Disproportionation pathways of aqueous hyponitrite radicals (HN 2O2/N2O2-)

Article abstract of DOI:10.1021/jp803230c

Pulse radiolysis and flash photolysis are used to generate the hyponitrite radicals (HN₂O₂W₂O₂^-) by one-electron oxidation of the hyponitrite in aqueous solution. Although the radical decay conforms to simple second-order kinetics, its mechanism is complex, comprising a short chain of NO release-consumption steps. In the first, rate-determining step, two N₂O₂^- radicals disproportionate with the rate constant 2k = (8.2 ± 0.5) × 10 ⁷ M^-1 s^-1 (at zero ionic strength) effectively in a redox reaction regenerating N₂O₂^2- and releasing two NO. This occurs either by electron transfer or, more likely, through radical recombination-dissociation. Each NO so-produced rapidly adds to another N₂O₂^-, yielding the N₃O ₃^- ion, which slowly decomposes at 300 s^-1 to the final N₂O + NO₂^- products. The N ₂O₂^- radical protonates with pK_a = 5.6 ± 0.3. The neutral HN₂O₂ radical decays by an analogous mechanism but much more rapidly with the apparent second-order rate constant 2k = (1.1 ± 0.1) × 10⁹ M^-1 s ^-1. The N₂O₂^- radical shows surprisingly low reactivity toward O₂ and O₂^-, with the corresponding rate constants below 1 × 10⁶ and 5 × 10⁷ M^-1 s^-1. The previously reported rapid dissociation of N₂O₂^- into N₂O and O^- does not occur. The thermochemistry of HN₂O ₂/N₂O₂^- is discussed in the context of these new kinetic and mechanistic results.

Full text of DOI:10.1021/jp803230c

J. Phys. Chem. A 2008, 112, 8295–8302
8295
•
•-
Disproportionation Pathways of Aqueous Hyponitrite Radicals (HN2O2 /N2O2 )
†
‡
,†
Gregory A. Poskrebyshev, Vladimir Shafirovich, and Sergei V. Lymar*
Chemistry Department, BrookhaVen National Laboratory, Upton, New York 11973, and Chemistry Department
and Radiation and Solid State Laboratory, New York UniVersity, New York, New York 10003
ReceiVed: April 14, 2008; ReVised Manuscript ReceiVed: July 2, 2008
2 2 2 2
Pulse radiolysis and flash photolysis are used to generate the hyponitrite radicals (HN /N O
O • •-) by one-
electron oxidation of the hyponitrite in aqueous solution. Although the radical decay conforms to simple
second-order kinetics, its mechanism is complex, comprising a short chain of NO release-consumption steps.
•
-
In the first, rate-determining step, two N
2
O
2
radicals disproportionate with the rate constant 2k ) (8.2 (
7
-1 -1
2-
0
.5) × 10 M
s
(at zero ionic strength) effectively in a redox reaction regenerating N
2
O
2
and releasing
two NO. This occurs either by electron transfer or, more likely, through radical recombination-dissociation.
•
-
-
Each NO so-produced rapidly adds to another N
2
O
2
2
, yielding the N
3
O
3
ion, which slowly decomposes at
radical protonates with pK ) 5.6 ( 0.3. The neutral
2
radical decays by an analogous mechanism but much more rapidly with the apparent second-order
-
1
-
•-
3
00 s to the final N
HN
O •
rate constant 2k ) (1.1 ( 0.1) × 10 M s . The N
2
O + NO
2
products. The N
O
2
a
2
9
-1 -1
•-
2
O
2
radical shows surprisingly low reactivity toward
•-
6
7
-1 -1
O
2
and O
2
, with the corresponding rate constants below 1 × 10 and 5 × 10 M s . The previously
into N O and O does not occur. The thermochemistry of HN
is discussed in the context of these new kinetic and mechanistic results.
•-
•-
•
reported rapid dissociation of N
N O
2 2
2
O
2
2
2 2
O /
•-
Introduction
The hyponitrite radical anion (N2O2•-) and its conjugate acid
•
(
HN2O2 ) can be obtained in aqueous solution either by one-
2
-
electron oxidation of hyponitrite (N2O2 ; hence, the radical
name) or by one-electron reduction of nitric oxide
). In the latter case, the precursor can be either nitroxyl ( HNO,
singlet ground state) or nitroxyl anion ( NO , triplet ground
1
2-4
(Figure
1
1
3
-
state). Spin prohibition in their acid-base equilibration gives
•
•-
rise to two independent pathways leading to HN2O2 /N2O2
;
although NO is added in both pathways, the rates differ
4-7
dramatically. It is due to these complications that considerable
confusion has persisted in the literature for over three decades
concerning the fundamental properties of the aqueous hyponitrite
radical derived from pulse radiolysis experiments involving
reduction of NO. Recently, we have suggested a major revision
that includes the radical absorption spectra, acidity, redox
properties, and stability with respect to dissociation back into
Figure 1. Formation pathways and absorption spectra of the hyponitrite
radicals.
1,4
•
environmental processes. We have estimated that HN2O2 /
N2O2• are both strongly oxidizing and moderately reducing,
which engenders the expectation of rich chemistry. At the same
time, the reactivity of these radicals remains largely unexplored.
In this work, we use pulse radiolysis and flash photolysis to
generate the hyponitrite radicals and investigate their decay
mechanism, which occurs through a surprisingly complex redox
disproportionation.
-
1
nitric oxide and nitroxyl.
The hyponitrite radicals are also known in the gas phase
8
,9
10-13
•-
and in solid argon and neon.
Although both cis-[ONdNO]
and trans-[ONdNO]• have been generated from various
-
precursors and stabilized toward spontaneous isomerization in
the cryogenic inert-gas matrixes, theory predicts lower energy
for the trans isomers of both HN2O2• and N2O2
.
•- 1,14-17
Experimental Section
Considering these results, it is reasonable to assume that the
radicals prepared by the one-electron oxidation of hyponitrite,
which has a trans configuration, will retain the parent’s trans
structure, as shown in Figure 1.
Lying between NO and N2O in terms of the nitrogen oxidation
state, the hyponitrite radical is a potentially important species
in the redox transformations of nitrogen in biological (nitrifica-
tion/denitrification, assimilatory nitrite reduction) and aquatic
Sample Solutions. Analytical-grade buffers, HClO4, and
NaOH and Milli-Q purified (ASTM type I) water were used
throughout. Sodium trans-hyponitrite (Na2N2O2 ·xH2O from
1
Aldrich, x ) 3.0 ( 0.2) was used as received. Its alkaline (pH
12-13) or acidic (pH 2-3) stock solutions were prepared daily
and kept on ice. The hyponitrite concentrations were assayed
spectrophotometrically at the 248 nm absorption maximum of
2
-
-1
-1 18
N2O2 using ε248 ) 6550 M cm . An aliquot of the stock
*
To whom correspondence should be addressed. Tel: 631-344-4333.
was diluted to the desired pH, saturated with N2O or its mixtures
with NO or O2 (as described in detail elsewhere ), and used for
Fax: 631-344-5815. E-mail: lymar@bnl.gov.
4
†
‡
Brookhaven National Laboratory.
New York University.
pulse radiolysis experiments within 1 h.
1
0.1021/jp803230c CCC: $40.75  2008 American Chemical Society
Published on Web 08/15/2008

8
296 J. Phys. Chem. A, Vol. 112, No. 36, 2008
Poskrebyshev et al.
Kinetic Measurements. The pulse radiolysis was carried out
with 2 MeV electrons from a Van de Graaff accelerator; pulse
widths were in the range of 0.06-0.8 µs. Either one or three
passes of analyzing light through a 2 cm long quartz cell were
used in the detection optical path. For the kinetics recorded on
a time scale of less than 100 µs, the analyzing Xe-arc light
source was pulsed. All experiments were done with temperature
stabilization at 25 ( 0.5 °C. The radiation yields of radicals
(
here and throughout in a number of radicals per 100 eV of
•
•
absorbed energy) were taken as G(OH ) ) 6.1 and G(H ) )
0
•
•-
.6 in N2O-saturated solutions, G(OH ) ) 2.8 and G(O2 ) )
•
•-
3
)
.4 in O2-saturated solutions, and G(OH ) ) 6.0 and G(O2
0.6 in solution saturated with a 9:1 or 4:1 N2O to O2 gaseous
)
mixture. Dosimetry was performed with a N2O-saturated 10 mM
4
-1
-1
KSCN solution using Gε ) 4.87 × 10 ions (100 eV)
M
-
1
•-
cm for the (SCN)2 radical at 472 nm. Continuous gamma
irradiation was performed in a well-type Co-60 source with
about 2 s accuracy of exposure at a 110 Gy/min dose rate.
Kinetic modeling was carried out with the INTKIN software
developed at the Brookhaven National Laboratory by H. A.
1
9
Schwarz.
Flash photolysis was done under O2 saturation in a flow cell,
as described elsewhere.
mixed with buffer also containing S2O8 prior to entering the
cell. A 308 nm excimer laser was employed to cleave S2O8
5
-7
2-
An alkaline solution of N2O2 was
Figure 2. Spectral changes during the hyponitrite radical decay in
acidic (pH 3, top) and alkaline (pH 12, bottom) solutions. Note the
difference in time scales and the appearance of species absorbing near
380 nm in alkali.
2
-
2
-
•
-
into a pair of SO4 radicals, which then were used to oxidize
hyponitrite.
the N2O2• decay. The decay traces at 290 nm can be very well
fitted to second-order kinetics
-
Results
Decay Kinetics. Previously, we have shown that all primary
•
-
•
A0 - A
radicals from water radiolysis (OH , eaq , and H ) can be
employed to generate the hyponitrite radical in a N2O-saturated
solution of hyponitrite under both alkaline and acidic condi-
f
At )
+ Af
(5)
1 + 2kapp
[
R × t
]
0
in both acid and alkali (Figure 3), where A0 and Af are the initial
and final absorbances, respectively, and the third fitting param-
eter 2kapp[R]0 yields the apparent second-order rate constant,
kapp, if the initial concentration of radicals can be evaluated from
A0 and the light path l, that is, [R]0 ) A0/ε290l. Using the values
1
-
•
tions. Omitting rapid reactions converting eaq and H into
•
20
OH , the principal reactions producing the hyponitrite radicals
were
2
2
-
•
•-
-
N O + OH f N O + OH
2
2
2
•-
of molar absorptivities ε290(N2O2 ) ) 5900 ( 300 and
•
-1
-1
k ) 8.8 × 10 M s-1 (1)
9
-1
ε290(HN O ) ) 3100 ( 300 M cm measured in the previous
2
2
1
1
work and varying the radiation dose, we have obtained a good
-
•
•-
HN O2 + OH f N O + H O
linearity of 2kapp[R]0 with [R]0 (Figure 3), from which 2kapp)
2
2
2
2
8
-1 -1
•-
(
5.4 ( 0.3) × 10 M
s
for the N2O2 radical at a 1 M
k ) 4.2 × 10 M s-1 (2)
9
-1
9
-1 -1
2
ionic strength and 2kapp ) (2.2 ( 0.2) × 10 M
s
for the
HN2O2• radical, with most of the error coming from the
uncertainties in molar absorptivities. It would thus appear that
the radicals decay through simple recombination reactions
•
•
2
H N O + OH f HN O + H O
2
2
2
2
2
k ) 4 × 10 M s-1 (3)
7
-1
3
•
-
•-
with their relative contribution governed by the acid dissociation
N O + N O f products
(6)
(7)
2
2
2
2
-
21
constants pKa(H2N2O2) ) 7.2 and pKa(HN2O2 ) ) 11.5. We
also observed a characteristic UV absorption spectrum of the
hyponitrite radical (Figure 1) and interpreted its changes with
pH as the evidence for the acid-base equilibrium
•
•
2
HN O + HN O f products
2
2
2
This formulation is supported by the observed independence
of 2kapp on the ionic strength in acidic media and by the
Debye-H u¨ ckel dependence corresponding to a reaction between
the two singly charged species in alkali (Figure 4, inset).
The second-order decay fits remain good throughout the pH
region from 2.5 to 12, and the pH dependence of the 2kapp/εapp
value is shown in Figure 4. Its titration curve appearance is in
accord with the equilibrium in reaction 4. Introducing, for
generality, a cross-recombination reaction
•
+
•-
HN O a H + N O
2
(4)
2
2
2
for which a pK4 of 5.5 was obtained. Both N2O2 and HN2O2•
exhibit maxima in the near-UV region (Figure 1), and their
decay is best observed around 290 nm, where interference from
both the absorption and bleaching of the starting material is
negligible.
1
•-
As shown in Figure 2, the decay in acidic solution occurs
monotonically at all wavelengths and is much more rapid than
the decay in alkaline solution, where the appearance of a long-
lived intermediate is observed at around 380 nm concurrent with
•
-
•
N O + HN O f products
(8)
2
2
2
2
we obtain for the pH dependence

Disproportionation of Hyponitrite Radical
J. Phys. Chem. A, Vol. 112, No. 36, 2008 8297
2
4
+ 2
+ 2k K4[H+
2
kapp
2k K + 2k7[
H
]
]
SCHEME 1: Alternative Recombination Mechanisms
Yielding the Stoichiometry of Reaction 12
6
8
a
)
(9)
•-
2
•
2
+]
+]
εapp
HN O )[
[
ε
{ (
N O
)
K + ε
(
H
}(K + H
)
2
4
2
4
The molar absorptivities are known (Figure 1) and the magni-
tudes of k6 and k7 are set by the plateau regions, leaving K4 and
k8 as the only fitting parameters, and the fit with pK4 ) 5.6 and
8
-1 -1
k8 ) 5 × 10 M
s
is shown in Figure 4. Within its estimated
uncertainty of (0.3, this pK4 value is the same as that
determined previously from the absorption spectra. Because
1
k7 is so much larger than k6, the fit is extremely insensitive to
k8 as long as it is kept below k7, but k8 can be decreased to zero
a
Initial and final species are in blue, and intermediates are in red.
without critically spoiling the fit. It is however reasonable that
k8 should lie somewhere between k6 and k7.
Although we have so far treated the hyponitrite radical decay
in terms of simple recombination reactions, it will become clear
in the following section that the apparent bimolecular rate
constants determined from the data in Figures 3 and 4, while
related to reactions 6-8, are not their actual rate constants.
Intermediates and Mechanism. As seen in Figure 2 for the
alkaline solution, an intermediate whose absorption rises
concurrently with N2O2• decay, peaks and then decays is
observed between 330 and 420 nm. We will now show that
-
-
this intermediate can be identified as the N3O3 anion, and its
formation and decay can be described by the following
mechanism
•-
•-
2-
N O + N O f N O + 2NO
rate-limiting
2
2
2
2
2
2
(
6a)
•
-
-
N O + NO f N O
3
rapid
slow
(10)
(11)
2
2
3
-
-
Figure 3. Observed rate constants, kapp, for the hyponitrite radical decay
N O f N O + NO
3 3 2 2
obtained by fitting the transient absorptions at 290 nm, A
order rate law in eq 5. Upper panel: alkaline solution containing 10
mM NaOH and 1.64 (O) or 3.9 mM (0) Na at a 1 M ionic
strength (NaClO ). Lower panel: acidic solution at pH 2.7-3.3 adjusted
with HClO and containing 0.5 (O) or 1 mM (0) of Na . The
t
, to the second-
where reaction 6a is rate-limiting for the formation of N3O3-.
This mechanism, depicted in Scheme 1A, effectively amounts
2
2 2
N O
•-
4
to redox disproportionation of the N2O2 radical. Beginning
N
2
O
2
•
4
2
with OH , the overall stoichiometry is
straight lines give the linear fits corresponding to the bimolecular rate
8
-1 -1
9
-1 -1
•
2-
-
-
constants 2kapp ) 5.4 × 10 M
s
(upper) and 2.2 × 10 M
s
4
OH + 3N O f 4OH + 2NO + 2N O (12)
2 2 2 2
(
lower). The insets show typical transient absorption decay kinetics
From the decay of N2O22 absorption in a N2O-saturated
-
and their second-order fits (red lines) under both alkaline and acidic
conditions.
solution under continuous radiolysis with a Co-60 gamma
2
-
source, we have obtained ∼0.7 for the ratio of N2O2
•
consumption to OH production rates (Supporting Information
Figure S1), which is close to the 3/4 ratio prescribed by the
stoichiometry in reaction 12.
-
Previously, the N3O3 anion has been observed in the pulse
2
-4
radiolysis of aqueous NO solutions,
where the anion is
produced as a result of NO reduction by the hydrated electron
followed by concatenation of two more NO radicals, that is,
the reaction sequence
eaq- + 3NO f NO + 2NO f N O + NO f N O3
3
-
•-
-
2
2
3
(
13)
Figure 4. Dependence upon pH of the observed rate constant for the
decay of the hyponitrite radical through reactions 6-8. The values of
The absorption spectrum of the so-created “authentic” N3O3-
anion with its distinct peak at 380 nm is reproduced as a line in
Figure 5. At sufficient concentration of NO, reaction sequence
3 is complete on a microsecond time scale, following which
the N3O3 anion decays exponentially through reaction 11 with
2kapp/εapp were obtained by fitting the transient absorptions to the second-
order rate law in eq 5. The dashed curve corresponds to eq 9 (see
text for equation parameters that were used). Inset shows the ionic
1
strength, µ, dependence of k in the alkaline region (pH > 9) and
6
-
1/2
its Debye-H u¨ ckel fit (line): log 2k
µ ), with the limiting value 2k
6
1
6
) log 2k
6
(µf0) + 1.02 µ /(1
8
3
,4
1
/2
a characteristic time of 3.3 ms in neutral solutions. As with
the “authentic” N3O3 , the intermediate in the recombination
+
M
(µf0) ) (1.64 ( 0.10) × 10
-
-1
-
s
.

8
298 J. Phys. Chem. A, Vol. 112, No. 36, 2008
Poskrebyshev et al.
-
3
ε
(
N O
)
Amax
(
τ6′ + tmax)2
3
)
×
+
•-
2
2
τ τ
[
N O ]0
6′ 11
2
τ - τ - t
•-
2
11
6′
max
ε
(
N O
)
(14)
2
τ11
where [N2O2•-]0 and ε(N2O2•-) are the initial concentration and
molar absorptivity for the hyponitrite radical and τ6′ )
1
/(2k6′[N2O2• ]0) and τ11 ) 1/k11 are the characteristic times of
-
reactions 6′ and 11, respectively. This equation reflects the fact
-
•-
that N3O3 and N2O2 are the only species in the system
absorbing between 330 and 420 nm where the maxima are
observed in the transient kinetics. Application of eq 14 to these
-
Figure 5. Absorption spectra of the N
3
O
3
transient obtained through
-
4
NO reduction with eaq (reaction 13, solid line), from the absorption
maxima on the transient kinetics (Figure 6) for a N O-saturated solution
using eq 14 (blue up triangles), and from the initial absorption
2
2-
•-
1
of N O
2 2
kinetics with the previously determined N2O2 spectrum and
2-
-
4
•-
(
Figure 6) for a N
O
2 2
solution saturated with a 99:1 N
2
O/NO mixture
N3O3 lifetime (3.3 ms) and with the values for [N2O2 ]0 and
τ6′ independently measured at 290 nm at the prevailing ionic
strength yields a spectrum that closely matches “authentic”
(
reactions 10 and 11, red down triangles).
-
N3O3 (Figure 5).
The same spectrum was obtained when a small amount of
2
-
NO was added to the solution of N2O2 prior to irradiation
Figure 5). This was done by bubbling a mixture of N2O and
(
NO (99:1), which yields a solution 19 µM in NO. This
concentration is too low for NO to compete with N2O for the
2
-
•
1,4
hydrated electron or with N2O2 for the OH radical, but it
•-
is sufficient to efficiently compete with recombination of N2O2 ,
assuring complete scavenging of this radical via reaction 10.
As shown in Figure 6, the 380 nm intermediate absorbance
appears promptly after the pulse, and its amplitude more than
doubles that obtained in the absence of NO because now every
N2O2• is intercepted through reaction 10. The intermediate
decay is purely first-order. The rate of this decay and the
spectrum of initial absorbance identify the intermediate as
Figure 6. Typical kinetic traces recorded at 380 nm following pulse
radiolysis (dose of 2.8 Gy or 0.29 µM of radicals per unit G value) of
-
a 1.3 mM Na
Lower trace: N
integration fit to reactions 6′ and 11 with 2k6′ ) 2.7 × 10 M
2
N
2
O
2
alkaline (66 mM NaOH, µ ) 70 mM) solution.
2
O-saturated solution. The red line shows the numerical
8
-1 -1
s
and
-
2
-1
N3O3 .
k
11 ) 3.3 × 10 s ; essentially the same fit (indistinguishable without
Additional evidence for this assignment comes from the
temperature dependence for the intermediate decay rate (Figure
magnification) is produced by using reactions 6a, 10, and 11 with 2k6a
8
-1 -1
9
-1 -1
2
)
1.4 × 10 M
s
, k10 ) 5.4 × 10 M
s
, and k11 ) 3.3 × 10
-
1
-
s
. Upper trace: solution saturated with a 99:1 N
2
O/NO mixture (19
6, inset). The decay of N3O3 obtained through reaction
µM of dissolved NO). The green line gives a first-order decay fit with
sequence 13 in the NO-saturated solution is strongly activated;
2
-1
a 3.0 × 10 s rate constant. The inset shows Arrhenius dependence
13
we obtain Ea ) 15.2 ( 0.2 kcal/mol and A ) (2.8-5.6) × 10
for the decay at 380 nm observed above pH 7 in a NO-saturated solution
-
1
s
for the Arrhenius parameters. This activation energy is in
excellent agreement with the previously reported value, and
the pre-exponential factor is consistent with a simple bond
with no added Na
saturated with a 99:1 N
O-saturated 2 mM Na
activation energy E ) 15.2 kcal/mol and pre-exponential factor A )
2
N
2
O
2
(red up triangles), a 2 mM Na
O/NO mixture (green down triangles), and a
2 2 2
N O solution (crosses). The straight line gives
2 2 2
N O solution
3
2
N
2
a
breaking. Essentially the same activation is observed for the
1
3
-1
•-
3
.9 × 10
s .
3
80 nm intermediate produced from the N2O2 radical either
through reactions 6a and 10 in the absence of added NO or by
reaction 10 only in the presence of small amounts of added
NO (Figure 6, inset).
reaction has an absorption maximum at 380 nm (Figure 5) and
decays on the millisecond time scale (Figure 6).
An alternative recombination mechanism kinetically indis-
tinguishable from that in Scheme 1A and yielding the same
overall stoichiometry (reaction 12) can be envisioned, in which
Analysis of Scheme 1A can be simplified by recognizing that
9
-1 -1 4
reaction 10 (previously estimated k10 ) 5.4 × 10 M s ) is
much more rapid than reaction 6a under the prevailing condi-
tions and by replacing reactions 6a and 10 with a single
bimolecular reaction; that is, reaction 6 can be reformulated as
-
N3O3 is formed directly by the NO group transfer
•-
•-
3
-
-
N O + N O f NO + N O
3
(6b)
2
2
2
2
3
•-
•-
2-
-
3
N O + N O f 0.5N O + N O
(6′)
3
-
2-
2
2
2
2
2
2
3
2 2
followed by the coupling of two NO to regenerate N O
Considering stoichiometry, the apparent rate constant for this
3
-
3
-
2-
NO + NO f N O
(15)
2
2
reaction is related to the rate constant of reaction 6a as k6′ ≈
2k6a. The validity of this approach is illustrated in Figure 6 with
and by the decay of N3O3- via reaction 11. This scenario is
shown in Scheme 1B, and the principal difference between this
and the mechanism starting with reaction 6a is in the interme-
diacy of NO instead of NO. To distinguish between these
two cases, experiments were performed in the presence of
the help of numerical simulations. Although the kinetics due to
reactions 6′ and 11 can be integrated, the resultsan infinite
seriessis not particularly useful. However, a simple expression
3
-
-
for determining the molar absorptivity of N3O3 from the
-
amplitude (Amax) and time (tmax) of the kinetic absorption
maximum (like the one appearing in Figure 6, lower trace) can
be derived
dissolved oxygen, and the formation of peroxynitrite (ONOO )
exhibiting a characteristic UV spectrum (λmax ) 302 nm, ε )
-
1
-1 22
1670 M cm ) has been examined.

Disproportionation of Hyponitrite Radical
J. Phys. Chem. A, Vol. 112, No. 36, 2008 8299
known pKa of ONOOH and the fact that the acid form, whose
characteristic lifetime is about 1 s, is solely responsible for the
25
overall peroxynitrite decay at around neutral pH. The formation
kinetics and magnitudes of residual absorption plateaus at 30-50
ms in Figure 8 can be reproduced by numerical simulations
using reactions 6a, 10, 11, and 19 for both O2-containing
solutions (for the simulated fits, see Supporting Information
Figure S2). The computed peroxynitrite yields (expressed as
-
•-
the [ONOO ]t)50ms/[N2O2 ]t)0 ratio) are 51 and 8% for
solutions saturated with O2 and N2O/O2, respectively. The yield
close to 50% in the former case arises because the rate constants
for reactions 10 and 19 are about the same and the initial
O •
-
Figure 7. Transient absorption spectra measured at 5 µs (N
2
2
-
radical, solid black), 0.4 ms (mainly N
3
O
3
, dash-dot blue), and 20
-saturated solution
containing 50 mM persulfate and 6.6 mM hyponitrite at pH 9.2 (10
concentrations of N2O2• and O2 are also comparable. The
-
•-
ms (dash green) after laser flash photolysis of an O
2
•
-
near 10% yield for the latter case simply reflects the [O2 ]t)0/
•
-
•-
mM Borax). The inset shows kinetics at 300 (N
2
O
2
decay, black)
formation and decay, blue); note the time axis
[N2O2 ]t)0 ) 0.6/6.1 ratio of initial concentrations. We thus
-
and 380 nm (N
break.
3
O
3
conclude that the pattern of peroxynitrite formation is fully
consistent with the mechanism shown in Scheme 1A plus
reaction 19.
3
Should reaction 6b occur, its NO- product would rapidly
An assumption implicit in this analysis is the absence of
•
-
interference from direct reactions between the N2O2 radical
combine with O2 to quantitatively yield peroxynitrite in the
oxygen-saturated (1.3 mM in O2) solutions
and either O2 or O2• , for example, the electron transfers
-
5
3
-
-
k ) 2.7 × 10 M s-1
9
-1
•-
•-
NO + O f ONOO
N O + O f 2NO + O
(20)
(21)
2
16
2
2
2
2
•
2
-
•-
2-
2
(16)
N O + O2 f N O + O2
2
2
•
-
This possibility was investigated with a previously described
2 2
A comparison of the N O decay kinetics in Figure 9 shows
•-
method of generating the N2O2 radical by laser flash pho-
that this is indeed the case. Because the decay reaction is second-
order in character, radiation doses have been applied such that
very nearly equal amounts of N O are generated in all N O-
1
tolysis, in which the photochemical cleavage of persulfate yields
•
-
•-
the SO4 radicals that rapidly oxidize hyponitrite
2
2
2
saturated, N2O/O2-saturated, and O2-saturated solutions. It is
seen that the kinetics are practically indistinguishable in the first
two cases, indicating that O2 does not appreciably react with
2
8
-
f 2SO4•-
S2O
(
+hν
)
λ ) 308 nm
(17)
HN O2 + SO4•- f N O
-
•-
2
+ SO42- + H+
2
2
•-
N2O2 . For the O2-saturated solutions, where large amounts
k ) 6.4 × 10 M s-1 (18)
8
-1
•-
•-
of O2 are also produced, the N2O2 decay appears even slower
than that in the other two cases. However, this effect is only
apparent and is explicable by the simultaneous formation of
1
8
The results shown in Figure 7 clearly demonstrate that no
peroxynitrite formation is observed at around 300 nm upon the
decay of all transients. Thus, reactions 6b and 16 do not occur;
that is, Scheme 1B does not represent the actual mechanism.
Figure 7 also shows that decay of the UV band of N2O2• is
accompanied by formation of the 380 nm absorption due to
N3O3 ; the N2O2 decay follows second-order kinetics.
The validity of Scheme 1A has been examined by pulse
radiolysis of the hyponitrite solutions saturated either with pure
O2 or with a 9:1 N2O to O2 gaseous mixture. In both cases,
essentially all radiolytically generated H atoms are scavenged
-
•-
ONOO and by the excess of unreacted O2 , whose absorptions
•-
overlap with that of N2O2 . In fact, this explanation is born
out by the kinetic simulations, confirming the absence of
-
•-
•-
interference from a direct reaction between O2 and N2O2
under our experimental conditions. Specifically, a good fit can
be obtained using only reactions 6a, 10, 11, and 19; inclusion
of reactions 20 and 21 seriously spoils the fit if their rate
-
•-
6
7
-1 -1
constants exceed 1 × 10 and 5 × 10 M s , respectively.
Discussion
•
-
by O2 to produce superoxide (O2 ), but the solvated electrons
generate O2 only in the O2-saturated solution; in the N2O/O2
mixture, almost 99% of eaq is converted into OH by reaction
with N2O. If reaction 6a takes place and NO is released,
•
-
Collectively, all of the data presented are consistent with the
disproportionation mechanism that involves a short chain
(Scheme 1A), where the rate-determining step is reaction 6a
producing NO as an intermediate. Applying our previous
-
•
2
3
24
peroxynitrite can be produced by combination with superoxide
o
•-
2-
estimates for the reduction potentials E (N2O2 /N2O2 ) ) 0.96
NO + O2 f ONOO-
•
-
k ) 4.8 × 10 M s-1
9
-1
o
•-
1
19
and E (2NO/N O ) ) -0.38 V versus NHE, we calculate
2
2
that this reaction is exergonic by 1.34 eV. In contrast, recently
(19)
o
•-
2-
o
published potentials E (N2O2 /N2O2 ) ) -0.4 and E (2NO/
-
26,27
and peroxynitrite formation should be much more extensive
under oxygen saturation. This follows because the radiation
yields of a limiting reactant in reaction 19 are 2.8 (NO) and 0.6
N2O2• ) ) -0.1 V based on ab initio calculations
predict
that reaction 6a is endergonic by 0.3 eV, which appears to be
inconsistent with the fairly high reaction rate. We will thus
consider the energetics in detail.
•-
(
O2 ) for solutions saturated with O2 and N2O/O2, respectively.
This is indeed the case, as shown by the magnitudes and the
Scheme 2 shows a thermodynamic cycle composed of
acid-base and redox reactions of hyponitrite. It is clear that
with our free-energy estimates, the cycle is closed within 0.01
eV; in contrast, the theoretically computed values leave the cycle
open by 1.3 eV (30 kcal/mol). This number gives a cumulative
error of computational results because the standard free-
spectra of the persisting residual absorption in Figure 8, left
panel, which belongs to peroxynitrite. This assignment is
corroborated by the rate of the residual absorption decay and
its pH dependence (Figure 8, right panel), which are both
characteristic of peroxynitrite; the pH dependence reflects the

8
300 J. Phys. Chem. A, Vol. 112, No. 36, 2008
Poskrebyshev et al.
Figure 8. (Left panel) Kinetic traces recorded at 300 nm following pulse radiolysis (doses 26-28 Gy) of 4.8 mM Na
.2). Trace 1 (green): N O-saturated solution; trace 2 (blue): solution saturated with a 9:1 N O/O mixture; trace 3 (red): O
residual absorbances at around 50 ms correspond to approximately 10 and 50% peroxynitrite yields per N
-saturated solutions, respectively. The inset compares the spectra of residual absorbances (solid symbols) with normalized spectra of authentic,
2
N
2
O
2
2
in 5 mM Borax (pH
-saturated solution. The
for the N O/O -saturated and
9
2
2
2
•-
2
O
2
2
2
O
2
-
chemically synthesized ONOO (red and blue dashed lines) and ONOOH (brown dashed line). Circles: O
O/O -saturated alkaline solution. Triangles: O -saturated acidic (pH 3) solution. All data are normalized to a 1 cm optical path. (Right panel)
Dependence upon pH of the first-order rate constant for the decay of residual absorbance due to ONOO /ONOOH in the O
2
-saturated alkaline solution. Squares:
N
2
2
2
-
2
-saturated 2 mM
) 6.8 and k ) 1.18 s . The inset shows typical
+
+
-1
Na
2
N
2
O
2
solutions. The titration-like curve corresponds to kdecay ) k[H ]/(K
a
+ [H ]) with pK
a
decay kinetics at 300 nm and pH 6.78 (phosphate) and its exponential fit (red line).
SCHEME 3: Possible Pathways for the N2O2• + N2O2•-
-
Reaction
superacid for no apparent reason. From this analysis, the largest
error in the theoretically derived energetics is clearly associated
•
-
Figure 9. Comparison of the initial stages of the N
as observed in a 2 cm cell at 290 nm following pulse radiolysis of 4.8
mM Na in 5 mM Borax buffer (pH 9.2). Blue trace: N O-saturated
solution (radiation dose 12.5 Gy). Green trace: solution saturated with
a 4:1 N O/O mixture (dose 12.7 Gy). Red trace: O -saturated solution
dose 25.6 Gy); the black line superimposed with this trace gives a
2
O
2
radical decay
2
-
•-
with the step of oxidizing N2O2 to N2O2 . On the basis of
the electrostatics and electronic structure considerations, Houk
2
N
2
O
2
2
26
2
2
2
and co-workers argued that their computed numbers are more
reasonable then our results specifically for this step. In our view,
the failure of the ab initio calculations in the thermodynamic
(
numerical simulation fit that includes reactions 6a, 10, 11, and 19.
2
-
•-
cycle test, primarily due to the N2O2 -to-N2O2 step, makes
arguing at the electronic structure level moot and leaves our
prior estimates as the only self-consistent data set.
SCHEME 2: Thermochemical Cycle Involving
Hyponitrite Species and Their Radicals in Aqueous
a
Solution
It is perhaps not surprising that despite a very large driving
force, reaction 6a between two radicals occurs with a rate that
is some 2 orders of magnitude below the diffusion control
because the reaction involves bond(s) breaking and, possibly,
making. Probable mechanisms for this reaction are shown in
Scheme 3, where pathways A and B represent intermolecular
electron transfer occurring in a collision complex. In the
adiabatic pathway B, an electron is transferred simultaneously
with the N-N bond breaking, which would require large nuclear
reorganization. For pathway A, a significant nuclear reorganiza-
tion energy is also expected due to drastic differences between
the N-N bond lengths and N-N-O angles in reactants and
products (2.26 Å and 97°, 1.36-1.40 Å and 111°, and 1.26 Å
a
The numbers correspond to free-energy changes (in eV/molecule,
1
M standard states, NHE reference) in the arrow directions. Entries in
roman are our estimates; italicized entries in parentheses are theoretical
results by Houk and co-workers;
the experimental pK values for hyponitrous acid.
1
26,27
the number in bold is derived from
a
and 112° for N2O2, N2O2• , and N2O2 , respectively
-
2-
16,17,29,30
).
energy change of 1.12 eV associated with full deprotonation
In addition, the nascent oxidized product for this pathway, trans-
of H2N2O2 is set by the measured pKa(H2N2O2) ) 7.2 and
N2O2 dimer, is unstable with respect to two separated NO; in
the gas phase, the free-energy difference is 0.6-0.8 eV,
-
21
28,31-33
pKa(HN2O2 ) ) 11.5.
The dashed diagonal arrow in Scheme 2 corresponds to
which makes pathway A less exoergic than pathway B. The
third possibility does not involve electron transfer at all and
could occur via a radical dimerization, followed by homolytic
eliminations of the NO end groups (case C). However, our data
give no evidence, either spectral or kinetic, for an intermediate,
so that if the reaction does occur through pathway C, the steady-
•
deprotonation of the HN2O2 radical, and its pKa ) 5.6 derived
from the data in Figure 4 is consistent with our free-energy data.
2
6,28
With the computed values
which side of the cycle is used and comes out as either 3.3 or
18 for the right and left side, respectively. The -18 number
however, this pKa depends on
-
•
2-
is particularly troublesome for it would make HN2O2
a
state concentration of the N4O4 intermediate has to be small

Disproportionation of Hyponitrite Radical
J. Phys. Chem. A, Vol. 112, No. 36, 2008 8301
HN O f N O + HNO (23)
has been reported along with a pKa of 3.1 for HN3O3.2,3 Our
due either to its short lifetime or low equilibrium constant for
its formation, or both. The absence of oxygen interference with
3
3
2
2
•-
the N2O2 decay on a millisecond time scale, evident in Figure
, implies that reactions 20 and 21 are slow despite their
observations are in general agreement with these results. Despite
9
•
the increased rate of the HN2O2 recombination (Figure 4), we
o
considerable driving force; we estimate ∆r20G ≈ -5 kcal/mol
-
no longer observe the formation kinetics for HN3O3/N3O3 ; only
o
and ∆r21G ≈ -26 kcal/mol. The assumption of large nuclear
vestiges of their decay can be detected (Figure 2, upper panel).
The decay rate sharply increases below pH 4 in a manner
consistent with the reported pKa and k23 (Supporting Information
Figure S3). However, we do observe formation of peroxynitrous
acid, ONOOH, in the quantitatively correct amounts when
oxygen is present and superoxide is generated (Figure 8, left
panel inset), which implies the intermediacy of NO. On the basis
reorganization barriers for both reduction and oxidation of
•
-
N2O2 would be in qualitative accord with their low rates. In
view of this discussion, we are inclined to favor the radical
recombination pathway C in Scheme 3.
It has been suggested that the hyponitrite radical is unstable
toward unimolecular decomposition
•
of this fact, we believe that the recombination of HN2O2 occurs
ONNO•- f N O + O
•-
(22)
2
by a mechanism analogous to that shown in Scheme 1A, that
is, the rate-limiting radical disproportionation
-1 3
with the rate constant k22 ) 350 s . Thermodynamically, this
reaction is reasonable; we calculate it to be downhill by 8 kcal/
•
2
•
2
mol in free energy. However, our kinetic data on the N2O2•
-
HN O + HN O f H N O + 2NO
(7a)
2
2
2
2
2
decay show no first-order component of this magnitude in the
essentially pure second-order process, as evidenced by a
negligible intercept in the upper panel of Figure 2. Moreover,
is followed by a rapid combination reaction
•
2
HN O + NO f HN O
(24)
2
3
3
•
the occurrence of reaction 22 would regenerate OH (pKa ≈
and finally by the HN3O3 decay in reaction 23. If k24 . k7a, the
rate constant k7a is 1/4 of the slope in the lower panel in Figure
3
4,35
1
2),
triggering a chain decomposition of hyponitrite. This
effect should be especially prominent under continuous, low-
intensity radiation when the self-recombination pathways for
8
-1 -1
3
, or k7a ) 5.5 × 10 M s .
•
-
N2O2 are minimized. And yet, our gamma radiolysis experi-
ment gives a hyponitrite decomposition radiation yield that is
Acknowledgment. Research at Brookhaven National Labo-
ratory was carried out under the auspices of the U.S.
Department of Energy under Contract DE-AC02-98CH10886
from the Division of Chemical Sciences, Office of Basic
Energy Sciences. Helpful comments from Drs. Norman Sutin
and Harold Schwarz are appreciated.
•
below the yield for OH and is in accord with the stoichiometry
of reaction 12 (Supporting Information Figure S1). Thus,
reaction 22 does not occur with a detectable rate. The most
apparent reason for this is that the reaction is not merely
dissociation along the N-O bond; a very large nuclear
reorganization is required to form a linear N2O product from a
nonlinear reactant. An instructive analogy here is the unimo-
lecular dissociation of monoprotonated hyponitrite anion, HON-
Supporting Information Available: Details of kinetic
simulations (tabulated rate constants, spectral properties, and
radiation yields); gamma radiolysis results; expanded view of
kinetic fits for Figure 8; and pH dependence of HN3O3/N3O3
decay. This material is available free of charge via the Internet
at http://pubs.acs.org.
-
-
o
NO , to yield N2O and OH . Although more exergonic (∆G
-30 kcal/mol) than reaction 22, this decomposition occurs
-
≈
1,36
with a 16 min half-life. Reaction 22 is probably even slower,
and it appears that previous speculations concerning the role of
N2O2 as a OH-releasing species in biological environments
should be reconsidered.
•-
37,38
References and Notes
(
1) Poskrebyshev, G. A.; Shafirovich, V.; Lymar, S. V. J. Am. Chem.
Soc. 2004, 126, 891.
2) Gr a¨ tzel, M.; Taniguchi, S.; Henglein, A. Ber. Bunsen-Ges. Phys.
The data for the solution without deliberately added NO in
Figures 5 and 6 can be used for estimating the upper limit of
reversibility in reaction 10. Specifically, simulations show that
(
Chem. 1970, 74, 1003.
-
it is not possible to maintain the molar absorptivity of N3O3
(3) Seddon, W. A.; Fletcher, J. W.; Sopchyshyn, F. C. Can. J. Chem.
1
2
973, 51, 1123.
at 380 nm within 10% of its nominal value and obtain even
remotely satisfactory fits to the absorption kinetics at this
(
4) Lymar, S. V.; Shafirovich, V.; Poskrebyshev, G. A. Inorg. Chem.
005, 44, 5212.
-
wavelength if the rate constant for decomposition of N3O3 back
(5) Shafirovich, V.; Lymar, S. V. Proc. Natl. Acad. Sci. U.S.A. 2002,
•
-
-1
99, 7340.
to N2O2 and NO exceeds some 3000 s . Thus, the dissociation
(
6) Shafirovich, V.; Lymar, S. V. J. Am. Chem. Soc. 2003, 125,
-
7
constant K-10 does not exceed 5 × 10 M (a rather conservative
6
547.
4
upper limit), in agreement with our previous estimate, and the
(7) Lymar, S. V.; Shafirovich, V. J. Phys. Chem. B 2007, 111,
6861.
o
-
free energy of formation, ∆fG , for aqueous N3O3 is below
(
(
8) Posey, L. A.; Johnson, M. A. J. Chem. Phys. 1988, 88, 5383.
9) Li, R.; Continetti, R. E. J. Phys. Chem. A 2002, 106, 1183.
7
3 kcal/mol. On the other hand, the occurrence of spontaneous
-
N3O3 decomposition via reaction 11 sets the lower limit for
∆
(
10) Jacox, M. E.; Thompson, W. E. J. Chem. Phys. 1990, 93, 7609.
(11) Andrews, L.; Zhou, M. F.; Willson, S. P.; Kushto, G. P.; Snis, A.;
o
o
-
fG , so that 18 < ∆fG (N3O3 ) < 73 kcal/mol.
Panas, I. J. Chem. Phys. 1998, 109, 177.
Most of the data presented above pertain to the deprotonated
(12) Andrews, L.; Zhou, M. F. J. Chem. Phys. 1999, 111, 6036.
(13) Lugez, C. L.; Thompson, W. E.; Jacox, M. E.; Snis, A.; Panas, I.
hyponitrite radical, primarily because information on the
recombination mechanism that could be obtained for the
J. Chem. Phys. 1999, 110, 10345.
(14) Mebel, A. M.; Morokuma, K.; Lin, M. C.; Melius, C. F. J. Phys.
Chem. 1995, 99, 1900.
•
radical’s conjugate acid (HN2O2 ) predominating below pH 5
is more limited. In the acidic pH region, considerable experi-
mental difficulties arise from drastic decreases in both the
(
15) Bunte, S. W.; Rice, B. M.; Chabalowski, C. F. J. Phys. Chem. A
1
997, 101, 9430.
-
absorption and lifetime of the N3O3 intermediate, undoubtedly
(16) Snis, A.; Panas, I. Chem. Phys. 1997, 221, 1.
(
17) Fuster, F.; Dezarnaud-Dandine, C.; Chevreau, H.; Sevin, A. Phys.
due to formation of the HN3O3 species. These effects have been
Chem. Chem. Phys. 2004, 6, 3228.
18) Polydoropoulos, C. N.; Voliotis, S. D. Anal. Chim. Acta 1968, 40,
170.
4
reported previously, and the rate constant of k23 ) 1.6 × 10
(
-
1
s
for the dissociation

8
302 J. Phys. Chem. A, Vol. 112, No. 36, 2008
Poskrebyshev et al.
(
19) The Numerical Integration method used in this program is the
(28) Dutton, A. S.; Fukuto, J. M.; Houk, K. N. Inorg. Chem. 2004, 43,
DVODE package written by P. N. Brown and A. C. Hindmarsh of Lawrence
Livermore National Laboratory, and G. D. Byrne of Exxon Research and
Engineering Co.
1039.
(
29) McKellar, A. R. W.; Watson, J. K. G.; Howard, B. J. Mol. Phys.
1
995, 86, 273.
20) In alkaline solutions, these reactions are H + OH f eaq- + H2O
-
(
(
30) Kukolich, S. G.; Sickafoose, S. M. Mol. Phys. 1996, 89, 1659.
-
-
and eaq + N2O + H2O f N2 + OH + OH. In acidic solutions, the
-
+
(31) Dkhissi, A.; Soulard, P.; Perrin, A.; Lacome, N. J. Mol. Spectrosc.
997, 183, 12.
following reactions occur: eaq + H f H and H2N2O2 + H f H2O + N2
OH. For detail, see: Poskrebyshev, G. A.; Lymar, V. J. Am. Chem. Soc.
004, 126, 891.
21) Bonner, F. T.; Hughes, M. N. Comments Inorg. Chem. 1988, 7,
15.
1
+
(32) Urban, B.; Strobel, A.; Bondybey, V. E. J. Chem. Phys. 1999, 111,
2
8
939.
(
(
(
33) Park, J. K.; Sun, H. Chem. Phys. 2001, 263, 61.
34) Hickel, B.; Corfitzen, H.; Sehested, K. J. Phys. Chem. 1996, 100,
2
(
(
22) Hughes, M. N.; Nicklin, H. G. J. Chem. Soc. A 1968, 450.
23) Buxton, G. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. B. J.
1
7186.
Phys. Chem. Ref. Data 1988, 17, 513.
(
991.
(35) Poskrebyshev, G. A.; Neta, P.; Huie, R. E. J. Phys. Chem. A 2002,
106, 11488.
(36) Hughes, M. N.; Stedman, G. J. Chem. Soc. 1963, 1239.
24) Lymar, S. V.; Poskrebyshev, G. A. J. Phys. Chem. A 2003, 107,
7
(
(
25) Goldstein, S.; Lind, J.; Merenyi, G. Chem. ReV. 2005, 105, 2457.
26) Dutton, A. S.; Fukuto, J. M.; Houk, K. N. Inorg. Chem. 2005, 44,
(
37) Ohshima, H.; Gilibert, I.; Bianchini, F. Free Radical Biol. Med.
999, 26, 1305.
38) Kirsch, M.; de Groot, H. J. Biol. Chem. 2002, 277, 13379.
1
4
024.
(
687.
(
27) Dutton, A. S.; Fukuto, J. M.; Houk, K. N. Inorg. Chem. 2005, 44,
7
JP803230C