Pressure dependence of peroxynitrite reactions. Support for a radical mechanism

Article abstract of DOI:10.1021/ic981359f

Activation volumes (ΔV^?) have been determined for several reactions of peroxynitrite using the stopped-flow technique. Spontaneous decomposition of ONOOH to NO₃- in 0.15 M phosphate, pH 4.5, gave ΔV^? = 6.0 ± 0.7 and 14 ± 1.0 cm³ mol^-1 in the presence of 53 μM and 5 mM nitrite ion, respectively. One-electron oxidations of Mo(CN)₈^4- and Fe(CN)₆^4-, which are first order in peroxypitrite and zero order in metal complex, gave ΔV^? = 10 ± 1 and 11 ± 1 CM³ mol^-1, respecfively, at pH 7.2. The limiting yields of oxidized metal complex were found to decrease from 61 to 30% of the initially added peroxynitrite for Mo(CN)₈^3- and from 78 to 47% for Fe(CN)₆^3- when the pressure was increased from 0.1 to 140 MPa. The bimolecular reaction between CO₂ and ONOO^- was determined by monitoring the oxidation of Fe(CN)₆^4- by peroxynitrite in bicarbonate-containing 0.15. M phosphate, pH 7.2 for which ΔV^?. = -22 ± 4 cm³ mol^-1. The Fe(CN)₆^3- yield decreased by ～20% upon increasing the pressure from atmospheric to 80 MPa. Oxidation of Ni(cyclam)²⁺ by peroxynitrite, which is first order in each reactant, was characterized by ΔV^? = -7.1 ± 2 cm³ mol^-1, and the thermal activation parameters ΔH^? = 4.2 ± 0.1 kcal mol^-1 and ΔS^?. = -24 ± 1 cal mol^-1 K^-1 in 0.15 M phosphate, pH 7.2. These results are discussed within the context of the radical cage hypothesis for peroxynitrite reactivity.

Full text of DOI:10.1021/ic981359f

528
Inorg. Chem. 2001, 40, 528-532
Pressure Dependence of Peroxynitrite Reactions. Support for a Radical Mechanism
John W. Coddington, Scot Wherland, and James K. Hurst*
Department of Chemistry, Washington State University, Pullman, Washington 99164-4630
ReceiVed NoVember 24, 1998
Activation volumes (∆Vq) have been determined for several reactions of peroxynitrite using the stopped-flow
technique. Spontaneous decomposition of ONOOH to NO3- in 0.15 M phosphate, pH 4.5, gave ∆Vq ) 6.0 ( 0.7
and 14 ( 1.0 cm3 mol-1 in the presence of 53 µM and 5 mM nitrite ion, respectively. One-electron oxidations
4-
of Mo(CN)8 and Fe(CN)64-, which are first order in peroxynitrite and zero order in metal complex, gave ∆Vq
) 10 ( 1 and 11 ( 1 cm3 mol-1, respectively, at pH 7.2. The limiting yields of oxidized metal complex were
3-
found to decrease from 61 to 30% of the initially added peroxynitrite for Mo(CN)8 and from 78 to 47% for
3-
Fe(CN)6 when the pressure was increased from 0.1 to 140 MPa. The bimolecular reaction between CO2 and
4-
ONOO- was determined by monitoring the oxidation of Fe(CN)6 by peroxynitrite in bicarbonate-containing
0.15 M phosphate, pH 7.2, for which ∆Vq ) -22 ( 4 cm3 mol-1. The Fe(CN)63- yield decreased by ∼20% upon
increasing the pressure from atmospheric to 80 MPa. Oxidation of Ni(cyclam)2+ by peroxynitrite, which is first
order in each reactant, was characterized by ∆Vq ) -7.1 ( 2 cm3 mol-1, and the thermal activation parameters
∆Hq ) 4.2 ( 0.1 kcal mol-1 and ∆Sq ) -24 ( 1 cal mol-1 K-1 in 0.15 M phosphate, pH 7.2. These results are
discussed within the context of the radical cage hypothesis for peroxynitrite reactivity.
Introduction
“indirect” pathway, which is first order in peroxynitrite and zero
order in the oxidizable species, indicating that rate-limiting
unimolecular activation precedes the oxidation-reduction
step.13,15-20 In contrast, two-electron oxidations appear to occur
by a “direct” pathway that is first order in both peroxynitrite
and reductant.12,16,21-24 Maximal yields of oxidized products
obtained by the indirect pathway are always significantly less
than calculated for stoichiometric conversion by the oxidant.
This property indicates the existence of at least two intermediates
along the reaction pathway, only one of which undergoes
bimolecular reactions, i.e., Scheme 1, where S is a one-electron
reductant.
There are two prevailing opinions concerning the nature of
the reaction intermediates.6,12,25 One is that “X” and “Y” are
conformational isomers whose reactivities differ from each other,
and the other is that “X” and “Y” are caged radical pairs and
free radicals formed from homolytic cleavage of the O-O bond,
i.e., Scheme 2. One focal point in the debate has been the
volume of activation (∆Vq) measured for the decomposition of
ONOOH. This was originally measured to be ∆Vq = +1.7 cm3
Recognition that peroxynitrite may be both a major patho-
genic agent in human diseases associated with oxidative stress1
and a natural microbicidal agent2-5 has triggered a renaissance
of interest in its chemical properties.6 This powerful oxidant,
which is thought to be formed by coupling of NO• and O2•- in
tissues or cells that simultaneously generate these radicals, reacts
by a relatively complex mechanism whose fundamental nature
is being actively debated.6
The peroxynitrite anion (ONOO-) is relatively unreactive,
but upon protonation1 or addition of Lewis acids7-11 (including
CO2 and certain metal ion complexes) it oxidizes susceptible
molecules by one- and/or two-electron processes12 or, in the
absence of reacting partners, it rapidly decomposes to form
NO3- as the stable product in acidic media.1,13,14 Kinetic studies
have shown that most one-electron oxidations occur by an
* Address correspondence to this author. E-mail: hurst@wsu.edu. Fax:
(509) 335-8867.
(1) Beckman, J. S.; Koppenol, W. H. Am. J. Physiol. Cell Physiol. 1996,
271, C1424-C1437, and references therein.
(2) Evans, T. J.; Buttery, L. D. K.; Carpenter, A.; Springall, D. R.; Polak,
J. M.; Cohen, J. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 9553-9558.
(3) Hurst, J. K.; Lymar, S. V. Acc. Chem. Res. 1999, 32, 520-528.
(4) Denis, M. J. Leukocyte Biol. 1994, 55, 682-684.
(5) Kuhl, S. J.; Rosen, H. Curr. Opin. Infect. Dis. 1995, 8, 181-185.
(6) Various viewpoints have been summarized in invited papers published
collectively as a discussion forum in Chem. Res. Toxicol. 1998, 11,
709-721.
(15) Lymar, S. V.; Jiang, Q.; Hurst, J. K. Biochemistry 1996, 35, 7855-
7861.
(16) Goldstein, S.; Czapski, G. Inorg. Chem. 1995, 34, 4041-4048.
(17) Goldstein, S.; Czapski, G. Inorg. Chem. 1997, 36, 5113-5117.
(18) Crow, J. P.; Spruell, C.; Chen, J.; Gunn, C.; Ischiropoulos, H.; Tsai,
M.; Smith, C. D.; Radi, R.; Koppenol, W. H.; Beckman, J. S. Free
Radical Biol. Med. 1994, 16, 331-338.
(19) Goldstein, S.; Squadrito, G. L.; Pryor, W. A.; Czapski, G. Free Radical
Biol. Med. 1996, 21, 965-974.
(7) Lymar, S. V.; Hurst, J. K. Chem. Res. Toxicol. 1998, 11, 714-715.
(8) Lymar, S. V.; Hurst, J. K. J. Am. Chem. Soc. 1995, 117, 8867-8868.
(9) Uppu, R. M.; Squadrito, G. L.; Pryor, W. A. Arch. Biochem. Biophys.
1996, 327, 335-343.
(20) Beckman, J. S.; Beckman, T. W.; Chen, J.; Marshall, P. A.; Freeman,
B. A. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 1620-1624.
(21) Radi, R.; Beckman, J. S.; Bush, K. M.; Freeman, B. A. J. Biol. Chem.
1991, 266, 4244-4250.
(10) Groves, J. T.; Marla, S. S. J. Am. Chem. Soc. 1995, 117, 9578-9579.
(11) Stern, M. K.; Jensen, M. P.; Kramer, K. J. Am. Chem. Soc. 1996,
118, 8735-8736.
(22) Pryor, W. A.; Jin, X.; Squadrito, G. L. Proc. Natl. Acad. Sci. U.S.A.
1994, 91, 11173-11177.
(12) Pryor, W. A.; Squadrito, G. L. Am. J. Physiol. 1995, 268, L699-
(23) Radi, R. Chem. Res. Toxicol. 1998, 11, 720-721.
(24) Goldsein, S.; Meyerstein, D.; van Eldik, R.; Czapski, G. J. Phys. Chem.
A 1997, 101, 7114-7118.
L722.
(13) Lymar, S. V.; Hurst, J. K. Inorg. Chem. 1998, 37, 294-301.
(14) Pfeiffer, S.; Gorren, A. C. F.; Schmidt, K.; Werner, E. R.; Hansert,
B.; Bohle, D. S.; Mayer, B. J. Biol. Chem. 1997, 272, 3465-3470.
(25) Koppenol, W. H.; Moreno, J. J.; Pryor, W. A.; Ischiropoulos, H.;
Beckman, J. S. Chem. Res. Toxicol. 1992, 5, 834-842.
10.1021/ic981359f CCC: $20.00 © 2001 American Chemical Society
Published on Web 01/04/2001

Pressure Dependence of Peroxynitrite Reactions
Inorganic Chemistry, Vol. 40, No. 3, 2001 529
Scheme 1
Scheme 2
Figure 1. Pressure dependence of the first-order rate constant for
decomposition of 0.35 mM ONOOH at 19.5 °C in 0.15 M phosphate
-
buffer, pH 4.5, containing 53 µM (filled squares) and 5.0 mM NO2
(open squares). Error bars are equal to twice the standard deviation of
four measurements at each pressure; individual rate constants are listed
in Table S1. The lines are linear least-squares fit to the data points.
mol-1 using a stopped-flow pH-jump technique at pH values
ranging from 5.6 to 6.2 and temperatures of 15-23 °C. It was
argued that this value is too small to accommodate a mechanism
that involves O-O bond homolysis to form a radical pair, but
is consistent with an internal rearrangement mechanism.26
However, pressure-dependent pulse radiolysis experiments,
made in 0.15 M formate or low phosphate (4 mM) buffers, have
given a self-consistent value for ONOOH decomposition of ∆Vq
= +10 cm3 mol-1 over the pH range 4.0-7.2.24,27 As discussed
by these researchers,27 this value is consistent with a radical
mechanism. They also demonstrated that ∆Vq was phosphate-
dependent in the pH range 5.6-6.2 and dropped to 1-2 cm3
mol-1 when the phosphate concentration was raised to 0.10 M;27
although the phosphate dependence accounted for the apparent
discrepancy in the two studies, the molecular basis for this effect
is obscure.
In this paper, we report results from our extensive studies of
the pressure dependence of ONOOH reactions measured by
stopped-flow spectrophotometry. These data support the radical
mechanism, provide evidence for discrete ONOOCO2- adduct
formation in the reaction between ONOO- and CO2, and, more
generally, are diagnostic of “direct” vs “indirect” reactions of
peroxynitrite.
Kinetic Measurements. Rate measurements at ambient pressure
were made using a Hi-Tech SF-40 stopped-flow spectrophotometer;
high-pressure measurements were made using a Hi-Tech HPS-2000
syringe drive unit contained within a high-pressure vessel, as previously
described.33 Both instruments were interfaced to a Hi-Tech SF-40C
spectrophotometer control unit. One drive syringe contained 0.3 M
phosphate buffer plus desired reductants and/or bicarbonate ion; the
other contained the peroxynitrite stock solution that had been ap-
propriately diluted with 10 mM NaOH. In cases where formation of
the ONOOCO2- adduct was to be avoided, the buffer was sparged with
N2 immediately before reaction to remove any accumulated CO2. The
final pH was determined both by simulating the mixing conditions
outside the stopped-flow apparatus and by measuring the pH of the
product solution obtained from the waste syringe. These measurements
gave values that were within 0.1 pH unit agreement, and measurements
indicated that the buffered reactant solution increased by only 0.1-
0.2 pH unit upon mixing with alkaline peroxynitrite. For the high-
pressure experiments, the reaction temperature was monitored with a
platinum thermistor which was located inside the vessel. Pressures were
performed in random order for cases where data for multiple pressures
were obtained with a single filling of the syringes. The interval between
the preparation of the reactant solutions and the collection of the last
data point was always less than 40 min, during which time decomposi-
tion of peroxynitrite was <5%. Kinetic data, collected for a minimum
of 3-4 half-lives, followed a single-exponential decay profile. Peroxy-
nitrite decomposition in the absence of oxidizable metal complexes
was monitored at 280 nm; reactions with Mo(CN)84-, Fe(CN)64-, and
Ni(cyclam)2+ were monitored at 388 nm (ꢀ388 ) 1200 M-1cm-1),34 420
nm (ꢀ420 ) 1000 M-1cm-1),35 and 360 nm (ꢀ360 ) 5100 M-1cm-1),36
respectively, corresponding to the absorption maxima of the oxidized
products. Product yields, calculated from the measured total absorbance
changes at these wavelengths, are expressed as the amount of oxidized
complex per total amount added peroxynitrite.
Experimental Section
Materials. All chemicals were reagent grade and were used as
received from commercial sources unless otherwise noted; water was
purified using a Milli-Q system. Stock solutions of peroxynitrite were
prepared from acidified hydrogen peroxide and sodium nitrite using a
tandem mixing apparatus,28 and then stored at -80 °C. Sodium nitrite
was present in a 10% excess to minimize contamination by unreacted
H2O2. The stock solution contained ∼140 mM peroxynitrite, determined
spectrophotometrically (ꢀ302 ) 1670 M-1cm-1),29 and 21 mM residual
nitrite, determined using the Greiss assay.30 The metal-complex salts,
K4Mo(CN)8 and Ni(cyclam)(ClO4)2, were prepared using procedures
described in the literature.31,32
Results
Decomposition of ONOOH. At pH 4.5, where virtually all
of the peroxynitrite is protonated (pKa ) 6.8), the first-order
rate constant for ONOOH decomposition decreased with
increasing pressure; typical results are plotted in Figure 1.
Activation volumes calculated from the magnitude of the change
are ∆Vq ) 6.0 ( 0.7 cm3 mol-1 for solutions containing no
(26) Kissner, R.; Nauser, T.; Bugnon, P.; Lye, P. G.; Koppenol, W. H.
Chem. Res. Toxicol. 1997, 10, 1285-1292.
(27) Goldstein, S.; Meyerstein, D.; van Eldik, R.; Czapski, G. J. Phys.
Chem. A 1999, 103, 7114-7118.
(28) Saha, A.; Goldstein, S.; Cabelli, D.; Czapski, G. Free Radical Biol.
Med. 1998, 24, 653-659.
added NO2 (residual [NO2-] ) 53 µM), and ∆Vq ) 14 ( 1
-
(29) Hughes, M. N.; Nicklin, H. G. J. Chem. Soc. A 1968, 450-452.
(30) Greenbert, A. E.; Connors, J. J.; Jenkins, D. Standard Methods for
the Examination of Water and Wastewater, 15th ed.; American Public
Health Association, United Book Press: Baltimore, 1995; pp 4-83,
4-84, 4-99.
(33) Murguia, M. A.; Wherland, S. Inorg. Chem. 1991, 30, 139-144.
(34) Perumareddi, J. R.; Liehr, A. D.; Adamson, A. W. J. Am. Chem. Soc.
1963, 85, 249-259.
(35) Jorgensen, C. K. Acta Chim. Scand. 1956, 10, 518-534.
(36) Zeigerson, E.; Ginzburg, G.; Meyerstein, D.; Kirschenbaum, L. S. J.
Chem. Soc., Dalton Trans. 1 1980, 1243-1247.
(31) Furman, N. H.; Miller, C. O. Inorg. Synth. 1950, 3, 160-63.
(32) Bosnich, B.; Tobe, M. L.; Webb, G. A. Inorg. Chem. 1965, 4, 1109-
1112.

530 Inorganic Chemistry, Vol. 40, No. 3, 2001
Coddington et al.
Figure 2. Pressure dependence of the first-order rate constant for the
reaction in 0.15 M phosphate, pH 7.2, between 0.5 mM ONOOH and
5.0 mM Mo(CN)84- (filled squares) at 21 °C or 5.0 mM Fe(CN)64- at
25 °C (open squares) and percent yields of Mo(CN)83- (filled circles)
and Fe(CN)63- (open circles). Individual data points are listed in Table
S2. The lines are the results of linear least-squares analyses.
Figure 3. Pressure dependence of the pseudo-first-order rate constant
for the reaction of 70 µM ONOOH with 1.0 mM Ni(cyclam)2+ at 20
°C (open circles) and the first-order rate constant for reaction of
-
ONOOCO2 (0.3 mM peroxynitrite, 6 mM total carbonate) with 5.0
4-
mM Fe(CN)6 at 25 °C (filled circles). Both reactions were done in
0.15 M phosphate, pH 7.2. Error bars equal to twice the standard
deviation are given for pressures where more than three determinations
cm3 mol-1 for solutions containing 5 mM NO2-. The rate
constant, measured at atmospheric pressure, was identical to
previously determined values (k ) 0.8 s-1 at 19.5 °C) and was
invariant within (10% to NO2- concentrations measured over
the range 53 µM-10 mM. Previously determined thermal
activation parameters for peroxynitrite decomposition at pH 4
were made. Individual data points are listed in Tables S3 (ONOOCO2-
)
and S4 (Ni(cyclam)2+). The lines are the results of linear least squares
analyses.
the thermal activation parameters of ∆Hq ) 4.2 ( 0.1 kcal
mol-1 and ∆Sq ) -24 ( 1 cal mol-1 K-1 (Figure S3). The
rate constant also increased with pressure (Figure 3), for which
∆Vq ) -7.1 ( 2 cm3 mol-1 (Figure 3). The yield of oxidized
product was independent of pressure, within experimental
uncertainty, and agreed with the value reported earlier for the
reaction at atmospheric pressure, i.e., 70-90%.16
are ∆Hq ) 21 kcal mol-1 and ∆Sq ) -13 cal mol-1 K-1 37
.
Oxidation of Cyanide Complexes of Mo and Fe. One-
electron oxidations of Mo(CN)84- and Fe(CN)64- are first order
in ONOOH and zero order in metal complexes; the observed
rate constants are identical to those for ONOOH self-
decomposition.13,16 These properties are diagnostic of the indirect
pathway. The maximum yields of the oxidized metal complexes
at atmospheric pressure are slightly less than 80% (Figure 2),16
which correspond well to recent measurements of maximal
yields for other “indirect” reactions.38,39 At pH 7.2, the observed
first-order rate constant decreased with pressure over the range
0.1-140 MPa (Figure 2), corresponding to ∆Vq ) 10 ( 1 cm3
Discussion
Radical Nature of the Reactions. Reactions of ONOOH
were initially interpreted in terms of free radical mechanisms;20,40
however, involvement of the OH• radical was subsequently
questioned on thermodynamic grounds, which led to the
development of alternative reaction models, which were based
upon differing reactivities of the cis and trans isomeric forms.12,25
Although these reaction models have been enthusiastically
received in some circles, recent studies have clearly demon-
strated the radical nature of reactions involving both ONOOH
and ONOOCO2-. Specifically, Merenyi and Lind41 and we42
have shown that the highly complex nature of ONOOH
decomposition in neutral solutions can be quantitatively simu-
lated by a radical model, the unique feature of which is that the
rate constants for all elementary steps haVe been independently
determined. This demonstration resolves the debate over whether
OH• formation is thermodynamically feasible;25,43,44 moreover,
it establishes that the radical reaction will take place under the
prevailing experimental conditions. The radical mechanism has
been criticized on the basis that the ONOOH decomposition
rate constant failed to decrease upon increasing the medium
viscosity,45 an effect often observed because increasing the
viscosity retards primarily escape of the radicals from their
solvent cage (the kD step in Scheme 2). In this particular case,
4-
mol-1 for Mo(CN)8 and ∆Vq ) 11 ( 1 cm3 mol-1 for
Fe(CN)64-. Over the same pressure range, the yield of oxidized
metal complexes also decreased proportionately from 61 to 30%
for Mo(CN)83- and from 78 to 47% for Fe(CN)63- (Figure 2).
4-
The pressure dependence of Fe(CN)6 oxidation was also
studied at pH 4.5, up to 125 MPa, with very similar results,
i.e., ∆Vq ) 12 ( 1 cm3 mol-1, with the yield decreasing as the
pressure increased from 68 to 43% (Figure S1).
4-
The oxidation of Fe(CN)6 was also studied at pH 7.2 in
phosphate-buffered solutions containing 6 mM bicarbonate ion.
Under the prevailing conditions, the CO2 concentration (0.34
mM) exceeded that of peroxynitrite by ∼10%, so the reaction
proceeded almost entirely via rate-limiting formation of
ONOOCO2-.8 In this case, the apparent first-order rate constant
increased dramatically with increasing pressure up to 82 MPa,
for which ∆Vq ) -22 ( 4 cm3 mol-1 (Figure 3). However,
3-
the Fe(CN)6 yields decreased from ∼70 to ∼50% over the
same pressure range (Figure S2).
Oxidation of Ni(cyclam)2+. A previously reported study
revealed that the rate of oxidation of Ni(cyclam)2+ by peroxy-
nitrite was first order in both reagents, indicating a direct
bimolecular reaction.16 From the temperature dependence of the
rate constant, determined over the range 5-45 °C, we calculated
(40) Mahoney, L. R. J. Am. Chem. Soc. 1970, 92, 5262-5263.
(41) Merenyi, G.; Lind, J. Chem. Res. Toxicol. 1998, 11, 243-246.
(42) Coddington, J. W.; Lymar, S. V.; Hurst, J. K. J. Am. Chem. Soc. 1999,
121, 2438-2443.
(43) Merenyi, G.; Lind, J. Chem. Res. Toxicol. 1997, 10, 1216-1220.
(44) Koppenol, W. H.; Kissner, R. Chem. Res. Toxicol. 1998, 11, 87-90.
(45) Pryor, W. A.; Jin, X.; Squadrito, G. L. J. Am. Chem. Soc. 1996, 118,
3125-3128.
(37) Merenyi, G.; Lind, J.; Goldstein, S.; Czapski, G. Chem. Res. Toxicol.
1998, 11, 712-713.
(38) Gerasimov, O. V.; Lymar, S. V. Inorg. Chem. 1999, 38, 4317-4321.
(39) Hodges, G. R.; Ingold, K. U. J. Am. Chem. Soc. 1999, 121, 10695-
10701.

Pressure Dependence of Peroxynitrite Reactions
Inorganic Chemistry, Vol. 40, No. 3, 2001 531
however, one can show46 from measured kD/kN and k-C/kN
47
Table 1. Experimentally Determined Volumes of Activation
∆Vq
∆(∆Vq)b,c
rate-constant ratios that the decomposition rate should be
insensitive to the medium viscosity;48 hence, the observed
insensitivity to viscosity is consistent with a radical mechanism
(for ONOOH). Using EPR spectroscopy, Augusto and co-
obs
b
reactant
pHa
(cm3 mol-1
)
(cm3 mol-1
)
ONOOH
ONOOH
4.5
4.5
4.5
7.2
7.2
7.2
7.2
+6
+14d
+12
+10
+11
-7
•-
4-
4-
workers have measured the formation of the CO3 radical in
ONOOH + Fe(CN)6
+13
+14
+16
∼0
- 49,50
ONOOH + Fe(CN)6
near-quantitative yield from ONOOCO2
,
based upon
4-
ONOOH + Mo(CN)8
limiting oxidation yields for its “indirect” reactions.49 Earlier
research in our13 and Goldstein and Czapski’s51 laboratories has
established that the reactive form of the adduct has chemical
properties and reactivity characteristics consistent with those
of the CO3•- + NO2• radicals, and Goldstein and Czapski have
also shown that product yields from “indirect” one-electron
oxidations by ONOOCO2- decreased markedly with increasing
medium viscosity; this is consistent (in this case) with a radical
mechanism.52 Finally, Lehnig has observed chemically induced
dynamic nuclear polarization (CIDNP) in the 15N NMR
ONOOH + Ni(cyclam)2+
ONOOCO2- + Fe(CN)6
-22
+18
4-
a In 0.15 M phosphate; other conditions for individual reactions as
listed in the text. b Error limits are (10-15%. c Defined as ∆Vq
-
D
∆Vq (Scheme 2). d Contained 5.0 mM added NO3
.
-
N
during cage escape than during internal rearrangement. This
result suggests that solvent restructuring constitutes a substantive
energetic barrier to separation of the radical pair; although
reasonable, this conclusion is not in accord with the frequently
invoked assumption that, since the viscosity of water is nearly
constant over this pressure range, kD will be pressure indepen-
dent. This behavior markedly contrasts that of the direct
bimolecular reaction between ONOOH and Ni(cyclam)2+, the
product yield for which is near stoichiometric and pressure
independent, i.e., ∆(∆Vq) = O (Table 1).
spectrum of NO3 during decomposition of O15NOOH and
-
O15NOOCO2-, which can be interpreted as indicating for-
mation of 15NO3 within the radical pairs {NO2 ,OH•} and
-
•
{NO2 ,CO3•-}.53 Thus, it seems appropriate to discuss the
•
pressure dependencies of the reactions of ONOOH and
ONOOCO2- within the context of the radical pair mechanisms,
i.e., Scheme 2. (Note that the rate-limiting step for the
ONOOCO2- reactions is adduct formation, whereas protonation
is a pre-equilibrium step preceding rate-limiting peroxo O-O
bond homolysis in the reactions of ONOOH.)
Interpretation of the Pressure Dependencies of Reaction
Rate Constants. (i) Bimolecular Reactions. The experimentally
measured bimolecular rate constant (kobs) for the formation of
ONOOCO2 is given by kobs ) k2/{(1 + [H+]/Ka)(1 +
-
Interpretation of the Pressure Dependencies of Product
Yields. According to Scheme 2, the product yields that are
attained when all of the radicals escaping the cage are scavenged
by reductants are dictated by the partitioning of the caged radical
pair between cage escape (kD) and rearrangement to NO3- (kN).
The pressure dependence of the rate-constant ratios can be
expressed as ∂ln(kD/kN)/∂P ) (∆VqN - ∆VqD)/RT (≡-∆(∆Vq)/
RT, where ∆(∆Vq) is the difference in volumes of activation
for cage escape and rearrangement within the cage). From the
pressure dependence of the maximal product yields (Ym), given
by the expression Ym ) kD/(kN + kD), one can calculate ∂ln(kD/
Kb/[H+])},8 where k2 is the intrinsic rate constant for reaction
between ONOO- and CO2, Ka ) 2 × 10-7 M is the acid
dissociation constant for ONOOH, and Kb ) 1 × 10-6 M is
the constant for the CO2 hydration-dehydration equilibrium,
CO2 + H2O h H+ + HCO3-. Under the experimental
conditions (pH 7.2), this equation reduces to kobs ≈ k2[H+]/Kb.
Since, in this case, the second acid dissociation constant (Ka′)
for phosphate, H2PO4 h H+ + HPO42-, is equal to [H+], it
-
follows that kobs ≈ k2Ka′/Kb, and ∆Vqobs ≈ ∆Vq(k2) + ∆V°(Ka′)
- ∆V°(Kb). The volumes of reaction for proton dissociation
from H2PO4- and CO2 hydrolysis are ∆V°(Ka′) ) -24 and -27
cm3 mol-1, respectively.54 Thus, the effects of pressure upon
kN)/∂P, hence ∆(∆Vq). Results obtained for radical scavenging
4-
of ONOOH decomposition products by Fe(CN)6
and
the two equilibria are mutually compensating and, from ∆Vq
Mo(CN)84-, and for ONOOCO2 decomposition products by
Fe(CN)64-, are listed in Table 1. For ONOOH decomposition,
∆(∆Vq) is independent of the identity of the scavenging ion
and the reaction pH; this case is expected for the mechanism
given in Scheme 2. In all cases, the values are large positive
numbers, which indicates that greater volume expansion occurs
-
obs
) -22 cm3 mol-1, one calculates ∆Vq(k2) = -25 cm3 mol-1
.
This value is consistent with a reaction that involves association
of a neutral molecule with an ion for which there is no
significant change in charge density, hence, only minor con-
tributions from solvent electrostriction. It therefore supports
proposals that CO2-catalyzed reactions of ONOO- involve
intermediary formation of a discrete ONOOCO2 adduct.8,55
-
•
•
-
(46) Under the usual reaction conditions (see text), the NO2 and CO3
free radicals are effectively scavenged, so that k-D ≈ 0. Application
of the steady-state approximation to the caged radical pair (Scheme
2) then gives ko ) kC(kD + kN)/(k-C + kD + kN), where ko is the
pH-independent first-order rate constant for ONOOH decomposition.
In water, kN/kD = 1.5 (determined from product yields for the “indirect”
reactions) and kN/k-C = 1,47 so that kobs = 0.6 kC. Assuming that kD
exhibits Debye-Stokes-Einstein behavior and the other microscopic
rate constants are insensitive to viscosity, increasing the viscosity by
20-fold would increase kN/kD to ∼30 for which kobs = 0.5 kC, i.e.,
causing the observed rate constant to decrease by less than 20%.
(47) Merenyi, G.; Lind, J.; Goldstein, S.; Czapski, G. J. Phys. Chem. A
1999, 103, 5685-5691.
The rate law for reaction of Ni(cyclam)2+ with peroxynitrite
is d[Ni(cyclam)3+]/dt ) k2′[ONOOH][Ni(cyclam)2+];16 the
combination of a small ∆Hq with a large negative ∆Sq suggests
that reaction occurs by an inner-sphere mechanism involving
ONOOH coordination prior to electron transfer.56 Following
algebraic arguments analogous to those described above, for
reaction of ONOOH with Ni(cyclam)2+ at pH 7.2 we obtain
the relationships kobs ≈ k2′Ka′/Ka and ∆Vq ≈ ∆Vq(k2′) +
obs
∆V°(Ka′) - ∆V°(Ka). Using ∆Vq ) -7 cm3 mol-1 and
obs
(48) Additional factors which can reduce the magnitude of experimentally
determined viscosity effects in this system (microheterogeneity,
reactions of secondary radicals) are discussed in ref 52.
∆V°(Ka) ) -7 cm3 mol-1 27
,
one obtains ∆Vq(k2′) ≈ +10 cm3
mol-1. This large positive value may reflect loss of coordinated
(49) Bonini, M. G.; Radi, R.; Ferrer-Sueta, G.; Ferreira, A. M. D. C.;
Augusto, O. J. Biol Chem. 1999, 274, 10802-10806.
(54) Isaacs, N. S. Liquid-Phase High-Pressure Chemistry; Wiley: New
York, 1981.
(55) Houk, K. N.; Condroski, K. B.; Pryor, W. A. J. Am. Chem. Soc. 1996,
118, 13002-13006.
(56) See, e.g.: Drljaca, A.; Hubbard, C. D.; van Eldik, R.; Asano, T.;
Basilevsky, M. V.; le Noble, W. J. Chem. ReV. 1998, 98, 2167-2289.
(50) See also: Meli, R.; Nauser, T.; Koppenol, W. H. HelV. Chim. Acta
1999, 82, 722-725.
(51) Goldstein, S.; Czapski, G. J. Am. Chem. Soc. 1998, 120, 3458-3463.
(52) Goldstein, S.; Czapski, G. J. Am. Chem. Soc. 1999, 121, 2444-2447.
(53) Lehnig, M. Arch. Biochem. Biophys. 1999, 368, 303-18.

532 Inorganic Chemistry, Vol. 40, No. 3, 2001
Coddington et al.
water from the axial ligand positions of the complex ion
accompanying ONOOH binding; typically, release of a single
coordinated H2O is assumed to contribute +13 cm3 mol-1 to
the overall ∆Vq(k2′).57
presence of these scavengers are therefore expected to cor-
respond to the steps leading to O-O bond homolysis, i.e., at
pH 7.2, kobs ≈ koKa′/Ka and ∆Vq ≈ ∆Vq + ∆V°(Ka′) -
obs
o
∆V°(Ka). Using ∆Vq ) +10 cm3 mol-1 (Table 1) and the
obs
previously indicated ∆Vo values, one obtains ∆Vqo ≈ +27 cm3
mol-1, independent of whether Fe(CN)64- or Mo(CN)64- is the
radical scavenger. This value is substantially larger than both
(ii) Unimolecular Reactions. The one-electron reductants,
Fe(CN)64-, Mo(CN)84-, and NO2-, are efficient scavengers of
the OH• radical.58,59 The mechanism given in Scheme 2 was
kinetically modeled using reported rate constants59 for reactions
the ∆Vq determined at pH 4.5 in this study and the pH-
o
of OH• and NO2 . These calculations indicated that, even in the
•
independent values determined by Goldstein and co-workers
in low phosphate environments.27 These data therefore support
earlier conclusions26,27 that phosphate ions somehow modulate
the reaction and suggest in particular that the interacting ion is
HPO42-. Ion-specific effects involving the NO2- ion may also
be important, since the difference in values obtained in media
-
absence of other scavengers, NO2 at levels above 1 µM will
compete effectively with NO2• for OH•. Thus, k-D is negligible
and the steady-state solution to Scheme 2 can be cast in the
form ko ) kC(1 + kD/kN)/(1 + k-C/kN + kD/kN), where ko is the
pH-independent first-order rate constant for ONOOH decom-
position. At pH 4.5, ko = kobs, and therefore ∆Vq = ∆Vq
.
obs
containing 0.05 and 5.0 mM NO2-, i.e., +6 and +14 cm3 mol-1
,
o
Assuming that the pressure dependencies of both k-C and kN
are small, the equality47 k-C ≈ kN will hold at all pressures,
and ∆Vqo ≈ ∆VqC + RT∂ln[(1 + kD/kN)/(2 + kD/kN)]/∂P. Based
upon the measured pressure dependence of product yields for
the “indirect” reactions, the latter term is approximately -1.5
respectively (Table 1, Figure 1), appear to be outside the range
of experimental uncertainty.
Concluding Remarks. The pressure dependencies measured
for several “indirect” reactions of ONOOH are consistent with
a mechanism in which peroxo O-O bond homolysis is the initial
reaction step. These reactions are clearly distinguishable from
CO2-catalyzed oxidations by ONOO-, whose pressure depend-
encies support conclusions based upon the rate law8 that reaction
is initiated by associative interaction of the oxidant and catalyst.
Although most of the data can be accommodated by the scheme
given in Figure 2, the observations that the product yields of
cm3 mol-1. The average of values obtained for ∆Vq for
obs
reactions at pH 4.5 (Table 1) is +11 cm3 mol-1, which is nearly
identical to the values in low phosphate environments reported
by Goldstein et al.27 Thus, the H2PO4 ion does not appear to
-
influence the reaction pathway. One also notes that ∆Vq is
obs
-
4-
independent of whether NO2 or Fe(CN)6 is the radical
scavenger, as expected. From the expression for ∆Vq , one
o
the “indirect” reactions are pressure dependent, and that the ∆Vq
o
obtains ∆VqC ≈ +12 cm3 mol-1. This value is somewhat larger
than the estimated values of ∆VqC for single bond O-O cleavage
in nonpolar environments, which are typically +3-5 cm3
values for several reactions are sensitive to the ionic composition
of the medium, are difficult to rationalize within the context of
this simple dynamical model and may indicate the existence of
additional intermediates, e.g., solvent-separated radical pairs.
mol-1 60
, but it is certainly consistent with rate-limiting formation
of two neutral fragments via bond cleavage.
In weakly alkaline solutions, a competing pathway for
ONOOH decomposition that gives rise to formation of NO2
Acknowledgment. Funding for this research was provided
by the National Institutes of Health under grant AI-15843 (to
J.K.H.).
-
and O2 becomes important. This pathway, which is thought to
be initiated by reaction of OH• with ONOO-,14,42 is completely
4-
suppressed upon addition of the Fe(CN)6 ion and other
Supporting Information Available: A compilation of individual
rate constants under various reaction conditions (Tables S1-S4), data
efficient radical scavengers.42 Measured values of ∆Vqobs in the
3-
plots of the pressure dependence of the Fe(CN)6 yield and rate
(57) See, e.g.: Metelski, P. D.; Swaddle, T. W. Inorg. Chem. 1999, 38,
301-307. Jolley, W. H.; Stranks, D. R.; Swaddle, T. W. Inorg Chem.
1990, 29, 1948-1951, and references therein.
4-
constant for Fe(CN)6 oxidation by ONOOH (Figure S1), and the
3-
-
4-
Fe(CN)6 yield from reaction between ONOOCO2 and Fe(CN)6
(Figure S2), and the temperature dependence of the rate constant for
reaction between ONOOH and Ni(cyclam)2+ (Figure S3). This material
is available free of charge via the Internet at http://pubs.acs.org.
(58) Logager, T.; Sehested, K. J. Phys. Chem. 1993, 97, 6664-6669.
(59) Buxton, J. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. B. J. Phys.
Chem. Ref. Data 1988, 17, 513-886.
(60) Neuman, R. C., Jr.; Bussey, R. J. J. Am. Chem. Soc. 1970, 92, 2440-
2445.
IC981359F