Kinetics and mechanisms of aqueous ozone reactions with bromide, sulfite, hydrogen sulfite, iodide, and nitrite ions

Article abstract of DOI:10.1021/ic000919j

Reactions of ozone with Br^-, SO₃^2-, HSO₃^-, I^-, and NO₂^-, studied by stopped-flow and pulsed-accelerated-flow techniques, are first order in the concentration of O₃(aq) and first order in the concentration of each anion. The rate constants increase by a factor of 5 × 10⁶ as the nucleophilicities of the anions increase from Br^- to SO₃^2-. Ozone adducts with the nucleophiles are proposed as steady-state intermediates prior to oxygen atom transfer with release of O₂. Ab initio calculations show possible structures for the intermediates. The reaction between Br^- and O₃ is accelerated by H⁺ but exhibits a kinetic saturation effect as the acidity increases. The kinetics indicate formation of BrOOO^- as a steady-state intermediate with an acid-assisted step to give BrOH and O₂. Temperature dependencies of the reactions of Br^- and HSO₃^- with O₃ in acidic solutions are determined from 1 to 25 °C. These kinetics are important in studies of annual ozone depletion in the Arctic troposphere at polar sunrise.

Full text of DOI:10.1021/ic000919j

4436
Inorg. Chem. 2001, 40, 4436-4442
Kinetics and Mechanisms of Aqueous Ozone Reactions with Bromide, Sulfite, Hydrogen
Sulfite, Iodide, and Nitrite Ions
Qian Liu, Lynn M. Schurter, Charles E. Muller, Simone Aloisio, Joseph S. Francisco, and
Dale W. Margerum*
Department of Chemistry, Purdue University, West Lafayette, Indiana 47907-1393
ReceiVed August 11, 2000
Reactions of ozone with Br-, SO32-, HSO3-, I-, and NO2-, studied by stopped-flow and pulsed-accelerated-flow
techniques, are first order in the concentration of O3(aq) and first order in the concentration of each anion. The
rate constants increase by a factor of 5 × 106 as the nucleophilicities of the anions increase from Br- to SO3
.
2-
Ozone adducts with the nucleophiles are proposed as steady-state intermediates prior to oxygen atom transfer
with release of O2. Ab initio calculations show possible structures for the intermediates. The reaction between
Br- and O3 is accelerated by H+ but exhibits a kinetic saturation effect as the acidity increases. The kinetics
indicate formation of BrOOO- as a steady-state intermediate with an acid-assisted step to give BrOH and O2.
Temperature dependencies of the reactions of Br- and HSO3- with O3 in acidic solutions are determined from 1
to 25 °C. These kinetics are important in studies of annual ozone depletion in the Arctic troposphere at polar
sunrise.
Introduction
estimated from the S(IV) reactions from pH -0.30 to 4.02,
-
where HSO3 and SO2 are the main species present.5 In the
Ozone reacts rapidly with a wide range of inorganic ions in
aqueous solution.1 The kinetics of its reactions with bromide,
sulfite, hydrogen sulfite, iodide, and nitrite are important in
environmental studies and in water treatment processes.2 The
reaction of O3 with Br- is of great concern in ozonation
processes because it is the first step leading to the generation
of BrO3-. Bromate ion has been identified as a carcinogen and
neurotoxin, and a maximum contamination level of 0.01 mg/L
has been set by the US EPA.3 In addition, the reactions of O3
with Br- and O3 with S(IV) species are involved in the ozone
depletion chemistry in the Arctic troposphere.4 The kinetics of
these reactions5-8 are revisited with focus on the reactions at
low pH and low temperature because of their possible role in
reactions in Arctic aerosols, sea ice, and snowpack. The reaction
of O3 with I- in seawater affects the global flux of iodine9,10
and may be important in taste and odor problems in drinking
water.11 The reaction is very fast, and until now the rate constant
has only been estimated.10 Similarly, the reaction between O3
present work, pulsed-accelerated-flow (PAF) spectrometry12,13
is used for the direct determination of the O3 rate constants with
I- and SO32-. The reaction of O3 with NO2- is important both
in water treatment processes and in atmospheric chemistry. The
reaction kinetics were previously studied1,9,14 only below pH
5. We use stopped-flow methods to extend the pH range from
1.5 to 10.6.
Comparison of the kinetics and rate constants for these five
anions with O3 leads to the proposal that adduct formation occurs
followed by oxygen-atom transfer with the direct elimination
of O2. Ab initio calculations show feasible structures for the
O3-anion adducts.
Experimental Section
-
Reagents. Solutions of Br-, I-, SO32-, HSO3-, and NO2 are
prepared from their respective sodium salts immediately prior to use
in pre-ozonated, freshly boiled, argon-purged water and are protected
-
with argon until use. The water used to prepare the SO32- and HSO3
2-
and SO3 is extremely rapid and the rate constant has been
solutions requires thorough argon purging (>1 h) to adequately remove
oxygen and protect against S(IV) loss. Solution acidity is adjusted with
standardized HClO4, H2SO4, carbonate-free NaOH, and the following
buffers: HC2H3O2 (pKa ) 4.4115 at µ ) 0.50 M, pKa ) 4.2615 at µ )
0.10 M), NaH2PO4 (pKa ) 6.5716 at µ ) 0.50 M, pKa ) 6.7216 at µ )
0.10 M), and NaHCO3 (pKa ) 9.7012,17 at µ ) 0.50 M, pKa ) 10.0018
(1) Hoigne´, J.; Bader, H.; Haag, W. R.; Staehelin, J. Water Res. 1985,
19, 993-1004.
(2) Hoigne´, J. Chemistry of Aqueous Ozone and Transformation of
Polluants by Ozonation. Handbook of Enironmental Chemistry;
Hrubec, J., Ed.; Springer-Verlag: Berlin, 1998; Vol. 5, Part C, pp
83-141.
(3) US EPA. National Drinking Water Regulations: Disinfectants and
Disinfection Byproducts. Fed. Regist. 1998, 63, 69390-69476.
(4) Oum, K. W.; Lakin, M. J.; Finlayson-Pitts, B. J. Geophys. Res. Lett.
1998, 25, 3923-3926.
(12) Fogelman, K. D.; Walker, D. M.; Margerum, D. W. Inorg. Chem.
1989, 28, 986-993.
(13) Bowers, C. P.; Fogelman, K. D.; Nagy, J. C.; Ridley, T. Y.; Wang,
Y.; Evetts, S. W.; Margerum, D. W. Anal. Chem. 1997, 69, 341-
438.
(5) Erickson, R.; Yates, L.; Clark, R.; McEwen, D. Atmos. EnViron. 1977,
11, 813-817.
(6) Haruta, K.; Takeyama, T. J. Phys. Chem. 1981, 85, 2382-2388.
(7) Haag, W. R.; Hoigne´, J. EnViron. Sci. Technol. 1983, 17, 261-267.
(8) Maahs, H. G. Atmos. EnViron. 1983, 17, 341-345.
(9) Garland, J. A.; Curtis, H. J. Geophys. Res. 1981, 86, 3183-3186.
(10) Garland, J. A.; Elzerman, A. W.; Penkett, S. A. J. Geophys. Res. 1980,
85, 7488-7492.
(14) Penkett, S. A. Nature Phys. Sci. 1972, 240, 105-106.
(15) Urbansky, E. T.; Cooper, B. T.; Margerum, D. W. Inorg. Chem. 1997,
36, 1338-1344.
(16) Smith, R. M.; Martell, A. E. Critical Stability Constants; Plenum: New
York, 1976; Vol. 4.
(17) Odier, M.; Plichon, V. Anal. Chim. Acta 1971, 55, 209-220.
(18) Frydman, N.; Nilsson, G.; Rengemo, T.; Sillen, L. G. Acta Chem.
Scand. 1958, 12, 878-884.
(11) Bishel, Y.; von Guten, U. EnViron. Sci. Technol. 1999, 33, 4040-
4045.
10.1021/ic000919j CCC: $20.00 © 2001 American Chemical Society
Published on Web 07/19/2001

Aqueous Ozone Reactions
Inorganic Chemistry, Vol. 40, No. 17, 2001 4437
at µ ) 0.10 M). Solutions of NaClO4 are prepared from the
recrystallized solid and standardized gravimetrically for use in adjusting
the ionic strength of all solutions. The pH measurements are taken
shortly after the experiments are conducted, and the measured pH values
are corrected to p[H+] ()-log[H+]) on the basis of electrode calibration
data.
Generation of Ozone. Ozone is generated by passing research grade
oxygen through a coaxial ozonator tube, which is powered by a 9 kV,
30 mA neon sign transformer. The molar absorptivity value for ozone
in solution was reviewed by Maahs8 and determined spectrophoto-
metrically in this study to be 3000 M-1 cm-1 at 260 nm. The ozone/
oxygen mixture is bubbled via gas dispersion tubes through acidic
reaction solutions. The concentration of ozone in solution is measured
immediately prior to use by a Perkin-Elmer Lambda 9 UV-vis
spectrophotometer, and solutions are kept acidic under red light to
minimize decomposition.
absorbance of the reaction, A∞ is the absorbance at infinite time, and b
is the path length of the cell. The upper limit of PAF measurements is
approached in acquiring data for the very fast reactions of O3 with
SO32- and with I- ions. Because of this fact, the reactions are run under
second-order unequal-concentration conditions with O3 as the limiting
reactant. For irreversible second-order reaction conditions,20 the second-
order apparent rate constant is defined by the double-reciprocal
relationship in eq 5, with Mexptl defined by eq 6, where q ) CA/CB (CA
1
1
1
k12
)
+
(5)
(6)
(7)
k12,app km,12
V
1 - qe-R
1 - q
Av - AA(1 - q) - Ap
AA + (1/q)(AB - Ap)
1 - q
R
Mexptl
)
)
ln
(
) (
)
Kinetic Methods. Stopped-flow methods are used to measure O3
reactions with Br-, HSO3-, and NO2- under pseudo-first-order condi-
tions with excess concentrations of these nucleophiles. Under these
conditions, solvent decomposition of O3 is negligible during the kinetic
measurements. First-order conditions are advantageous, where possible,
because the relative O3 concentrations are measured during the reaction
so that volatility is not a problem.
bk12,appCA(1 - q)
R )
V
> CB). Iteration of eqs 6 and 7 yields k12,app, and a plot of 1/k12,app vs
1/V gives a slope of 1/km,12 and an intercept of 1/k12 (reciprocal of the
second-order rate constant for the reaction).
For faster reaction conditions, where qe-R , 1(<0.05), the iteration
on R can be eliminated and Mexptl is represented by eq 8. Substitution
of eq 5 into eq 8 yields eq 9, which enables a direct calculation of k12
from the inverse of the slope of a plot of Mexptl vs V.
Data for NO2- reactions are collected by monitoring the disappear-
ance of O3 at 260 nm on a Dionex-Durrum Model D-110 stopped-
flow spectrophotometer. Data for the reactions of O3 with Br- and
-
HSO3 are collected on an Applied PhotoPhysics stopped-flow spec-
-
trophotometer, Model SX.18MV. The appearance of Br3 at 266 nm
(ꢀBr ) 40 900 M-1 cm-1
)
19 is monitored for the O3 reaction with Br-.
-
AV - AV(1 - q) - Ap
3
V
1
Mexptl
)
)
ln
bkappCA 1 - q
(8)
(9)
The disappearance of O3 at 260 nm is monitored for the reaction
between O3 and HSO3
AV + (1/q)(AB - Ap)
-
.
Pseudo-first-order rate constants (kobsd) are obtained from fits of ln-
(At - A∞)/(A0 - A∞) vs time. Kinetic data are fit to the first-order model
represented by eq 1, where [Xn-] ) [NO2-], [Br-], [HSO3-] and k12 is
ln[(1 - q)-1
)
]
1
V
Mexptl
)
+
(
km,12 k12
CAb
It has been shown that reliable k12 values are obtained for second-
order unequal conditions where the initial half-life, t1/2, as defined by
eq 10, is greater than 22 µs. Our smallest t1/2 value is 26 µs.
-d[O3]
dt
) kobsd[O3] kobsd ) k12[Xn-
]
(1)
the second-order reaction rate constant. The reactions are followed for
at least 5 half-lives. All kinetic traces are an average of 5 runs.
Temperatures are controlled to ( 0.2 °C.
2CA - CB
1
1
(t1/2)i )
ln
(10)
(
) (
)
k12 CA - CB
CA
PAF methods are used to measure the extremely fast reactions of I-
2-
and SO3 with O3 at 25.0 ( 0.2 °C. In these cases, second-order
Computation. Calculations are performed using the GAUSSIAN
9421 suite of programs to determine what are the plausible structures
for the reaction intermediates. Equilibrium geometries are optimized
using the Becke three-parameter hybrid functional combined with the
Lee, Yang, and Parr correlation (B3LYP)22 density functional theory
method. This method is used with the 6-31G(d) basis set in the full
optimization of the geometries. Harmonic vibrational frequencies are
calculated at this level of theory to verify if the resulting structures are
true minimums.
unequal-concentration conditions are necessary to obtain the rate
constants. On the PAF Model IV instrument,13 integrating observation
over an optical path length of 2.05 cm during a 0.22 s pulse of
continuous flow enables the collection of 250 points of intensity vs
time data per trial. The solutions are mixed by employing a decelerated
flow velocity of 12 to 3 m/s. The mixing constant (kmix) in eq 2 is a
kmix ) kmV
(2)
(3)
Results and Discussion
1
1
1
)
+
kapp kmV kr
O3 Reaction with Br-. The reaction between O3 and Br- is
followed under pseudo-first-order conditions with Br- (0.50-
100 mM) in large excess over O3 (∼10-5 M) at 25.0 ( 0.2 °C,
p[H+] ) 5.9, and µ ) 0.46 M. Under conditions of excess Br-,
Br3- forms rapidly once Br- is oxidized to Br2. The rate of O3
function of the flow velocity (V) and a proportionality constant (km).
The apparent rate constant (kapp) is defined by the double-reciprocal
relationship in eq 3, where kr is the first-order reaction rate constant.
For pseudo-first-order reaction conditions where kapp is greater than
4000 s-1, eq 4 describes the relationship between absorbance and
(20) Gerritsen, C. M.; Gazda, M.; Margerum, D. W. Inorg. Chem. 1993,
32, 5739-5748.
(21) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.; Johnson,
B. G.; Robb, M. A.; Cheeseman, J. R.; Keith, T.; Petersson, G. A.;
Montgomery, J. A.; Raghavachari, K.; Al-Laham, M. A.; Zakrzewski,
V. G.; Ortiz, J. V.; Foresman, J. B.; Cioslowski, J.; Stefanov, B. B.;
Nanayakkara, A.; Challacombe, M.; Peng, C. Y.; Ayala, P. Y.; Chen,
W.; Wong, M. W.; Andres, J. L.; Replogle, E. S.; Gomperts, R.;
Martin, R. L.; Fox, D. J.; Binkley, J. S.; Defrees, D. J.; Baker, J.;
Stewart, J. P.; Head-Gordon, M.; Gonzalez, C.; Pople, J. A. Gaussian
94, revision B.3; Gaussian, Inc.: Pittsburgh, PA, 1995.
AV - A∞
A0 - A∞ bkapp bkm bkr
V
1
V
Mexptl
)
)
)
+
(4)
velocity. Mexptl represents the degree of reaction in the observation path,
AV is the absorbance at a particular solution velocity, A0 is the initial
(19) Wang, T. X.; Kelley, M. D.; Cooper, J. N.; Beckwith, R. C.; Margerum,
D. W. Inorg. Chem. 1994, 33, 5872-5878.
(22) Becke, A. M. J. Chem. Phys. 1993, 98, 5648-5652.

4438 Inorganic Chemistry, Vol. 40, No. 17, 2001
Liu et al.
Table 1. Parameters and Calculated Rate Constants for O3 Reaction
with Br- at µ ) 1.0 Ma
temp,
°C k1, M-1 s-1 k2/k-1, M-1 k3/k-1 kH2O, M-1 s-1
k
H+, M-1 s-1
25.0 1.10 × 104
15.0 7.59 × 103
5.0 2.41 × 103
0.47
0.33
0.64
0.024 2.58 × 102
0.022 1.63 × 102
0.042 97
3.64 × 103
1.99 × 103
9.77 × 102
a ∆Hqk1 ) 50 ( 10 kJ/mol and ∆Sqk1 ) 2 ( 50 J K-1 mol-1. ∆Hq
+
H
) 43.0 ( 1.2 kJ/mol and ∆Sq ) -32 ( 1 J K-1 mol-1, at 1.0 M
+
H
[H+]. ∆HqH O ) 31.2 ( 0.6 kJ/mol and ∆SqH O ) -94 ( 2 J K-1 mol-1
.
2
2
k1, k2/k-1, and k3/k-1 values are determined by fitting the data to eq 11.
+
kH O values are calculated from eq 12, and kH values are calculated
2
from eq 11 at 1.0 M [H+].
The rate expression is derived in eq 11 based on a steady-
state approximation in [BrOOO-]. The water path rate constant
(kH O) (eq 12) is valid at very low [H+].
2
Figure 1. Temperature and [H+] dependence of the O3 reaction with
Br-. Conditions: [Br-] ) 100 mM, µ ) 1.0 M (NaClO4), acidity is
adjusted with HClO4, except for open squares, where H2SO4 is used.
k2
k3
k1
[H+] +
(
)
kobsd
[Br-]
k-1
k-1
)
(11)
(12)
loss (eq 1) equals the rate of [Br2]T () [Br2] + [Br3-]) formation.
Individual kinetic traces give excellent first-order fits in [O3].
The plot of kobsd vs [Br-] is linear, indicating a first-order
dependence on [Br-]. Linear regression of the data yields k12
equal to 248 ( 1 M-1 s-1 from the slope and an intercept (with
a very small correction for O3 loss in the absence of Br-)
statistically equal to zero.
k2
k3
1 +
[H+] +
k-1
k-1
k1
kH O
)
2
k
-1 + 1
k3
The acid dependence study is carried out at 1.0 M ionic
strength for p[H+] ) 0-3 with perchloric or sulfuric acid (pKa
) 1.32).16 Sulfuric acid is chosen because of its high levels in
Experimental data collected at three temperatures (25.0, 15.0,
and 5.0 °C) are fit with eq 11 using a nonlinear curve-fitting
program with a Marquardt-Levenberg algorithm (SigmaPlot).
As shown in Figure 1, the experimental results correspond to
the proposed mechanism at all temperatures, showing strong
support for the presence of a steady-state intermediate species.
-
Arctic aerosols and to test if HSO4 acts as a general acid to
accelerate the reaction. Figure 1 shows both the HClO4 results
(black squares) and the H2SO4 results (open squares) at 25.0
°C. Four facts are evident: (1) the reaction rate is assisted by
[H+]; (2) the rate constants show evidence of a saturation effect
Three parameters (k1, k2/k-1, and k3/k-1) are determined from
+
as [H+] increases; (3) HSO4 does not accelerate the reaction
-
the data at each temperature. Values for kH O and for kH
-
2
() kobsd/[Br-] at [H+] ) 1.0 M) are summarized in Table 1.
as a general acid (the H2SO4 results are similar to the HClO4
results); (4) the nonzero intercept is consistent with a water
pathway.
Haruta and Takeyama6 showed that [H+] assists the reaction;
however, the [H+] dependence we find deviates greatly from
the linear relationship they reported. Our study covers a much
wider [H+] range (10-3 to 1.0 M) compared to their range of
10-3 to 10-1.2 M, and the second-order experimental conditions
they used require accurate knowledge of the initial O3 concen-
trations to calculate the rate constants.
The kH O value obtained in this [H+] dependence study is 258
2
( 18 M-1 s-1 at 25.0 °C and µ ) 1.0 M, which is within
experimental error of the value of 248 ( 1 M-1 s-1 (µ ) 0.46
M, 25.0 °C) obtained from the [Br-] dependence study. A small
difference may be caused by the difference in ionic strength
for the two cases. Both values are somewhat larger than the
hydrogen-ion-independent path rate constant of 211 M-1 s-1
(27 °C, µ ) 0.1 M) reported by Haruta and Takeyama6 and
significantly larger than the value of 160 M-1 s-1 reported by
Haag and Hoigne´.7
On the basis of our kinetic data, a mechanism is proposed as
shown in Scheme 1, where O3 reacts with Br- to form the
steady-state intermediate BrOOO- in the first step. This
intermediate subsequently reacts with H+ or H2O to give O2
and HOBr.
Activation enthalpies and entropies are calculated (eq 13) for
k1, the water path, and the proton path from the data in Figure
1. The ∆Hqk1 value is 50 ( 10 kJ mol-1, the proton path gives
k
T
∆Sq ∆Hq
kB
ln
) ln
+
-
(13)
( )
( )
h
R
RT
Scheme 1
∆Hq ) 43.0 ( 1.2 kJ mol-1, and the water path gives ∆Hq
+
O3 + Br- { k1 } BrOOO-
H
H2O
) 31.2 ( 0.6 kJ mol-1. The ∆Sq values are also given in Table
k-1
1. Under possible Arctic conditions at 273 K and 1.0 M acid,
k2
the predicted rate constants for Br- reaction with O3 is kH
)
BrOOO - + H + 98 HOBr + O2
+
6.81 × 102 M-1 s-1
.
k3
9
O3 Reaction with SO32-. S(IV) is distributed among SO2
(aq), HSO3-, and SO32- as a function of pH.23 Erickson et al.5
showed that O3 reacts with S(IV) from p[H+] ) -0.30 to 4.02
with a first-order dependence in [O3] and a first-order depen-
BrOOO- + H2O
8 HOBr + O2 + OH-
HOBr + Br - + H+ {fast} Br2 + H2O
Br2 + Br - {fast} Br3
-
(23) Brandt, C.; van Eldik, R. Chem. ReV. 1995, 95, 119-190.

Aqueous Ozone Reactions
Inorganic Chemistry, Vol. 40, No. 17, 2001 4439
Figure 2. p[H+] dependence of the O3 reaction with SO32-. Condi-
tions: initial [O3] ) (2.46-3.89) × 10-5 M, initial [SO32-]T ) 3.00 ×
10-5 to 3.00 × 10-4 M, µ ) 0.1 M, 25.0 °C, [acetate]T ) 10 mM for
p[H+] ) 4.5-5.06, [phosphate]T ) 10 mM for p[H+] ) 5.65-7.49,
[carbonate]T ) 20 mM for p[H+] ) 8.73-9.64.
-
Figure 3. [S(IV)]T dependence of the O3 reaction with HSO3
.
Conditions: [O3] ) ∼10-5 M, [HSO3-]T ) 0.48-2.97 mM, µ ) 1.0
M, 25.0 °C, acidity is adjusted with HClO4.
dence in S(IV). The kinetic contribution of SO2 was not
evaluated, but it was negligible above pH 2.
In the present work, the kinetics of the O3/S(IV) reaction are
investigated at p[H+] ) 4.50-9.64 at 25.0 ( 0.2 °C and µ )
0.10 M, with [O3] ) (2.46-3.89) × 10-5 M and [SO32-]T )
3.00 × 10-5 to 3.00 × 10-4 M. In this p[H+] range, the
concentration of SO2(aq) is negligible (pKa1 ) 1.90, pKa2
)
6.30)23 and the experimental rate constant k12 due to SO32- and
-
HSO3 is expressed by eq 14. The experimental second-order
kSO 2- Ka2 + kHSO - [H+]
3
3
k12
)
(14)
Figure 4. Temperature and [H+] dependence of the O3 reaction with
HSO3-. Conditions: µ ) 1.0 M, acidity is adjusted with HClO4.
Ka2 + [H+]
rate constants obtained from eqs 5-7 are plotted as a function
25 °C and p[H+] values of 0.49, 0.75, and 1.20. The reaction is
first order in [S(IV)]T at all p[H+] values. Figure 4 shows the
result of the [H+] dependence study for [H+] ) 0.05-0.35 M
at four temperatures. At µ ) 1.0 M, pKa1 values of 1.90, 1.79,
1.71, and 1.63 are used for 25.0, 15.0, 8.1, and 1.2 °C,
of [H+] in Figure 2. A nonlinear curve fit of the data using eq
14 yields a value of (1.3 ( 0.1) × 109 M-1 s-1 for kSO with
2-
3
a negligible HSO3 contribution (i.e., kHSO is less than 106
-
-
3
M-1 s-1). The rate constant for SO32- with O3 agrees with the
value of 1.0 × 109 M-1 s-1 at 22 °C obtained from data
extrapolated from pH 5.2.1
respectively, based on ∆H° ) -18 kJ mol-1 24
For the same
.
temperatures and ionic strength values the respective pKa2 values
are 6.30, 6.15, 6.06, and 5.96, based on ∆H° ) -12.3 kJ
2-
Our value for kSO is significantly smaller than the value of
3
2.2 × 109 M-1 s-1 reported by Erickson et al.5 They studied
mol-1 25
.
At 25.0 °C the kSO value of 1.3 × 109 M-1 s-1 is
2-
-
3
this reaction below pH 4.02, where the HSO3 species is
-
used in the fit of data to eq 15, which yields kHSO ) (8.0 (
dominant and the kinetics of HSO3- and SO32- reactions with
O3 both contribute. In our study, the p[H+] range (4.50-9.64)
is chosen such that the SO32- reaction with O3 is the dominant
path and its rate constant is more accurately evaluated. The
3
0.1) × 105 M-1 s-1 at µ ) 1.0 M and a kSO value that is
2
statistically equal to zero. Hoigne´ et al.1 reported a value of 3.2
× 105 M-1 s-1 at 22 °C based on the same pK1 value but with
a pKa2 value of 7.2. The ionic strength was not given. The
contribution of the SO32- path to the overall rate is only 1.3%
at p[H+] ) 1.20 and 0.3% at p[H+] ) 0.49 at 25.0 °C. At lower
temperatures, measurements at the highest acidity are used to
2-
disagreement in kSO values may also be attributable to the
3
difference in ionic strength, which is controlled at 0.10 M in
our study, whereas the ionic strength was not mentioned in the
previous studies.1,5
2-
-
O3 Reaction with HSO3-. We also studied the kinetics of
the O3 reaction with S(IV) under more acidic conditions (p[H+]
) 0.49-1.20) by stopped-flow spectroscopy under pseudo-first-
order conditions with [HSO3-]T (from 0.48 to 2.97 mM) in large
excess over O3 (∼10-5 M) and µ ) 1.0 M. In this p[H+] range,
evaluate kHSO by first neglecting the contributions of SO3
.
3
This yields kHSO ) (5.6 ( 0.2) × 105 M-1 s-1 at 15.0 °C,
-
3
(3.6 ( 0.2) × 105 M-1 s-1 at 8.1 °C, and (2.43 ( 0.04) × 105
M-1 s-1 at 1.2 °C. These values are almost the same as the
values of kHSO ) (5.23 ( 0.03) × 105 M-1 s-1 at 15.0 °C,
-
3
(3.69 ( 0.05) × 105 M-1 s-1 at 8.1 °C, and (2.51 ( 0.05) ×
-
SO2(aq) and HSO3 are the major forms of S(IV) while the
105 M-1 s-1 at 1.2 °C obtained when the 25 °C value of kSO
2-
2-
2-
SO3 concentration is very low. However, the value of kSO
3
3
2-
is used to fit all the temperature data. The latter gives the best
is so large that SO3 may still make a small contribution to
the observed reaction rate. Equation 15 gives the rate expression,
fit of data (Figure 4), which indicates that the temperature
2-
2-
dependence of kSO is small. The values of kSO as a function
3
3
kSO 2-K Ka2 + kHSO -K [H+] + kSO [H+]2
of temperature are not measured because the PAF Model IV
instrument does not permit temperature variation.
kobsd
a1
a1
3
3
2
)
(15)
Ka1Ka2 + Ka1[H+] + [H+]2
[S(IV)]T
(24) van Eldik, R.; Harris, G. Inorg. Chem. 1980, 19, 880-886.
(25) Sille´n, L.; Martell, A. Stability Constants of Metal Ion Complexes;
Burlington House: London, 1971; Supplement 1.
taking into consideration the contributions of all three species.
Figure 3 shows the results of the [S(IV)]T dependence study at

4440 Inorganic Chemistry, Vol. 40, No. 17, 2001
Liu et al.
•
-
Table 2. Comparison of Experimental Rate Constants (25.0 °C, in M-1 s-1) of the Reactions between Inorganic Ions (Xn-) with O3 and CO3
with Predictions Based on a Marcus Outer-Sphere Electron-Transfer Model
O3 + Xn-
•CO3- + Xn-
couple
E°, V
k11
k12,pred
k12,obsd
k12,pred
k12,obsd
-
O3/O3
1.01a
1.59c
1.33a
1.04a
0.72d
4b
•CO3-/CO3
1.6b
2-
I/I-
2 × 108 b
29.1
1.2 × 109
5.83 × 105
1.3 × 109
8.0 × 105
248
1.9 × 106
1.0 × 104
9.2 × 106
2.5 × 108 c
6.6 × 105 c
2.9 × 107 c
0.3b
0.061
-
NO2/NO2
•SO3-/SO3
4b
1.3 × 103
2-
-
HSO3‚/HSO3
Br/Br-
1.92a
a Reference 34. b Reference 28. c References 32 and 33. d Sarala, R.; Islam, M. A.; Rabin, S. B.; Stanbury, D. M. Inorg. Chem. 1990, 29, 1133-
1142.
the expression for the observed rate constant, assuming that
-
NO2 is the only reactive form. This gives an excellent fit of
the data (Figure 5), with a second-order rate constant of (5.83
( 0.04) × 105 M-1 s-1, which is in reasonable agreement with
the value obtained at p[H+] ) 2. Both rate constants are higher
than previous values1,2 which were obtained at either lower pH
(pH 4)9 at 25 °C or in a relatively narrow pH range (from 1.8
to 5)1 at temperatures around 20-23 °C and then extrapolated
to high pH.
Lack of General Acid-Base Effects. Previous studies of
the kinetics of Br- oxidation by HOBr27 and of SO32- oxidation
by HOCl12 have shown the importance of general-acid assistance
in these halogen-cation transfer mechanisms. The O3 oxidation
Figure 5. [H+] dependence of the reaction of O3 with NO2
.
-
-
2-
of Br-, NO2 and SO3 are affected by the acidity of the
Conditions: [O3] ) (1.75-4.64) × 10-5 M, [NO2]T ) 0.2 mM, µ )
0.50 M, 25.0 °C, no buffer for p[H+] ) 1.46-3.02, acetate buffer for
p[H+] ) 3.52-4.83, phosphate buffer for p[H+] ) 5.49-7.52, carbonate
buffer for p[H+] ) 8.72-10.61, [buffer]T ) 0.050 M for all buffers.
solutions but not by buffer concentrations.
Conclusions
Mechanistic Comparisons. Table 2 compares the experi-
mentally determined rate constants for the reactions of O3 with
I-, NO2-, and SO32- with values predicted from a Marcus outer-
sphere electron-transfer mechanism. Values of the self-exchange
rate constants for O3/O3-, I/I-, NO2/NO2-, and SO3-/SO32- are
taken from Stanbury’s review.28 In these O3 reactions, the
experimental values (k12,obsd) are many orders of magnitude
greater than the predicted k12 values. For example, the k12,obsd
value for the O3/I- reaction is so large that electron transfer
can be ruled out because it would require the reverse reaction
to be vastly greater than the diffusion-controlled rate. As
Stanbury29-30 and Bennett31 have pointed out, reactions between
main-group species rarely proceed by an outer-sphere electron-
transfer mechanism. Greater orbital overlap can occur with
nonmetal species than when one reactant is a metal ion shielded
by ligands. They propose that greater orbital overlap can lead
to a lower energy transition state. Table 2 gives rate constants
from the work of Huie et al.32-33 for the reactions of the
carbonate radical ion (•CO3-), with I-, NO2-, and SO32-. The
The activation enthalpy and entropy for the HSO3- reaction
with O3, obtained from an Eyring plot (eq 13), give ∆Hq
-
HSO3
) 30.9 ( 0.9 kJ mol-1 and ∆Sq
) -28 ( 3 J K-1 mol-1
.
-
HSO3
Much higher values were reported by Erickson et al.5 from
studies at only two temperatures (25 and 16 °C).
O3 Reaction with I-. The PAF method is used to study the
reaction of O3 with I- at 25.0 °C by following the loss of O3 at
260 nm under second-order unequal-concentration conditions
([O3] ) (2.74-4.07) × 10-5 M, [I-] ) 3.34 × 10-5 to 5.57 ×
10-4 M, p[H+] ) 6.7, and µ ) 0.1 M). The average rate constant
is (1.2 ( 0.1) × 109 M-1 s-1. This is 240 times larger than the
previously specified lower limit from direct chemical studies
but agrees with a value of 1 × 109 M-1 s-1 estimated from
surface deposition measurements.10
O3 Reaction with NO2-. A nitrite dependence study is carried
out at p[H+] ) 2.0, µ ) 0.50 M, and 25.0 °C with [NO2-] in
10-fold excess over [O3]initial. Values of kobsd show a first-order
dependence in [NO2-] with a second-order rate constant of (6.08
( 0.06) × 105 M-1 s-1. The reaction is also studied for p[H+]
) 1.46-10.61 under pseudo-first-order conditions. Figure 5
shows the p[H+] dependence of kobsd. For p[H+] ) 1.46-3.02
there is no added buffer, but acetate, phosphate, and carbonate
buffers are used to study the reaction for p[H+] ) 3.5-10.6.
The studies show that the reactions have no buffer dependence.
•CO3 ion, generated by pulse radiolysis, is an even stronger
-
one-electron oxidizing agent (E° ) 1.59 V33) than is ozone (E°
) 1.01 V31). The predicted outer-sphere electron-transfer rate
(27) Beckwith, R. C.; Wang, T. X.; Margerum, D. W. Inorg. Chem. 1996,
35, 995-1000.
(28) Stanbury, D. M. Electron-Transfer Reactions; Advances in Chemistry
Series; American Chemical Society: Washington, DC, 1997; pp 165-
182.
-
There is an equilibrium between HNO2 and NO2 with pKa
) 2.97 for HNO2 at 25.0 °C and µ ) 0.50 M.26 Equation 16 is
(29) Stanbury, D. M.; de Maine, M. M.; Goodloe, G. J. Am. Chem. Soc.
1989, 111, 5496-5498.
-
Ka[NO2 ]T
(30) Stanbury, D. M.; Martinez, R.; Tseng, E.; Miller, C. E. Inorg. Chem.
1988, 27, 4277-4280.
kobsd ) k12
(16)
Ka + [H+]
(31) Bennett, L. E.; Warlop, P. Inorg. Chem. 1990, 29, 1975-1981.
(32) Huie, R. E.; Neta, P.; Ross, A. B. J. Phys. Chem. Ref. Data 1988, 17,
No. 3.
(26) Tummavuori, J.; Lumme, P. Acta Chem. Scand. 1968, 22, 2003-
(33) Huie, R. E.; Shoute, L. C. T.; Neta, P. Int. J. Chem. Kinet. 1991, 23,
541-552.
2011.

Aqueous Ozone Reactions
Inorganic Chemistry, Vol. 40, No. 17, 2001 4441
Figure 6. (a) Structure of the proposed BrOOO- intermediate from
ab initio calculations with bond distances in Å. Values of the charge
density are given in boldface and italic type. (b) Structure of the
proposed transition state [HOOOBr]q for the proton-assisted path.
constants are much larger than those for ozone, but the k12,obsd
•
-
2-
values for CO3 with I- and SO3 are factors of 5 and 45
smaller, respectively, than the k12,obsd values for O3. The k12,obsd
values for the •CO3- oxidations are larger than the k12,pred values,
but only by factors of 130 (for I-), 66 (for NO2-) and 2.9 (for
SO32-). This suggests that weak interactions (orbital overlap)
Figure 7. Structures of O3‚Xn- adducts from ab initio calculations
with bond distances in Å. Values of the charge density are given in
boldface and italic type.
•
may occur between these anions and CO3-, whereas much
stronger interactions occur with O3.
We propose that with O3 these nucleophiles form Lewis acid-
base adducts as kinetic intermediates before the transition state.
We also propose that these adducts react by oxygen-atom
transfer rather than by electron transfer. Hoigne´ has also
suggested oxygen-atom transfer reactions for NO2-, I-, and Br-
reactions with O3.2 Our kinetic data require a BrOOO-
intermediate in the reaction of Br- with O3. Ab initio calcula-
tions show that BrOOO- intermediates are possible. There are
two minimum energy conformations for the BrOOO- interme-
diate. In one Br- is weakly bound to O3, and in the other a
stronger Br-O bond forms at the expense of lengthening one
of the O-O bonds in O3 (Figure 6a). The calculations also show
that both the proton path and the water path could go through
a common transition state, as shown in Figure 6b, that leads to
the elimination of residual O2 and formation of HOBr. Aqueous
solvation will undoubtedly have a large effect on the stability
and reactivity of the intermediate; however, the calculated
structures serve as useful models for the mechanism.
Table 3. Nucleophilicity (n) for Inorganic Anions (Xn-) and
∆Hadduct Values Calculated for O3‚Xn-
∆Hadduct
,
Xn-
I-
NO2
n
kJ mol-1 c
5.04a
4.2b
-
-225.5
-487.9
-259.8
-2.9
2-
SO3
5.10a
-
HSO3
Br-
3.89a
a Swain, C. G.; Scott, C. B. J. Am. Chem. Soc. 1953, 75, 141.
b Margerum, D. W.; Schurter, L. M.; Hobson, J.; Moore, E. E. EnViron.
Sci. Technol. 1994, 28, 331-337. c This work.
Results from ab initio calculations are consistent with this
conclusion. In fact, the ab initio results suggest that the electron
transfer (eq 18) is very endothermic, while the atom transfer
(eq 17) is exothermic.
Structural changes during adduct formation also suggest that
atom transfer is more favorable. Ab initio calculations yield
structures of the intermediate adducts, as shown in Figures 6
and 7. The (BrOOO)- intermediate in Figure 6a has greatly
lengthened one of the original O-O bonds in O3, and its
cleavage will give BrO- and O2. The H+-assisted transition state
in Figure 6b again shows that this bond is greatly elongated.
Similarly by breaking the longest O-O bond, the (O3SOOO)2-
intermediate in Figure 7a, the (HOSO2OOO)- intermediate in
It is instructive to compare the energetics of atom-transfer
vs electron-transfer for an intermediate such as BrOOO- (eqs
17 and 18). In aqueous solutions, reaction 17 is preferred over
BrOOO- f BrO- + O2 (atom transfer)
BrOOO- f Br• + O3- (electron transfer)
(17)
(18)
reaction 18 by 402 kJ mol-1 (calculated from E° values34,35).
Hence, atom transfer is the overwhelmingly favorable path.
Figure 7b, and the (O2NOOO)- intermediate in Figure 7c are
poised to easily release O2 and SO42-, HOSO3-, and NO3
,
-
respectively. With such structural changes, it is difficult for
electron transfer to occur because more energy is required to
(34) Stanbury, D. M. AdVances in Inorganic Chemistry; Academic: New
York, 1989; Vol. 33, pp 69-138.
(35) Bard, A. J.; Parsons, R.; Jordan, J. Standard Potentials in Aqueous
Solution; Marcel Dekker: New York, 1985.
break the already formed X-O bond (X ) Br, S, N) and to
-
readjust the two O-O bond lengths to form O3
.

4442 Inorganic Chemistry, Vol. 40, No. 17, 2001
Liu et al.
Nucleophilicity and ∆Hadduct Correlation. If the reactions
of O3 with NO2-, HSO3-, SO32-, I-, and Br- all proceed via
the same oxygen atom transfer mechanism, then the rate constant
of these reactions should correlate with the nucleophilic
reactivities of NO2-, HSO3-, SO32-, I-, and Br- (Table 3) in
accord with the Swain-Scott relationship (eq 19). In this equation
log k12 ) sn + b
(19)
n is the nucleophilicity of the anions and s is the sensitivity of
the reaction rate constant to the strength of the nucleophile.
Figure 8 includes a Swain-Scott plot of log k12 vs n. The dashed
line is the linear regression through the data points (open
squares) with NO2-, HSO3-, SO32-, I-, and Br- on or near the
line. The correlation supports an atom-transfer mechanism for
these reactions. The slope of the plot gives s ) 5.1 and shows
that the ozone reaction is very sensitive to the nucleophilicity
of the anion.
Figure 8. Swain-Scott nucleophilicity correlation (open squares and
dashed line) and ∆Hadduct correlation (black circles and solid line) plots.
Also included in Figure 8 is a plot of log k vs -∆Hadduct
.
Acknowledgment. This work was supported by National
Science Foundation Grants ATM-9631572, CHE-9818214, and
CHE-9622683. We thank David Littlejohn (Lawrence Berkeley
Laboratory) for help in initiating these studies.
∆Hadduct is the heat of formation (from ab initio calculations)
for the adduct formed from O3 and the nucleophiles (Table 3).
The heat of formation of the adduct reflects the ability of the
nucleophiles to donate an electron pair to O3. The more negative
∆Hadduct, the stronger the interaction. An excellent correlation
is found between ∆Hadduct and the reaction rate constant for each
of the species. The solid line is the linear regression through
the data points (black circles). It matches very well with the
Swain-Scott nucleophilicity correlation plot. Both plots support
adduct formation and an oxygen atom transfer mechanism.
Supporting Information Available: Tables and figures giving
kinetic data. This material is available free of charge via the Internet
at http://pubs.acs.org.
IC000919J