13.6 ꢁ 0.2 ppm in the peroxone reactions corresponds
Angew. Chem. Int. Ed. 2004, 43, 4656 –4659
www.angewandte.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4657

Communications
resin support. Thus, resin-supported 1,2-DPH derivative 1
was prepared (Scheme 1).
The amount of HOOOH generated during the peroxone
reaction was determined from an analysis of the relationship
between the area of the OOOH resonance (13.6 ꢁ 0.2 ppm)
and the amount of HOOOH present in the NMR sample. This
relationship was determined by a standard method involving
interpolation of the ratio of the peak areas of the OOOH
resonance at 13.6 ꢁ 0.2 ppm and the methyl protons of an
external standard (phenyltrimethylsilane) to a standard
calibration curve. This analysis revealed that approximately
[
19]
2
2
9 mmol of HOOOH (correspondingto a concentration of
.9 mm in the [D ]acetone solution) was formed in this
6
specific peroxone reaction (10 mL reaction volume). This
quantity of H O3 was reproducibly generated when the
2
peroxone reaction was repeated under identical conditions
of HOOH concentration and O flow. This amount of H O
3
2
3
Scheme 1. Preparation of resin-bound diphenylhydrazine 1 and its use
to prepare HOOOH from O3.
generated in the peroxone process is significant, and therefore
the chemistry of the peroxone reaction may now have to be
revised to not only consider ozone and HOOH chemistry, but
also that of H O .
2
3
The resin-supported hydrazine 1 was prepared in two
steps by an initial addition of 4-phenylazophenol (2) to low
cross-linked (1% divinylbenzene) chloromethylpolystyrene
followed by reduction of the orange azo resin derivative 3
with tri-n-butyltin hydride in refluxingtoluene. Low-temper-
There are at least two mechanistically viable possibilities
for the origin of HOOOH in the peroxone reaction. By using
quantum chemical methods Xu and Goddard have calcu-
[
31]
lated that the enthalpy of reaction (DH ) for the thermal
r
3
reaction of HOOH and O to yield HOOOH and O in the
3
2
ꢀ
1
ature ozonation (ꢀ788C) of resin-1,2-DPH 1 in [D ]acetone
gas phase is ꢀ32.3 kcalmol , with the highest enthalpy of
6
ꢀ1
or [D ]tetrahydrofuran with a stream of ozone/oxygen led to
activation (DH ) being4.8 kcalmol (Table 1, entry 5). The
1
0
a
the formation of HOOOH in solution, with the only
contaminant beingHOOH (Fi ug re 2). The oxidized resin
turns bright orange during the ozonation, as it becomes
oxidized in part to 3, and can then be removed by filtration.
process involves two sequential single electron transfer
reactions that is formally the redox reaction between O
3
[
31]
and HOOH. Xu and Goddard
have
proposed that the reaction proceeds via a
planar seven-membered-ringbiradical
intermediate (with the spin multiplicity
of the triplet and singlet states being
Addition of the solution of HOOOH in [D ]acetone
6
(
generated by ozonation of 1) to the NMR tube containing
the peroxone reaction resulted in an increase in the area of
the OOOH resonance (13.6 ꢁ 0.2 ppm, Figure 1a), thus
strengthening the hypothesis that the unknown resonance
corresponds to HOOOH.
almost degenerate) to yield HOOOH
3
and O (Figure 3 and Table 1, entry 5).
Figure 3. Postu-
lated seven-mem-
bered-ring triplet
biradical inter-
mediate on the
pathway from
2
More recently, to explain hydrogen iso-
tope effects on the reaction kinetics of the
peroxone reaction in water, Lesko et al.
have concluded that HO3 , the conjugate
base of H O , may be generated from a
[
32]
ꢀ
2
3
H O and O into
2
2
3
3
reaction between the anion of hydrogen
H O and O .
2 3 2
peroxide and O (Table 1, entry 6).
3
The observation that HOOOH is generated in the thermal
peroxone reaction in low-water-content organic solvents at
low temperature brings with it clear questions. These include
the as yet, unknown generation of HOOOH under totally
aqueous conditions at room temperature of a peroxone
[
24,25]
mixture. Bielski
has shown that HOOOH, detected
chemically and spectroscopically, is quasistable under totally
aqueous conditions (t = ca. 2 s at pH 1.5, 298 K). Koller and
1
/2
[
21]
[31]
Plesnicar
and Xu and Goddard
have independently
1
calculated that HOOOH decomposes to yield O and H O in
2
2
a water-catalyzed reaction (Table 1 entry 7). Thus, although
1
not yet reported, O2 may well be generated during the
peroxone reaction in aqueous systems from the known
decomposition of H O . In addition, further reactions of
1
2
3
Figure 2. H NMR spectrum (600 MHz recorded at ꢀ608C) of the fil-
1
HOOOH with either itself, HOOH, O , or O to yield, as yet,
trate following low-temperature ozonation (ꢀ788C) of resin-1. Note
2
3
undefined products will add further complexity to this rich
and complex reaction process.
PhTMS in [D ]methanol is present as an external standard for quantifi-
4
cation of HOOOH.
4
658
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2004, 43, 4656 –4659

Angewandte
Chemie
[
23] B. Plesnicar, J. Cerkovnik, T. Tekavec, J. Koller, Chem. Eur. J.
000, 6, 809.
24] B. H. J. Bielski, J. Phys. Chem. 1970, 74, 3213.
25] B. H. J. Bielski, H. A. Schwartz, J. Phys. Chem. 1968, 72, 3836.
26] B. Plesnicar, T. Tuttle, J. Cerkovnik, J. Koller, D. Cremer, J. Am.
Chem. Soc. 2003, 125, 11553.
27] C. Spanka, P. Wentworth, K. D. Janda, Comb. Chem. High
Throughput Screening 2002, 5, 233.
28] P. Wentworth, Jr., K. D. Janda, Curr. Opin. Biotechnol. 1998, 9,
Experimental Section
2
Reaction between O and HOOH: All samples were prepared as
3
[
[
[
solutions in [D ]acetone (99.9%) or [D ]tetrahydrofuran (99.9%,
6
10
1
0 mL). The experiments were carried out by usingSchlenk
techniques. In brief, a stream of O in O was bubbled through an
HOOH solution (28m) in [D ]acetone (99.9%) or [D ]tetrahydro-
furan (99.9%, 10 mL) precooled to ꢀ788C for 20 s. The hydrogen
peroxide used in these studies was prepared either by the resin
method described in the SupportingInformation or by vacuum
3
2
6
10
[
[
109.
distillation of an aqueous H O2 (50% w/v containingstabilizers)
2
[
[
[
29] P. Wentworth, Jr., K. D. Janda, Chem. Commun. 1999, 1917.
30] P. Wentworth, Jr., Trends Biotechnol. 1999, 17, 448.
31] X. Xu, W. A. Goddard, Proc. Natl. Acad. Sci. USA 2002, 99,
15308.
solution to approximately one-third volume. These ozonolyzed
solutions were then degassed briefly by bubbling argon through the
solution of H O and transferred to a teflon-valved NMR tube. The
2
2
1
samples were then warmed to ꢀ608C and H NMR spectra were
measured. Each experiment was repeated at least twice and gave
essentially the same amount of HOOOH.
[
[
[
[
32] T. M. Lesko, A. J. Colussi, M. R. Hoffman, J. Am. Chem. Soc.
2
004, 126, 4432.
33] K. Sehested, H. Corfitzen, J. Holcman, E. J. Hart, J. Phys. Chem.
992, 96, 1005.
Generation of HOOOH in [D ]acetone with 1: Resin-supported
6
1
1
,2-diphenylhydrazine 1 was suspended in [D ]acetone (5 mL) and
6
34] K. Sehested, J. Holcman, E. Bjergbakke, E. J. Hart, J. Phys.
Chem. 1984, 88, 4144.
35] NDRL/NIST Solutions Kinetics Database, NIST Standard
Reference Database 40, http://kinetics.nist.gov/solution/
index.php, 2002.
cooled to ꢀ788C. A stream of ozone/oxygen gas was passed through
the suspension for 2 min. The resin was then removed by filtration
under argon and the filtrate was quickly transferred to an NMR tube
under a positive argon pressure.
[
[
[
36] H. Christensen, K. Sehested, H. Corfitzen, J. Phys. Chem. 1982,
Received: April 26, 2004
Published Online: August 12, 2004
86, 1588.
37] G. V. Buxton, C. L. Greenstock, W. P. Helman, A. B. Ross, J.
Phys. Chem. Ref. Data 1988, 17, 513.
38] X. Xu, P. R. Muller, W. A. Goddard III, Proc. Natl. Acad. Sci.
USA 2002, 99, 3376.
Keywords: deuterium · hydrogen peroxide · oxidation · ozone ·
.
solid-phase synthesis
[
[
[
[
1] V. Rothmund, A. Burgstaller, Monatsh. Chem. 1917, 38, 295.
2] H. Taube, W. C. Bray, J. Am. Chem. Soc. 1940, 62, 3357.
3] W. C. Bray, J. Am. Chem. Soc. 1938, 60, 82.
4] C.-H. Kuo, L. Zhong, M. E. Zappi, A. P. Hong, Can. J. Chem.
Eng. 1999, 77, 473.
[
5] R. J. Spanggord, D. Yao, T. Mill, Environ. Sci. Technol. 2000, 34,
450.
[
[
6] M. Elovitz, U. von Gunten, Ozone Sci. Eng. 1999, 21, 239.
7] M. Elovitz, U. von Gunten, H. Kaiser, Ozone Sci. Eng. 2000, 22,
123.
[
[
8] J. L. Acero, U. von Gunten, Ozone Sci. Eng. 2000, 22, 305.
9] R. Brunet, M. M. Bourbigot, M. Dore, Ozone Sci. Eng. 1984, 6.
[
10] J. Prousek, Chem. Listy 1996, 60, 229.
[
11] P. Wentworth, Jr., L. H. Jones, A. D. Wentworth, X. Zhu, N. A.
Larsen, I. A. Wilson, X. Xu, W. A. Goddard III, K. D. Janda, A.
Eschenmoser, R. A. Lerner, Science 2001, 293, 1806.
[
12] P. Wentworth, Jr., J. McDunn, A. D. Wentworth, C. Takeuchi, J.
Nieva, K. D. Janda, A. Eschenmoser, R. A. Lerner, Science 2002,
298, 2195.
[
13] P. Wentworth, Jr., A. D. Wentworth, X. Zhu, I. A. Wilson, K. D.
Janda, A. Eschenmoser, R. A. Lerner, Proc. Natl. Acad. Sci.
USA 2003, 100, 1490.
[
[
14] J. Weiss, Trans. Faraday Soc. 1935, 31, 668.
15] K. Sehested, H. Corfitzen, J. Holcman, E. J. Hart, J. Phys. Chem.
A 1998, 102, 2667.
[
16] J. Staehelin, R. E. Buhler, J. Hoigne, J. Phys. Chem. 1984, 88,
5
999.
17] R. E. Buhler, J. Staehelin, J. Hoigne, J. Phys. Chem. 1984, 88,
560.
[
2
[
[
[
[
[
18] A. Engdahl, B. Nelander, Science 2002, 295, 482.
19] See SupportingInformation for full details.
20] J. Cerkovnik, B. Plesnicar, J. Am. Chem. Soc. 1993, 115, 12169.
21] J. Koller, B. Plesnicar, J. Am. Chem. Soc. 1996, 118, 2470.
22] B. Plesnicar, J. Cerkovnik, T. Tuttle, E. Kraka, D. Cremer, J. Am.
Chem. Soc. 2002, 124, 11260.
Angew. Chem. Int. Ed. 2004, 43, 4656 –4659
www.angewandte.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4659