Kinetics and mechanism of styrene epoxidation by chlorite: Role of chlorine dioxide

Article abstract of DOI:10.1021/ic500512e

An investigation of the kinetics and mechanism for epoxidation of styrene and para-substituted styrenes by chlorite at 25 °C in the pH range of 5-6 is described. The proposed mechanism in water and water/acetonitrile includes seven oxidation states of chlorine (-I, 0, I, II, III, IV, and V) to account for the observed kinetics and product distributions. The model provides an unusually detailed quantitative mechanism for the complex reactions that occur in mixtures of chlorine species and organic substrates, particularly when the strong oxidant chlorite is employed. Kinetic control of the reaction is achieved by the addition of chlorine dioxide to the reaction mixture, thereby eliminating a substantial induction period observed when chlorite is used alone. The epoxidation agent is identified as chlorine dioxide, which is continually formed by the reaction of chlorite with hypochlorous acid that results from ClO produced by the epoxidation reaction. The overall stoichiometry is the result of two competing chain reactions in which the reactive intermediate ClO reacts with either chlorine dioxide or chlorite ion to produce hypochlorous acid and chlorate or chloride, respectively. At high chlorite ion concentrations, HOCl is rapidly eliminated by reaction with chlorite, minimizing side reactions between HOCl and Cl₂ with the starting material. Epoxide selectivity (>90% under optimal conditions) is accurately predicted by the kinetic model. The model rate constant for direct reaction of styrene with ClO₂(aq) to produce epoxide is (1.16 ± 0.07) × 10^-2 M ^-1 s^-1 for 60:40 water/acetonitrile with 0.20 M acetate buffer. Rate constants for para substituted styrenes (R = -SO₃ ^-, -OMe, -Me, -Cl, -H, and -NO₂) with ClO₂ were determined. The results support the radical addition/elimination mechanism originally proposed by Kolar and Lindgren to account for the formation of styrene oxide in the reaction of styrene with chlorine dioxide.

Full text of DOI:10.1021/ic500512e

Article
pubs.acs.org/IC
Kinetics and Mechanism of Styrene Epoxidation by Chlorite: Role of
Chlorine Dioxide
Jessica K. Leigh, Jonathan Rajput, and David E. Richardson*
Center for Catalysis, Department of Chemistry, University of Florida, Gainesville, Florida 32611, United States
S
* Supporting Information
ABSTRACT: An investigation of the kinetics and mechanism
for epoxidation of styrene and para-substituted styrenes by
chlorite at 25 °C in the pH range of 5−6 is described. The
proposed mechanism in water and water/acetonitrile includes
seven oxidation states of chlorine (−I, 0, I, II, III, IV, and V) to
account for the observed kinetics and product distributions. The
model provides an unusually detailed quantitative mechanism for
the complex reactions that occur in mixtures of chlorine species
and organic substrates, particularly when the strong oxidant
chlorite is employed. Kinetic control of the reaction is achieved by the addition of chlorine dioxide to the reaction mixture,
thereby eliminating a substantial induction period observed when chlorite is used alone. The epoxidation agent is identified as
chlorine dioxide, which is continually formed by the reaction of chlorite with hypochlorous acid that results from ClO produced
by the epoxidation reaction. The overall stoichiometry is the result of two competing chain reactions in which the reactive
intermediate ClO reacts with either chlorine dioxide or chlorite ion to produce hypochlorous acid and chlorate or chloride,
respectively. At high chlorite ion concentrations, HOCl is rapidly eliminated by reaction with chlorite, minimizing side reactions
between HOCl and Cl2 with the starting material. Epoxide selectivity (>90% under optimal conditions) is accurately predicted by
the kinetic model. The model rate constant for direct reaction of styrene with ClO2(aq) to produce epoxide is (1.16 0.07) ×
10−2 M−1 s−1 for 60:40 water/acetonitrile with 0.20 M acetate buffer. Rate constants for para substituted styrenes (R = −SO3 ,
−
−OMe, −Me, −Cl, −H, and −NO2) with ClO2 were determined. The results support the radical addition/elimination
mechanism originally proposed by Kolar and Lindgren to account for the formation of styrene oxide in the reaction of styrene
with chlorine dioxide.
dioxide has been investigated for a variety of organic functional
groups,5,17−19 and separately the kinetics of chlorine dioxide
formation in mixtures of other chlorine oxidation states have
been studied by many researchers.20−25
INTRODUCTION
■
Chlorine dioxide is a free radical known for its oxidative
selectivity and stability in aqueous solution over a wide pH
range. It is reversibly reduced to chlorite ion (eq 1, E° = 0.954
V vs normal hydrogen electrode at 25 °C),1−3 and much of the
oxidative chemistry of chlorine dioxide has been attributed to
Over the last 30 years, chlorine dioxide has largely replaced
chlorine as the dominant agent for the bleaching of wood pulp
because of its lower tendency to form chlorinated products in
the oxidation of organic compounds,8,9,26 significantly reducing
the release of organochlorine pollutants into the environment.
Several studies27−29 concerning the reactions in “ClO2 only”
(elemental chlorine free (ECF)) pulp bleaching have shown
that hypochlorous acid is produced as a byproduct of the
reaction with lignin, and HOCl is now considered to be
responsible for chlorination products in ECF pulp bleaching,
although at a much-reduced level compared to traditional
chlorine bleaching. Conjugated C−C double bonds are a
common feature of organic chromophores in natural pigments
in textiles and wood pulp,30 so the present work contributes to
an improved understanding of that complex oxidative
chemistry.
ClO2(aq) + e− ⇌ ClO2−(aq)
(1)
this one-electron transfer reaction.4,5 The relatively low
electrode potential leads to good oxidative selectivity when
compared to indiscriminate radical oxidants such as OH. In
contrast to Cl2, ClO2 does not hydrolyze in solution and is an
active oxidant in the pH range of approximately 4−10.6,7
Because of its selectivity and several other advantages of
chlorine dioxide over chlorine, chlorine dioxide has become a
widely used oxidant and disinfectant in industrial processes
such as wood pulp bleaching, water treatment, and many
others.6,8−10 It also has excellent antimicrobial properties,
which have led to its use as a decontaminant of medical devices,
laboratory equipment, and persistent biological warfare agents
in buildings.6,8,10 Its usefulness as a cold sterilant has received
increasing interest in recent years as a consequence of the
emergence of new “superbugs” such as MRSA, norovirus,
Legionella, and CRE.11−16 The reactivity of pure chlorine
Received: March 4, 2014
Published: June 13, 2014
© 2014 American Chemical Society
6715
dx.doi.org/10.1021/ic500512e | Inorg. Chem. 2014, 53, 6715−6727

Inorganic Chemistry
Article
−
As will be shown, the strong oxidant chlorite (ClO2 ) is an
important common reactant in mixtures that maximize the
advantage of chlorine dioxide reactivity with certain functional
groups (alkenes in the present case). Chlorite ion has a rich
oxidative chemistry with organic compounds, most notably in
the conversion of aldehydes to carboxylic acids.31−35 Chlorite is
also used in the textile bleaching industry, where it offers several
advantages over chlorine bleach, but the chemistry of chlorite
bleaching has not been investigated in appreciable detail to our
knowledge.36−38 Chlorite is not a direct oxidant for epoxidation
of alkenes but can serve as a source of oxidizing
equivalents.39−41 One goal in the present work was to
demonstrate how the selectivity and utility of chlorite can be
enhanced as an oxidant in epoxidation applications via a
thorough understanding of the reaction mechanism. We will
show how the high oxidizing power of the chlorite ion can be
successfully coupled to the selectivity of ClO2 for the
epoxidation of alkenes.
by introducing a small amount of chlorine dioxide in solution
via either direct addition or reaction of an aldehyde with
chlorite (which rapidly forms chlorine dioxide). The most likely
mechanism was assumed to be that of Kolar and Lindgren, in
which intermediate chlorine dioxide is the direct oxidant and
chlorite is the ultimate source of oxidizing equivalents. Jangam
and Richardson also proposed that hypochlorous acid formed
in the reaction is responsible for the production of undesirable
byproducts, but high concentrations of chlorite ion were used
to react with hypochlorous acid and minimize these side
reactions. It was noted that the proposed radical addition/
elimination mechanism is analogous to that57−59 for the
epoxidation of alkenes by organoperoxyl radicals (ROO·),
which are isoelectronic with ClO2.
We investigated the detailed kinetics and mechanism of the
selective epoxidation of styrene by excess chlorite in the pH
range of 5−6. To build a complete mechanism for the
epoxidation reaction it was first necessary to develop a
quantitative model for the underlying chlorine reactions
under our conditions. This chemistry is the source of
complications in the interpretation of chlorine dioxide reactions
with organic substrates, as chlorine species produced in the
reactions can oxidize and/or chlorinate the substrate, producing
numerous side products and reducing yields. A quantitative
kinetic understanding of the chlorine chemistry allows us to
account for reactions involving the formation of styrene oxide
and various reaction byproducts. We propose a mechanism that
can be used to predict selectivity for the epoxide and the
distribution of side-products under a wide range of initial
conditions. Notably, we find that it is necessary to account for
seven oxidation states of chlorine (−I, 0, I, II, III, IV, and V),
thereby vividly illustrating the well-known complexities of
chlorine oxidation chemistry. We also provide evidence for the
nature of the rate-determining step in the epoxidation reaction.
The resulting mechanism and associated rate constants can
serve as the basis for understanding and modeling complex
reactions of chlorite/chlorine mixtures with other reactants,
including biomolecules and pollutants.
Introduction of chlorite into equilibrated chlorine solutions
leads to formation of chlorine dioxide via oxidation of
chlorite22−25,42 and, to an extent dependent on acidity, the
disproportionation of chlorous acid.43−45 We have used the
previous mechanisms for the former reaction proposed by
Peintler and co-workers23 and Margerum and co-workers22 in
our model. We show here that it is also necessary to
incorporate other reactive oxidation states (specifically
chlorine(II), chlorine(III), and chlorine(IV)) into a fully
quantitative model for substrate oxidation that can account
for product distributions under various initial conditions.
Three general pathways for chlorine dioxide oxidation
reactions with electron-rich organic compounds have been
proposed in the literature.8,17,41,46 The initial reactions in
chlorine dioxide oxidations have typically been described as
either one-electron oxidations or hydrogen abstrac-
tions.8,17,46−53 The former mechanism is commonly invoked
for reactions with most organic compounds, leading to
formation of radical organocations in reactions of the oxidant
with neutral organic molecules. Hydrogen abstraction reactions
are relatively uncommon since the OClO−H homolytic bond
dissociation energy is rather low, estimated to be 71 kcal/
mol.54−56
A less commonly invoked O atom transfer pathway of
particular interest in the present study was proposed by Kolar
and Lindgren in 1982 and consists of radical addition of ClO2
to the substrate followed by elimination of the ClO radical.41 In
their investigation of the oxidation of excess styrene by chlorine
dioxide, Kolar and Lindgren observed the formation of styrene
oxide in addition to a number of chloro-substituted oxidation
products and chlorinated styrenes. Under the conditions of
their experiments, however, selectivity for styrene epoxidation
was very low. Geng et al.39 described an improved method for
epoxidation of styrene by chlorite in a mixed solvent with
heating and reported a yield of up to 77%. They showed that
chlorine dioxide is likely to be the reactive species and cited the
Kolar and Lindgren mechanism to explain the cis to trans
isomerization observed in the epoxidation of cis-stilbene.
In 2010, Jangam and Richardson reported that high
conversions (>99%) and selectivities as high as 89% could be
achieved for oxidation of a variety of alkene substrates
(including styrene) to epoxides at room temperature with
excess chlorite ion in appropriate cosolvent/buffer mixtures.40
They showed oxidation by chlorite ion alone proceeds only
after a lengthy induction period, but this delay was eliminated
EXPERIMENTAL SECTION
■
Reagents. Water was purified by using a Barnstead/Thermolyne E-
Pure Model 04361 system to achieve a specific resistance of 18.2 MΩ
cm. Aqueous chlorine dioxide solutions were prepared via the reaction
of acetic anhydride and sodium chlorite in purified water.60,61 The
purity of aqueous ClO2 solutions was assessed by ion chromatography,
and the stock solutions were stored away from light and refrigerated
between 0 and 4 °C. Chlorine dioxide concentrations were
standardized by UV−visible spectrophotometry (Hewlett-Packard
8453) at 359 nm (ε = 1250 M−1 cm−1) before each use in reaction
solutions. Technical grade sodium chlorite (80%) was purchased from
Fisher Scientific and recrystallized as described in the literature.22,23 Its
purity was verified by ion chromatography, and stock solutions were
standardized spectrophotometrically at 260 nm (ε = 154 M−1 cm−1).60
Sodium hypochlorite (5% active chlorine, Fisher Scientific) was
acidified, and hypochlorous acid was purified by fractional distillation
under reduced pressure.62,63 Stock solutions of HOCl were stand-
ardized spectrophotometrically at 292 nm (ε = 362 M−1 cm−1)60 and
buffered (acetic acid/sodium acetate) to obtain solutions of
hypochlorous acid. All other chemicals/reagents (styrene, deuterated
styrene, p-substituted styrene derivatives, acetic acid, sodium acetate,
and naphthalene) were purchased from commercial suppliers, and
their purity was assessed by NMR. No further purification was
necessary, and fresh stock solutions were prepared prior to
experiments. A Dionex Ion Chromatography (IC) System 1500 was
6716
dx.doi.org/10.1021/ic500512e | Inorg. Chem. 2014, 53, 6715−6727

Inorganic Chemistry
Article
Table 1. Proposed Mechanism for the Oxidation of Styrene by Chlorite and Chlorine Dioxide
a
b
In aqueous solution. Values also used for mixed solvents (see text). For 0.20 M acetate buffer, based on forward rate constant kfM1 = 434 M−1 s−1
c
(eq 4) and kM8/kfM1 = (3.7 0.2) × 103 M−1 (see text). Unless otherwise noted, rate constants are estimated for 60:40 water/acetonitrile with 0.20
M acetate buffer at T = 25 °C. Error limits are from literature cited or from this work as noted. Estimated standard error for reaction M5 obtained
from fits to experimental data for the chlorite/hypochlorous acid reaction with no substrate. Estimated errors for reactions 10, 13, and 14 obtained
d
e
from fits to substrate oxidation experimental data with kM5 = 2.86 × 106 M−2 s−1. This work. Best fit simulation rate constant is given for 60:40
f
H2O/CH3CN and 0.20 M acetate buffer. Literature value for M14 in aqueous solution is 7 × 109 M−1 s−1. See text for other conditions. kM6/kM9
=
g
h
(5.4 0.2) × 104 M−1. Set to this value in all fits. Error limits on rate constant could not be estimated from fit because of low yield of product. The
i
value given is an estimate used in fits. M15 set to the literature value above in all fits. The estimated value of kM14, therefore, is dependent on this rate
constant as it determines the M14/M15 branching ratio (see text). Value determined from experiments in D2O with 0.20 M acetate buffer.
j
used to monitor the concentrations of anionic inorganic chlorine
species (chlorite, chloride, and chlorate).
HOCl/ClO2 Reaction. The reactions of hypochlorous acid and
Naphthalene (5 mM) was added to the solution as an internal
standard for quantitative GC analysis. Solutions were buffered (by
adding the necessary amount of acetic acid), and the pH was verified
by a Fisher Scientific Accumet Basic model AB15 digital pH meter
equipped with a Thermo Orion 911600 electrode. Solution temper-
atures were maintained at 25.0 0.1 °C throughout the reactions. The
reactions were initiated with the addition of chlorine dioxide to the
solution to achieve a final concentration of ∼10 mM. Reactant and
product concentrations were analyzed over a 6 h period at regular
intervals. Diluted aliquots from the reaction mixture were analyzed to
determine organic and inorganic reactant/product concentrations over
the course of the reaction. Chlorite, chloride, and chlorate were
monitored by ion chromatography. UV−visible spectrophotometry
was used to monitor the chlorine dioxide concentration at 359 nm.
The organic species (styrene, styrene oxide, and various chlorinated
products) were characterized by gas chromatography−mass spectrom-
etry (GC-MS) and by chromatographic comparison with pure
standards by a Varian model CP-3800 gas chromatograph (GC)
equipped with a flame ionization detector and a J & W Scientific DB-
35MS Capillary Column, 0.5 μm, 30m × 0.53 mm. Diluted aliquots
from the reaction mixture were analyzed to determine the inorganic
reactant and product concentrations as described above. Organic
species concentrations were calculated using a ratio of the styrene peak
area to the naphthalene (internal standard) peak area:
−
sodium chlorite were carried out under excess chlorite conditions
(4.74 × 10−3 M sodium chlorite and 1.50 × 10−4 M hypochlorous
acid) in water or 60/40 water/acetonitrile. Solutions were prepared in
a 1.0 cm path length quartz cuvette, with a Teflon stopper and a total
reaction volume of 3.2 mL. Reaction temperatures were maintained at
25.0 0.1 °C via a Fisher Scientific Isotemp Refrigerated Circulator
(Model 901) and a thermostated cell holder in a Hewlett-Packard
8453 UV−visible spectrophotometer. Reactions were investigated in
0.50 M sodium acetate in water, 0.20 M sodium acetate in water, and
0.20 M sodium acetate in a water/acetonitrile mixture at three pH
values (5.25, 5.55, and 5.95−stock solutions of sodium acetate were
prepared and buffered by adding the appropriate amount of acetic
acid). The reactions were initiated by the addition of sodium
hypochlorite (in 0.50 or 0.20 M sodium acetate) to the reaction
solution. Reaction kinetics were evaluated by monitoring the
production of chlorine dioxide by UV−visible spectrophotometry at
359 nm. Control experiments confirmed that photochemical
decomposition in the spectrophotometric beam was negligible under
our conditions.
Styrene Oxidations. The reaction mixture for the oxidation of
styrene was 0.005−0.075 M sodium chlorite, 0.0125 M substrate, and
0.200 M sodium acetate in a 60:40 water/acetonitrile solvent.
6717
dx.doi.org/10.1021/ic500512e | Inorg. Chem. 2014, 53, 6715−6727

Inorganic Chemistry
Article
and chlorite ion.22−25 This step is central to the chain
mechanism proposed by Kolar and Lindgren41 and is of
particular importance to our study of the selective oxidation of
styrene under excess chlorite conditions described in detail
below.
⎛
⎝
styrene peak areat
⎜
⎜
[styrene]t = 12.5 mM ×
naphthalene peak area
t
⎞
styrene peak area0
⎟
⎟
÷
naphthalene peak area0
⎠
HOCl is formed at equilibrium by the well-known hydrolysis
of Cl2 (eq 2). The reaction of
[producta]t = (12.5 mM − [styrene]t )
Cl2 + H2O ⇌ HOCl + Cl− + H+
(2)
⎛
⎞
productapeak areat
⎜
×
⎟
⎟
−
2ClO2 + HOCl + H+ → 2ClO2 + Cl− + H2O
⎜
(3)
total product peak area
⎝
⎠
t
−
HOCl with ClO2 produces chlorine dioxide (eq 3), thereby
4-Styrenesulfonate Oxidations. The oxidation of 4-styrenesul-
fonic acid was studied in D2O solution, and the reaction mixture
contained 0.010−0.075 M sodium chlorite, 0.0125 M acetonitrile
(used as an internal standard), and 0.200 M sodium acetate buffered at
pH 5.25 with the addition of acetic acid. Before the addition of
substrate to the reaction mixture, approximately 0.010 M chlorine
dioxide was generated in situ by the reaction of 0.005 M benzaldehyde
and 0.015 M sodium chlorite.19,64 The oxidation reaction was initiated
by the addition of approximately 0.0125 M sodium 4-styrenesulfonate,
and the reaction was monitored over a 5−7 h period at regular
intervals. Inorganic products were monitored as described for styrene
reactions. Organic reactant and product concentrations were analyzed
by a 500 MHz Inova NMR spectrometer, and quantitative data were
obtained using a ratio of the integrations of the reactant/products to
the internal standard (acetonitrile, 0.0125 M).
reducing the concentration of HOCl in solution and increasing
the availability of the epoxidizing agent ClO2. It was our
hypothesis that the resulting low steady-state concentration of
HOCl, when compared to that of chlorine dioxide, reduces the
probability of side reactions of HOCl with the substrate that
would otherwise produce undesirable chlorinated byproducts.41
In addition, the lessened formation of Cl2 from HOCl and Cl−
by the reverse of eq 2 would presumably lead to reduced
amounts of other chlorination products. The addition of
chlorite to suppress HOCl side reactions in pulp bleaching has
been proposed by Ni and co-workers.69
The kinetics and stoichiometry of the HOCl/ClO2− reaction
and the related reaction of Cl2 with ClO2− have been described
extensively in the literature.22,23,25,70−73 In 1990 Peintler et al.
proposed a mechanism that modeled their experimental results
for chlorine dioxide formation under a variety of initial
conditions.23 In 2008, the 1990 mechanism was expanded by
Para-Substituted Styrenes. p-Chlorostyrene, p-methylstyrene, p-
methoxystyrene, and p-nitrostyrene were investigated in 60:40 water/
−
acetonitrile with [ClO2 ]0 = 75 mM, pH 5.25 sodium acetate buffer
(200 mM in water), and [ClO2]0 = 1.5 mM (p-methoxystyrene), 20
mM (p-nitrostyrene), or 10 mM (p-chlorostyrene, p-methylstyrene);
substrate concentrations were in the 5−12 mM range, as required to
achieve solution homogeneity. The rate of consumption of substrate,
the formation of organic products, and the rate of formation of ClO2
were typically monitored by GC and UV−visible spectrophotometry,
respectively (ClO2 concentrations for nitrostyrene could not be
determined due to the overlap of absorbance due to the substrate).
Second-order rate constants for the rate equation R = k [ClO2]-
[styrene] were estimated by determining the reaction rates at multiple
time intervals that had essentially constant [ClO2] and averaging. As a
check, the mechanism of Table 1 was also used to model the reaction
curves numerically, and the rate constants for M10−M12 were
optimized to achieve the best fit to the concentration versus time
curves for all of the substrates (this method alone was used to obtain
the p-nitrostyrene rate constant). For comparison, the reactions were
also investigated under different conditions in a solvent that allows
higher substrate concentrations and uses an alternative buffer, with
Korman
́
yos et al.25 to include additional steps involved in
catalysis by chloride ion. Margerum and co-workers have
reported the large impact of buffer catalysis on the rate of
chlorine dioxide formation (specifically for acetic acid/
acetate)22 and the rapid reaction of Cl2 with chlorite.42 In the
pH range employed in the present work, the contribution of the
acid-catalyzed disproportionation of chlorite to the production
of ClO2 is negligible.43−45
−
The mechanism we use for the HOCl/ClO2 reaction is
summarized in Table 1 (steps M1−M9). The equilibrium
constant for reaction M1 was set at KM1 = 6.1 × 10−4 M2. The
value of the forward rate constant for equilibrium M1(kfM1) was
calculated based on the work of Nicoson and Margerum42 that
shows the rate of chlorine disproportionation at 25.0 °C is
dependent on the acetate concentration, as summarized in eq 4.
−
[ClO2 ] = 170 mM, pH 7 phosphate buffer (100 mM in water), a
kfM1 = 22 4 s−1 + 2060 30[Ac−] s−1
(4)
20:55:25 acetonitrile/buffer/ethanol mixed solvent, 25 mM substrate,
and small amounts (∼2 mM) of ClO2 to eliminate the induction
period.
The rate constant for reaction M8 was determined from the
ratio of kM8/kfM1 = (3.7 0.2) × 103 M−1 as determined by
Nicoson and Margerum.42 The rate constant for reaction M9 is
Numerical Methods. All kinetic models were analyzed using
numerical methods to produce predicted concentration versus time
curves for reactants, products, and intermediates. Gear integration was
used,65 and the numerical integration models were supplemented by
fitting routines based on a combination of simplex66 and Marquardt67
optimizations to determine the best fit rate constants. Estimated
parameter errors (for model rate constants) were determined using
numerical matrix methods.68 Further details are given in the
Supporting Information.
set at 15 M−1 s−1, and kM6 was calculated using the ratio of kM6
kM9 = (5.4
/
0.2) × 104 M−1.25 For reversible reactions the
corresponding reverse reaction rate constants were set by the
value of the equilibrium constants.
Numerical integration of steps M1−M9 reproduced the
calculated concentration versus time curves in the work of
Peintler et al. (1990) and Korman
́
yos et al. (2008)23,25 when
−
ClO2 is in excess and no initial chloride is present. Although
■
the Korman
́
yos et al. mechanism incorporates the impact of Cl−
RESULTS AND DISCUSSION
−
The HOCl/ClO2 Reaction. Reactions involving multiple
on the rate constant for the direct reaction of HOCl with
−
chlorine oxidation states in solution result in significant
complexity in building a kinetic model for chlorite oxidations
of organic substrates. We began our modeling with a review of
the much-investigated reaction of hypochlorous acid (HOCl)
HClO2/ClO2 , we used the 1990 Peintler et al. mechanism
because the generation of Cl− in styrene epoxidation has a
negligible effect on the overall reaction rates for the
disappearance of reactants and generation of products (chloride
6718
dx.doi.org/10.1021/ic500512e | Inorg. Chem. 2014, 53, 6715−6727

Inorganic Chemistry
Article
catalysis is significant in the overall rate only at much higher
initial concentrations of chloride). Acetate/acetic acid (M4) is
used only as a buffer in our mechanism and was included in the
simulations. The reactions in the catalysis of M5 by acetic acid
described in detail by Jia et al.22 are not explicitly included in
our reaction steps but are modeled by appropriate choices for
the rate constants M1, M5, and M8, as discussed below.
Therefore, the choice of pKa for acetic acid74 is not crucial to
the mechanism used here.
As described below the conditions for the styrene
epoxidation (i.e., addition of a cosolvent, decreased buffer
concentration) required us to reset the value of the rate
constant for reaction M5 in Table 1 for our reaction conditions.
The mechanism of the sodium acetate/acetic acid effect on
reaction M5 was investigated by Jia et al.22 We used the
mechanism of Jia et al. to assess the appropriateness of a single
kM5 value to fit the observed yield of chlorine dioxide and its
rate of production under the conditions used for the
epoxidation of styrene. While an increase in pH will decrease
[HAc] with [Ac−] constant, thereby altering the value of kM5
according to the rate laws of Jia et al., the extent to which the
value of the effective rate constant for M5 is altered within the
pH range from 5.25 to 5.95 is <10%. Therefore, for simplicity, a
single value of kM5 was used to fit our experimental data, as in
the work of Peintler et al.23 Note that models for the reaction
outside of the pH range of this study would require inclusion of
the effects of pH on acetate catalysis.
actual uncertainty in the rate constant to be around 10% in the
pH range of 5−6.
Effect of Solvent Composition. Under the conditions
−
described above for the HOCl/ClO2 reactions (i.e., aqueous
sodium acetate buffer) a heterogeneous mixture resulted upon
addition of styrene and naphthalene (internal GC standard) to
the reaction solution. A homogeneous solution can be obtained
by using a cosolvent of 60:40 water/acetonitrile and a lower
concentration of buffer (0.20 M acetate). Since the rate of
chlorine dioxide production was observed to increase in
acetonitrile/water compared to water, it was also necessary to
accommodate this acceleration in the kinetic model.
To account for the cosolvent reaction conditions, only the
rate constant for reaction M5 was adjusted with the
understanding that the resulting kM5 value effectively
incorporates any shift in the relevant acid−base equilibria
(M1−M3) resulting from addition of acetonitrile. A value of
kM5 = (2.86 0.04) × 106 M−2 s−1 results from a fit of the
mechanism to all experimental results in the mixed solvent
reactions (Figure 1). Although limited in useful pH range, the
mechanism successfully models the kinetics and stoichiometry
−
of the HOCl/ClO2 reaction in the mixed solvent used for
epoxidations in this work.
Epoxidation of Styrene. The reaction of styrene with
chlorite was investigated in buffered 60:40 water/acetonitrile
with [Ac−] = 0.20 M. The characteristic induction period for
chlorite oxidation was avoided by adding ClO2 to the reaction
mixture.40 With excess chlorite, styrene oxide is the major
organic product, and small amounts of 1,2-dichloroethylben-
zene, 2-chloro-1-phenyl ethanol, and 2-chlorovinylbenzene are
also produced (an example GC is shown in Supporting
Information). Although phenylacetaldehyde was observed by
GC, it is well-known that rearrangement of styrene oxide to
phenylacetaldehyde occurs in the injector port.76,77 NMR was
used to monitor the reaction over the course of 6 h under the
same conditions (in D2O/d3-acetonitrile) to determine if the
aldehyde was also produced as a result of styrene oxidation by
chlorite. Phenylacetaldehyde was not observed in the NMR
analysis, confirming that the aldehyde observed by GC is a
result of styrene oxide rearrangement in the injector port.
Aldehyde and epoxide peaks in GC were therefore combined to
give the epoxide yield.
Although we did initial comparisons to the Peintler et al.
study at [Ac−] = 0.50 M (see Supporting Information), the
epoxidation reactions required an acetate buffer concentration
of 0.20 M. We find kM5 = (9.41 0.09) × 105 M−2 s−1 for
[Ac−] = 0.20 M based on fits to our experimental data; Figure 1
shows the fits obtained by using this value. The error limits
from fits for kM5 reflect the statistical analysis from the least-
squares model but do not account for the expected variation
with pH; therefore, it would be appropriate to consider the
−
Results for a typical kinetic experiment at high [ClO2 ] (74
mM) and pH 5.25 are shown in Figures 2 and 3. Chloride and
chlorate were the major inorganic chlorine products, and the
concentration of chlorine dioxide rose slowly from the initial
concentration over the course of the reaction. See Supporting
Information for kinetic data and fits at pH 5.88.
The mechanism of Table 1 was used to fit the styrene
oxidation kinetics by using kfM1 = 434 s−1 (eq 4), kM5 = 2.86 ×
106 M−2 s−1 (as obtained above), and kM8= 1.61 × 106 M−1 s−1
(kM8 = 3.7 × 103 M−1 kfM1). In addition to the previously
established inorganic reactions, reactions were added producing
styrene oxide, 1,2-dichloroethylbenzene, 2-chlorovinylbenzene,
and 2-chloro-1-phenyl ethanol (M10−M13 in Table 1,
respectively).
Fitting the pH dependence of reaction M13 required acid
catalysis, which is common for chlorohydrin formation from
HOCl and alkenes.78,79 The reactions of HOCl and Cl2 with
hydrocarbons can involve a number of reactive intermedi-
ates78,80 so the reactions M10−M13 must be considered
tentative; however, as discussed below they are adequate to
Figure 1. Experimental (symbols) and simulated (solid lines) [ClO2]
vs time in 100% water or 60:40 water/acetonitrile (mix). Conditions
−
(100% H2O): [ClO2 ]0 = 4.74 mM, [HOCl]0 = 0.161 mM, [Ac−] =
200 mM; simulations obtained using kM5 = 9.41 × 105 M−2 s−1. T = 25
−
°C. Conditions (60:40 H2O/CH3CN): [ClO2
]
= 4.74 mM,
0
[HOCl]0 = 0.15 mM, [Ac−] = 200 mM; simulated data obtained
using kM5 = 2.86 × 106 M−2 s−1. T = 25 °C. The pH values are those of
the aqueous buffer used in the mixed solvent.
6719
dx.doi.org/10.1021/ic500512e | Inorg. Chem. 2014, 53, 6715−6727

Inorganic Chemistry
Article
and rate constants were obtained by optimization for reactions
M10−M13 to give the best fit to the experimental data (Table
1). The observed stoichiometries of the reactions were as
predicted by the mechanism, and observed oxidation electron
equivalents for all inorganic and organic species were equal to
reduction electron equivalents within experimental errors. All
optimizations and statistical analyses for rate constant
parameters were done by fitting all measured reactant and
product concentrations in all data sets (for various pH values
and initial concentrations of reactants) simultaneously.
Uncertainties for the styrene reaction rate constants were
deduced using statistical methods described in the Supporting
Information are reported in Table 1; for reactions with low
product yield (chlorinations), the estimated rate constants have
large uncertainties, and the values reported in Table 1 are
simply those found in the fitting routines and should be
considered rough estimates. The ratio of kM14/kM15 in the
optimized fit was found to be 10.0 0.4, which is close to the
literature value for kM14/kM15 of ∼7 from independent
determinations of the rate constants in aqueous solutions.21,75
According to the mechanism proposed by Kolar and
Lindgren41 for the epoxidation of styrene, the reaction of 1
equiv of chlorine dioxide adds to the vinyl group, producing a
radical intermediate (1).
Figure 2. Experimental (symbols) and simulated (solid lines) organic
−
reactant/product concentrations: [ClO2]0 = 10.1 mM, [ClO2 ] = 73.6
mM, [styrene] = 12.5 mM, [Ac−] = 200 mM buffered at pH 5.25;
60:40 water/acetonitrile mixed solvent; simulated data obtained using
kM5 = 2.86 × 106 M−2 s−1. T = 25 °C. The pH value is that of the
aqueous buffer used in the mixed solvent. Abbreviations: S−Cl2 = 1,2-
dichloroethylbenzene, S-HOCl = 2-chloro-1-phenyl ethanol, and S−Cl
= 2-chlorovinylbenzene.
Free rotation around the C−C bond in 1 has been invoked in
prior studies39,40 to explain the observed isomerization in the
epoxidation of cis-stilbene to a mixture of cis- and trans-stilbene
oxide. The intermediate undergoes elimination of the chlorine
oxide radical (ClO) to form the epoxide (M10) as shown in
Figure 4.41 Alternatives to this mechanism for this step are
considered in a later section.
Figure 4. Epoxidation of styrene by radical addition/elimination
mechanism. This mechanism for reaction M10 as an elementary step is
supported by the low sensitivity of the reaction rate to para
substituents on the styrene.
Figure 3. Experimental (symbols) and simulated (solid lines)
inorganic reactant/product concentrations: [ClO2]0 = 10.1 mM,
−
[ClO2 ] = 73.6 mM, [styrene] = 12.5 mM, [Ac−] = 200 mM buffered
at pH 5.25; 60:40 water/acetonitrile mixed solvent; simulated data
obtained using kM5 = 2.86 × 106 M−2 s−1. T = 25 °C. The pH value is
that of the aqueous buffer used in the mixed solvent.
On the basis of the proposed mechanism, decreasing the
initial amount of chlorite in the reaction should increase the
extent of HOCl reaction with substrate via reaction M13,
resulting in decreased selectivity for the epoxide. For example,
with 5 mM initial chlorite the observed and simulated results
(Figure 5) show the expected decrease in the amount of
epoxide produced and an increase in 2-chloro-1-phenyl ethanol,
and the selectivity for epoxide is only ∼60%.
The corresponding inorganic data and fits (Figure 6) for a
low chlorite reaction show chlorate as the major product
instead of chloride, which is the dominant product with excess
chlorite. In addition, the initial chlorine dioxide is consumed
during the low chlorite reaction, while, in contrast, chlorine
model the observed byproduct distributions in these chlorite
oxidations.
Two additional known reactions were added to model the
fast reactions of the chlorine oxide radical with chlorine dioxide
and chlorite (M14 and M15). Changes to the ratio of the
literature rate constants21,75 for these reactions were allowed to
best fit the experimental data. For purposes of the fits, the value
of kM15 was set to a fixed value (Table 1) and kM14 was
optimized.
All reactions M1−M15 were used to obtain simulated
concentration versus time profiles via numerical integration,
6720
dx.doi.org/10.1021/ic500512e | Inorg. Chem. 2014, 53, 6715−6727

Inorganic Chemistry
Article
dioxide is generated under excess chlorite conditions (Figure
3).
Epoxidation of 4-Styrenesulfonic Acid. The oxidation of
−
4-styrenesulfonic acid by ClO2 under high and low ClO2
conditions was examined in aqueous solution to remove the
influence of added nonaqueous solvent. An induction period
was eliminated by using benzaldehyde to produce ClO2 via its
reaction with chlorous acid19,64 in situ prior to the initiation of
the reaction. The organic reactants and products were
monitored by NMR; therefore, D2O was used as the solvent.
To assess the effect of D2O on the background inorganic
chlorine reactions, the HOCl/ClO2− reaction was carried out as
described above but in D2O. A comparison of the rate of ClO2
production in H2O and D2O demonstrated that D2O had a
negligible effect on the background chlorine kinetics. Addition-
ally, calculations of pH and pKa values in D2O81 (vs H2O)
revealed only small variations from H2O to D2O. Therefore, it
was only necessary to change kM5 to 9.41 × 105 M−2 s−1,
−
Figure 5. Experimental (symbols) and simulated (solid lines) organic
reflecting the fit for the HOCl/ClO2 reaction in water with
−
0.20 M Ac− buffer as described above, and to optimize the
substrate-containing steps (M16 and M17 in Table 2) and M14
to simulate the experimental data.
reactant/product concentrations: [ClO2]0 = 9.73 mM, [ClO2 ] = 5.06
mM, [styrene] = 12.5 mM, [Ac−] = 200 mM buffered at pH 5.25;
60:40 water/acetonitrile mixed solvent; simulated data obtained using
kM5 = 2.86 × 106 M−2 s−1. T = 25 °C. The pH value is that of the
aqueous buffer used in the mixed solvent. Abbreviations: S−Cl2 = 1,2-
dichloroethylbenzene, S-HOCl = 2-chloro-1-phenyl ethanol, and S−Cl
= 2-chlorovinylbenzene.
The epoxide (4-oxirane-phenylsulfonate) and chlorohydrin
(4-(2-chloro-1-hydroxyethyl)phenylsulfonate) products were
observed for the oxidation of 4-styrenesulfonate along with 4-
(1,2-dihydroxyethyl)phenylsulfonate (the hydrolysis product of
the epoxide). Other chlorination products were not observed,
but the expected yields would not be measurable via NMR in
light of the lower sensitivity compared to GC. Good fits of the
available experimental data were obtained for both the excess
and low chlorite conditions (see Figures 7 and 8 and the
Supporting Information) using the rate constants in Table 2 in
place of M10 and M13. The ratio of kM14/kM15 under the
aqueous reaction conditions was found to be 4.2
0.2
(somewhat lower than the value of 10.4 in the 60:40 mixed
solvent used in the styrene oxidations). For purposes of fitting
M16, the observed amounts of epoxide and its hydrolysis
product (diol) were combined. The epoxide (4-oxirane-
phenylsulfonate) selectivities under low and excess chlorite
conditions are ∼60% and 90%, respectively.
Mechanism of Styrene Epoxidation. The overall
proposed mechanism is illustrated in Figure 9. The main goal
of our study was to model the much higher selectivity for the
room-temperature styrene epoxidation observed when chlorite
is in large excess, as observed in our previous work.40 Under the
optimal conditions with high chlorite concentrations described
above, the reaction is >90% selective for epoxide. Jangam and
Richardson40 report a similar selectivity of 89% for styrene
oxide under their conditions.
Figure 6. Experimental (symbols) and simulated (solid lines)
inorganic reactant/product concentrations: [ClO2]0 = 9.73 mM,
−
[ClO2 ] = 5.06 mM, [styrene] = 12.5 mM, [Ac−] = 200 mM buffered
at pH 5.25, 60:40 water/acetonitrile mixed solvent; simulated data
obtained using kM5 = 2.86 × 106 M−2 s−1. T = 25 °C. The pH value is
that of the aqueous buffer used in the mixed solvent.
Table 2. 4-Styrenesulfonic Acid Reactions and Rate Constants in 100% Aqueous Solution
6721
dx.doi.org/10.1021/ic500512e | Inorg. Chem. 2014, 53, 6715−6727

Inorganic Chemistry
Article
(M5) 2ClO2 + HOCl + H+ → 2ClO2 + Cl−+H2O
−
(7)
−
−
(net) 2ClO2 + S → SO + Cl− + ClO3
(8)
The second chain (eqs 9−12) arises as a consequence of the
higher chlorite concentration used in our conditions, diverting a
substantial fraction of ClO to reaction with chlorite (M15).
(M10) ClO2 + S → SO + ClO
(9)
−
(M15) ClO + ClO2 → ClO2 + OCl−
(10)
(M3) OCl− + H+ → HOCl
(11)
−
(M5) 2ClO2 + HOCl + H+ → 2ClO2 + Cl− + H2O
(12)
Figure 7. Experimental (symbols) and simulated (solid lines) organic
(net) S + 3ClO2 + H+ → SO + 2ClO2 + Cl− + H2O
−
−
reactant/product concentrations: [ClO2]0 = 7.39 mM, [ClO2 ] = 72.1
mM, [4-styrenesulfonic acid] = 13.1 mM, [Ac−] = 200 mM buffered at
pH 5.25 ; 100% D2O; simulated data obtained using kM5 = 9.41 × 105
M−2 s−1. T = 25 °C. Abbreviation: S-HOCl = 4-(2-chloro-1-
hydroxyethyl) phenylsulfonate). Epoxide concentration is sum of
epoxide and the diol hydrolysis product.
(13)
Reaction M15 was suggested by Kolar and Lindgren41 as an
alternative pathway in high chlorite conditions. The net
reaction of the second chain (eq 13) produces chloride as
the sole inorganic byproduct. The difference in predicted
stoichiometries for the chlorine products in eqs 8 and 13 allows
a quantitative assessment of the relative contributions of the
two chains in our study. The net reaction of eq 8 does not
consume ClO2, and eq 13 produces the increase in ClO2 noted
for high initial chlorite reactions.
The branching fraction based on the relative rates of steps
M14 and M15 is affected by the concentrations of ClO2 and
−
ClO2 . The ratio of RM15/(RM15 + RM14) when half of the
substrate has been consumed ([styrene] = [styrene]0/2) is
−
shown as a function of initial values for [ClO2]0 and [ClO2 ]0
in Figure 10. As [ClO2]0 values increase the branching fraction
generally decreases due to increased availability of ClO2 for
−
reaction M14 (eq 6). At very low values of [ClO2 ]0 (i.e., such
as 5 mM chlorite in Figure 10) with higher [ClO2]0 values,
reaction M14 dominates since chlorine dioxide is more readily
available to react with the ClO intermediate. However, with
−
both low [ClO2]0 (<5 mM) and low [ClO2 ]0 (5 mM) the
reaction does not go to completion, and reaction M15
eventually dominates. The small amount of ClO2 initially
present will react rapidly with the starting material, leaving
ClO2 in excess of ClO2 and, therefore, making the ClO2
more readily available to react with the ClO intermediate (eq
10) when half of the styrene has been consumed.
Figure 8. Experimental (symbols) and simulated (solid lines)
inorganic reactant/product concentrations: [ClO2]0 = 7.39 mM,
−
[ClO2 ] = 72.1 mM, [4-styrenesulfonic acid] = 13.1 mM, [Ac−] = 200
−
−
mM buffered at pH 5.25; 100% D2O; simulated data obtained using
kM5 = 9.41 × 105 M−2 s−1. T = 25 °C.
The overall stoichiometry depends on the relative con-
tributions of eq 8 and eq 13 to the reaction. The rates of these
two reactions under our highest chlorite conditions are roughly
equal, thereby accounting for the higher concentrations of Cl−
compared to ClO3 in the final reaction mixtures for high
chlorite reactions (Figure 3, for example).
In our complete mechanism the epoxidation by chlorite
occurs by two chain reactions. In the original Kolar and
Lindgren proposal, ClO reacts only with ClO2 (M14). For
reactions that included chlorite ion (at relatively low
concentrations) Kolar and Lindgren41 proposed the following
chain reaction (eqs 5−7), where S = styrene and SO = styrene
oxide, with the net reaction forming equal amounts of chloride
and chlorate (eq 8).
−
The high selectivity for epoxidation in a large excess of
−
chlorite is attributed to the reaction of ClO2 with HOCl,
ultimately forming chlorine dioxide and markedly lowering the
concentration of HOCl during the reaction. This diversion to
the ClO2 pool reduces the rate of side reaction M13 and
thereby reduces formation of the resulting byproducts to <10%
(M10) ClO2 + S → SO + ClO
(5)
−
(M14) ClO + ClO2 + H2O → HOCl + ClO3 + H+
of the overall conversion. The branching fraction RM5
(RM5+RM13) via eq 14
/
(6)
6722
dx.doi.org/10.1021/ic500512e | Inorg. Chem. 2014, 53, 6715−6727

Inorganic Chemistry
Article
Figure 9. Mechanism for the epoxidation of styrene by chlorite. Bold arrows represent the highest flux reactions when chlorite is in large excess and
the pH is 5−6. The double-headed arrow indicates 2 equiv of ClO2 are produced in reaction M6. The reaction details for M1−M15 are in Table 1.
optimal conditions examined in the pH range of 5−6, the
reactions involving Cl2 (M1, M8, M11, M12) are only minor
−
contributors. With high [ClO2 ] (∼0.1 M) the calculated
steady-state concentration of ClO2 (∼10−2 M) is greater than
[HOCl] (∼10−6 M) and [Cl2] (∼10−14 M), showing that the
rate of formation of epoxide can be favored over the formation
of chlorinated side products despite the much lower rate constant
for reaction of ClO2 with substrate (∼0.01 M−1 s−1 ) in
comparison to HOCl (∼15 M−1 s−1 at pH 5) and Cl2 (∼108
M−1 s−1). In the reaction mechanism, the strong oxidant
chlorite appears in several reactions, and via the net reactions of
eqs 8 and 13 it is the ultimate oxidant when [ClO2] ≪
[substrate].
The observations at low chlorite (for example, Figures 4 and
5) are readily explained by the shift in balance of the two chain
reactions, with eq 8 now dominating, and the decreased
trapping of intermediate HOCl by chlorite. The excess chlorate
produced in Figure 5 is a result of the relatively high [ClO2]0 at
Figure 10. Simulated branching fraction RM15/(RM15 + RM14) vs initial
the outset of the reaction coupled to the significant reaction of
HOCl with substrate instead of chlorite. Chlorine dioxide
initially present will react with styrene to produce the epoxide
and ClO radical (M10). With low chlorite in the reaction
mixture reaction M14 (ClO2 + ClO) is favored over M15
(ClO2 + ClO). Reaction M14 generates ClO3 and HOCl,
which accounts in part for the consumption of chlorine dioxide
shown in Figure 5. Furthermore, because [HOCl] is not rapidly
decreased by the chlorite-dependent reaction M5 in low
chlorite reactions, the reaction of styrene with HOCl (M13)
becomes significant, resulting in the higher observed yield of 2-
chloro-1-phenyl ethanol.
The uniformly good fit to the kinetics and product
distributions in both low and high chlorite reactions at various
pH values, in both water and mixed solvent, strongly supports
the proposed mechanism in Table 1. Further discussion of
predicted conversions and selectivities based on the simulations
of the mechanism in Table 1 is presented in the Supporting
Information.
Notably, beyond the epoxide there is no other product
apparent from the reaction of ClO2 with the substrate. All of
the undesirable side products arise from reactions of other
chlorine species formed with the substrate, particularly HOCl.
−
[ClO2] for varying initial [ClO2 ] calculated at t = 1/2[styrene]0.
Conditions: [styrene] = 12.5 mM, [Ac−] = 0.20 M.
−
BF = kM5[H+][ClO2−]/(kM5[H+][ClO2
+ kM13[styrene][H+])
]
−
−
= kM5[ClO2−]/(kM5[ClO2−] + kM13[styrene])
(14)
was calculated throughout the reaction using the reaction
conditions stated in Figures 2 and 3 and is ∼0.95, thereby
accounting for the high yield of epoxide (steps M1 and M7 can
be ignored as they have a negligible impact on [HOCl]).
Additional details are in the Supporting Information. Addition
of chlorite to suppress organochlorine products in ECF pulp
bleaching has been advocated by Ni and co-workers,69 and their
rationale is supported by our results.
Under high chlorite conditions, the instantaneous rates of the
reactions M1−M15 in the epoxidation model were calculated
and compared at all time points (see Supporting Information).
Comparing the rates, it is possible to identify the reactions with
the highest flux during the oxidation, and bold arrows were
used to note those steps in Figure 9. Essentially, under the
6723
dx.doi.org/10.1021/ic500512e | Inorg. Chem. 2014, 53, 6715−6727

Inorganic Chemistry
Article
This model serves to illustrate why oxidations using ClO2 alone
often display relatively poor selectivity, and it is also a likely
reason that industrial bleaching with ClO2 is often accompanied
by chlorination side reactions.27−29,69,82
Epoxide Yield As a Function of Chlorite and Initial
Chlorine Dioxide Concentrations. A useful quantitative
mechanism allows calculation of product yield as a function of
the initial reaction conditions. The mechanism of Table 1 was
tested by comparison of experimental and predicted of organic
product distributions as a function of initial chlorite and
chlorine dioxide concentrations. As shown in Figure 11
active chlorinating agent(s) and to probe the mechanistic
details of chlorohydrin and other byproduct formation.
Radical Addition/Elimination Versus Electron-Transfer
Epoxidation Mechanisms. Our proposed mechanism for
M10 involves the epoxidation of styrene via radical addition of
chlorine dioxide to the double bond, forming 1, followed by the
elimination of the ClO radical to form epoxide (Figure 6). Rav-
Acha and co-workers,83 however, have suggested that
epoxidation occurs via an outer-sphere electron-transfer
reaction initially forming the styrene radical cation (2) and
chlorite.
In their view, subsequent addition of water and a second
equivalent of chlorine dioxide to the radical cation yields the
epoxide and chlorous acid.83 Both Kolar et al. and Rav-Acha et
al. observed a decrease in the rate of epoxidation as the polarity
of the solvent decreases. Rav-Acha et al. attribute this solvent
effect to an increasing solvent barrier for the electron-transfer
reaction, while Kolar and Lindgren consider the solvent effect
to be attributable to a charge separation in the styrene-ClO2
adduct (1).41,83
While neither scenario can be ruled out based on
observations related to the polarity of the transition state in
the rate-determining step, the predicted stoichiometry for the
Rav-Acha et al. mechanism (consumption of 2 equiv of chlorine
dioxide and the production of 2 equiv of chlorite/chlorous
acid) is not observed experimentally. In addition, the
mechanism proposed by Rav-Acha includes steps involving
the reaction of starting material with HOCl; however, no step
in their mechanism provides for the production of chlorine(I).
In contrast, the Kolar and Lindgren addition/elimination
mechanism posits the production of the chlorine oxide radical,
Figure 11. Comparison of styrene oxide yields from experiments and
predictions via the mechanism in Table 1 at various initial
−
−
concentrations ClO2 and ClO2 . [ClO2] = 1−15 mM and [ClO2 ]
= 5−25 mM. See Supporting Information for details. The dashed line
represents theoretical perfect agreement between predicted and
experimental selectivities.
(conditions and results tabulated in Supporting Information),
good agreement between observed and predicted yield for
styrene oxide is found for a wide range of initial conditions (R2
= 0.992).
−
which goes on to react with ClO2 or ClO2 , forming HOCl
(M14 and M15).41,83
Our molecular mechanism for epoxidation of olefins by
chlorine dioxide is analogous to the well-studied epoxidation
reaction by isoelectronic alkylperoxy radicals (ROO·), which
proceeds via addition of ROO· to the alkene followed by
elimination of RO· radical.57−59 The heat of reaction for the
epoxidation of olefins with the peroxy radical is approximately
−20 kcal/mol.84,85 Using literature values for heats of formation
of ClO2 and ClO,54 we estimated ΔH° ≅ −23 kcal/mol for
reaction M10. It is reasonable that both exothermic epoxidation
reactions would proceed by a similar radical addition/
elimination mechanism. However, based on the results for
styrene alone we cannot eliminate an alternative mechanism for
M10 in which an initial endoergic outer-sphere electron-
transfer reaction is followed by exoergic chlorite addition and
decomposition of the adduct (Figure 12). In this case, the rate-
determining step would be an electron-transfer reaction as
opposed to adduct formation.
Effect of Substituents on the Rate of Epoxidation. We
further probed the nature of the rate-determining step for M10
by investigating the reaction of ClO2 with para-substituted
styrenes with X = MeO−, Me−, H−, Cl−, and −NO2 at the
para position. In the first two cases (X = MeO− and Me−),
reactions in our standard 60:40 water/acetonitrile solvent were
accompanied by some phase separation. We therefore reduced
initial concentrations of substrate in 60:40 water/acetonitrile so
Styrene Chlorination Reactions. The formations of
chlorinated products of styrene, namely, 1,2-dichloroethylben-
zene, 2-chloro-1-phenyl ethanol, and 2-chlorovinylbenzene, are
described by reactions M11−M13. 2-Chlorovinylbenzene could
in principle be formed as a result of dehydration of 2-chloro-1-
phenyl ethanol; however, from the experimental results the
amount of 2-chlorovinylbenzene is clearly not dependent on
the amount of HOCl available. Therefore, M12 is tentatively
written as a chlorination by Cl2. The formation of 1,2-
dichloroethylbenzene (M11) is the result of chlorination of
styrene by molecular chlorine. Simulations show that Cl2 is not
at equilibrium with HOCl due to the slow equilibration of
reaction M1, and its concentration is ∼10−14 M under all
conditions we studied.
The formation of 2-chloro-1-phenyl ethanol (M13) is written
as the product of an acid-dependent chlorination of styrene by
HOCl since the chlorohydrin concentration increases propor-
tionately with the product [HOCl][H+]. As illustrated recently
by Sivey et al., the active chlorinating agent in reactions of this
nature may be a less-abundant yet more-reactive chlorine
species (e.g., Cl2O or H2OCl+).78,80 Although we have been
successful in modeling the chlorination side reactions under a
number of initial conditions, further studies on the HOCl
reactions themselves are necessary to determine definitively the
6724
dx.doi.org/10.1021/ic500512e | Inorg. Chem. 2014, 53, 6715−6727

Inorganic Chemistry
Article
The energetics of a hypothetical outer-sphere electron-
transfer reaction (Figure 12) can be estimated from the
electrode potentials for the styrene derivatives86,87 and the
ClO2/ClO2 couple.1−3 Although the potentials in water and
−
mixed acetonitrile/water, respectively, are not known, it is
apparent that the hypothetical one-electron transfer reaction in
Figure 11 is highly endoergic. Given self-exchange rate
constants for the two couples, these outer-sphere reaction
rate constants for the substituted styrenes with ClO2 could be
estimated via the Marcus cross relation.88,89 However, the
styrene/styrene radical cation self-exchange rate constants are
not known, and the equilibrium constants in the solvent
mixtures are only estimated, so it is not possible to definitively
exclude the electron-transfer mechanism of Figure 12 on the
basis of a Marcus cross reaction analysis of the individual
predicted rate constants. While an outer-sphere reaction
mechanism could in principle be eliminated for these highly
endoergic reactions in view of the relatively high observed
forward rate constants, the uncertainty in the reaction free
energies requires that additional evidence be considered.
We can predict the relative rates of cross reactions with some
confidence by making the simplifying assumption that the
substituents do not substantially alter the self-exchange rates for
the styrenes and their radical cations. (Given the narrow size
range of the reactants, this assumption appears reasonable; see,
for example, the Marcus analysis of organic hydrazine radical
cation/neutral electron-transfer reactions by Nelsen and
Pladziewicz90 and radical anion/neutral self-exchange rates for
similarly substituted benzonitrile by Kowert et al.91) The
change in free energy for the electron-transfer reaction going
from −Cl to −OMe substituents (∼0.6 V)86 would predict a
large increase in the outer-sphere electron-transfer rate constant
via the cross relation (kOMe/kCl ≈105), but the observed ratio is
only ∼50 (also see Table S2 in Supporting Information). From
−Cl to −Me, the potential difference is ∼0.4 V, and the cross
relation prediction is kMe/kCl ≈ 1500, but a factor of only ∼6 is
observed. The kinetic impact of varying para substituents is
more in line with the energetic effects of the same substituents
on the relative stabilities of benzyl radicals.92 In the absence of
experimental data for the para-substituted styrene(0/+) self-
exchange reaction rates, caution is warranted for this analysis;
however, the substituent effects on the epoxidation reaction
rate constants are more consistent with an addition/elimination
mechanism (Figure 6) than they are with a rate-determining
outer-sphere electron-transfer reaction followed by addition of
chlorite (Figure 12).
Figure 12. Alternative elementary mechanism for epoxidation of
styrene via outer-sphere electron transfer followed by addition of
chlorite.
that we could make direct comparisons to the styrene results
from our detailed studies. Table 3 gives the best-fit rate
constants for M10 for the substituted styrenes using the full
mechanism of Table 1.
Table 3. Rate Constants for Reaction M10 for p-Substituted
Styrenes
a
Rate constants and errors based on an optimized fit for kM10 using the
b
mechanism in Table 1. Rate constants and estimated errors using R =
k[ClO2][sub]
We found that a 20:55:25 acetonitrile/water/ethanol mixture
prevented phase separation even with higher substrate
concentrations, so this solvent was also used to investigate
substituent effects using a more polar solvent. Ethanol
complicates the kinetic interpretation since we noted in control
reactions that the alcohol reacts with chlorine dioxide slowly
over the course of styrene oxidations. It was generally observed
that the concentration of ClO2 increased during the reactions
(as a result of the reaction of HOCl with chlorite), as expected
from the mechanism, but the stoichiometries were not reliable
due to the slow reaction of solvent with ClO2. We were able to
determine the direct rates of reaction of chlorine dioxide with
substituted styrenes by following the loss of styrene and
formation of the epoxides along with the concentration of ClO2
throughout the initial part of the reaction. In the more polar
solvent, methoxystyrene was completely consumed in <300 s,
and it was only possible to determine the minimum rate
constant. For p-nitrostyrene, the ClO2 UV−visible peak was
obscured by the strong absorbance in the same region by the
substrate. The other three observed reaction profiles were fit to
the simple rate law R = k[ClO2][sub]. As a further check, we
did a numerical fit to the reaction data using a simple kinetic
model in which the observed small [ClO2] increase during the
run was simulated, and a best-fit value for k was determined.
Both methods gave the same rate constants within ranges
quoted in Table 3.
CONCLUSION
■
This investigation shows how efficient chlorite oxidations of
alkenes can be achieved through control of pH and initial
chlorite and chlorine dioxide concentrations. The stoichiometry
and inorganic products are determined by the competition of
two chain reactions for the consumption of chlorite and
substrate, and both chains involve the epoxidation of the vinyl
group by ClO2. As proposed by Jangam and Richardson40 and
confirmed by this work, the addition of chlorine dioxide to the
initial reaction mixture initiates the chain reactions and
predictably contributes to the overall stoichiometry. If ClO2
is used as the only initial chlorine reagent in reactions with
alkenes, then the subsequent reactions produce HOCl in
abundance and result in copius amounts of chlorinated side
products.
6725
dx.doi.org/10.1021/ic500512e | Inorg. Chem. 2014, 53, 6715−6727

Inorganic Chemistry
Article
It is well-known that chlorinated organics are observed in
ClO2-only (ECF) bleaching plants, and Ni and co-workers28,29
suggested that addition of chlorite to the process reduces the
concentration of HOCl and therefore reduces the undesirable
chlorinated byproducts. Their explanation for the effect as well
as the pioneering work of Kolar and co-workers27 are validated
by the present studies using vinylbenzene substrates in place of
the complex mixture of substrates in lignin. The successful
quantitative modeling by the detailed mechanism of Figure 11
puts the earlier proposals on firm mechanistic grounds, at least
for reactions involving oxidation of double bonds. The model
quantitatively explains the formation of chlorinated hydro-
carbons in chlorine dioxide oxidations, produced, for example,
in a variety applications including water treatment and pulp
bleaching, by accounting for the formation of HOCl rather than
attributing that formation to reactions of substrate with
chlorine impurities.83,93−96
This work also supports a less commonly invoked pathway
for chlorine dioxide oxidations in which radical addition/
elimination occurs in lieu of an initial one-electron outer-sphere
electron-transfer reaction. The probability of the former
pathway, which is closely analogous to certain reactions of
peroxyl radicals, is significantly enhanced when the inter-
mediate radical adduct has special stability, as in the case of
styrene and its derivatives studied here.
graduates grant. The authors gratefully acknowledge the
University of Florida for supporting this research.
REFERENCES
(1) Holst, G. Sven. Papperstidn. 1945, 48.
(2) Flis, I. E. Zh. Fiz. Khim. 1958, 32.
■
(3) Troitskaya, N. V.; Mishchenko, K. P.; Flis, I. E. Russ. J. Phys.
Chem. 1959, 33.
(4) Dodgen, H.; Taube, H. J. Am. Chem. Soc. 1949, 71, 2501−2504.
́
(5) Hoigne, J.; Bader, H. Water Res. 1994, 28, 45−55.
(6) Alternative Disinfectants and Oxidants Guidance Manual, EPA
#815-R-99−014; U.S. Environmental Protection Agency: Washington,
DC, 1999.
(7) Latshaw, C. L. Tappi 1994, 163−166.
(8) Gordon, G.; Rosenblatt, A. A. Ozone: Sci. Eng. 2005, 203−107.
(9) Svenson, D. R.; Jameel, H.; Chang, H.; Kadla, J. F. J. Wood Chem.
Technol. 2006, 201−213.
(10) Sharp, R. J.; Roberts, A. G. J. Chem. Technol. Biotechnol. 2006,
1612−1625.
(11) Callahan, K.; Beck, N.; Duffield, E.; Shin, G.; Meschke, J. J.
Occup. Environ. Hyg. 2010, 7, 529−534.
(12) Davies, A.; Pottage, T.; Bennett, A.; Walker, J. J. Hosp. Infect.
2011, 77, 199−203.
(13) Lim, M. Y.; Kim, J.-M.; Ko, G. Water Res. 2010, 44, 3243−3251.
(14) Srinivasan, A.; Bova, G.; Ross, T.; Mackie, K.; Paquette, N.;
Merz, W.; Perl, T. M. Infect. Control Hosp. Epidemiol. 2003, 24, 575−
579.
We are presently investigating the use of mechanisms based
on the overall scheme in Figure 9 to provide quantitative
models for chlorite and/or chlorine dioxide oxidations of a
variety of other substrates, such as sulfides, phenols, and related
biological molecules. Other substrates will of course require the
introduction of additional steps for ClO2 reactions and
reactions of other chlorine intermediates with substrate to
account for product distributions and kinetics. It is encouraging
that the highly interdependent chemistry of multiple chlorine
oxidation states in complex mixtures can be modeled
quantitatively by a series of known reactions.
(15) Walsh, T. R.; Toleman, M. A. J. Antimicrob. Chemother. 2012,
67, 1−3.
(16) Pogue, J. M.; Mann, T.; Barber, K. E.; Kaye, K. S. Expert Rev.
Anti-Infect. Ther. 2013, 11, 383−393.
(17) Gordon, G.; Kieffer, R. G.; Rosenblatt, D. H. The Chemistry of
Chlorine Dioxide. In Progress in Inorganic Chemistry; Lippard, S. J., Ed.;
John Wiley & Sons, Inc.: Hoboken, NJ, 1972; Vol. 15, pp 202−286.
(18) Rav-Acha, C. Water Res. 1984, 18, 1329−1341.
(19) Lehtimaa, T.; Kuitunen, S.; Tarvo, V.; Vuorinen, T.
Holzforschung 2010, 64, 555−561.
(20) Tarvo, V.; Lehtimaa, T.; Kuitunen, S.; Alopaeus, V.; Vuorinen,
T.; Aittamaa, J. Ind. Eng. Chem. Res. 2009, 48, 6280−6286.
(21) Wang, L.; Margerum, D. W. Inorg. Chem. 2002, 41, 6099−6105.
(22) Jia, Z.; Margerum, D. W.; Francisco, J. S. Inorg. Chem. 2000, 39,
2614−2620.
ASSOCIATED CONTENT
* Supporting Information
■
S
(23) Peintler, G.; Nagypal
2954−2958.
(24) Yin, G.; Ni, Y. Can. J. Chem. Eng. 1998, 76, 921−926.
(25) Kormanyos, B.; Nagypal, l.; Peintler, G.; Horvath, A. K. Inorg.
Chem. 2008, 47, 7914−7920.
́
, I.; Epstein, I. J. Phys. Chem. 1990, 94,
Included here: introductory material, error analysis, HOCl/
ClO2− reaction, oxidation of styrene and 4-styrenesulfonic acid,
reaction selectivity and epoxidation mechanism, rate constant
correlation matrices. This material is available free of charge via
the Internet at http://pubs.acs.org.
́
́
́
(26) Volk, C. J.; Hofmann, R.; Chauret, C.; Gagnon, G. A.; Ranger,
G.; Andrews, R. C. J. Environ. Sci. Eng. 2002, 323−330.
(27) Kolar, J. J.; Lindgren, B. O.; Pettersson, B. Wood Sci. Technol.
1983, 17, 117−128.
(28) Ni, Y.; Kubes, G. J.; Heiningen, A. R. P. v. Nord. Pulp Pap. Res. J.
1993, 8, 350−352.
(29) Ni, Y.; Kubes, G. J.; Heiningen, A. R. P. v. Nord. Pulp Pap. Res. J.
1992, 7, 200−204.
AUTHOR INFORMATION
Corresponding Author
*E-mail: der@ufl.edu.
■
Notes
The authors declare no competing financial interest.
(30) Pettersen Roger, C. Chem. Solid Wood 1984, 207, 57−126.
(31) Krapcho, A. P. Organic Preparations and Procedures International
2006, 38, 177−216.
ACKNOWLEDGMENTS
■
(32) Lindgren, B. O.; Nilsson, T.; Husebye, S.; Mikalsen, Ø.;
Leander, K.; Swahn, C.-G. Acta Chem. Scand. 1973, 27, 888.
(33) Dalcanale, E.; Montanari, F. J. Org. Chem. 1986, 51, 567−569.
(34) Babu, B. R.; Balasubramaniam, K. K. Org. Prep. Proced. Int. 1994,
26, 123−125.
This Paper was inspired by H. Taube, who first investigated
chlorine dioxide as a student at the Univ. of Saskatchewan
(Spinks, J. W. T.; Taube, H. J. Am. Chem. Soc. 1937, 59, 1155−
115) and mentioned its potential practical usefulness to one of
us (D.E.R.) 31 years ago. Since that time, Taube’s prediction
has become a reality, and, among its many applications,
chlorine dioxide has significantly reduced chlorinated organic
pollution in modern wood pulp bleaching and has been used to
destroy anthrax. J.R. was supported by a National Science
Foundation International Research Experiences for Under-
(35) Launer, H. F.; Tomimatsu, Y. Anal. Chem. 1959, 31, 1385−
1390.
(36) Hefti, H. Text. Res. J. 1960, 30, 861−867.
(37) Higginbotham, R. S.; Leigh, R. A. J. Text. Inst., Proc. 1962, 53,
312−319.
(38) Abdel-Halim, E. S. Carbohydr. Polym. 2012, 90, 316−321.
6726
dx.doi.org/10.1021/ic500512e | Inorg. Chem. 2014, 53, 6715−6727

Inorganic Chemistry
Article
(39) Geng, X.; Wang, Z.; Li, X.; Zhang, C. J. Org. Chem. 2005, 70,
(79) Pinkston, K. E.; Sedlak, D. L. Environ. Sci. Technol. 2004, 38,
4019−4025.
9610−9613.
(80) Sivey, J. D.; McCullough, C. E.; Roberts, A. L. Environ. Sci.
Technol. 2010, 44, 3357−3362.
(40) Jangam, A.; Richardson, D. E. Tetrahedron Lett. 2010, 51, 6481−
6484.
(81) Kręzel, A.; Bal, W. J. Inorg. Biochem. 2004, 98, 161−166.
̇
(41) Kolar, J. J.; Lindgren, B. O. Acta Chem. Scand., Ser. B 1982, 36,
(82) Solomon, K. R. Pure Appl. Chem. 1996, 68, 1721−1730.
(83) Rav-Acha, C.; Choshen, E.; Blits, R.; Grafstein, O. A Proposed
Mechanism for the Reactions of Chlorine Dioxide with Aquatic
Organic Materials. In Water Chlorination: Health Effects and Environ-
mental Impacts; Lewis Publishers: Ann Arbor, MI, 1987; Vol. 6, pp
849−857.
599−605.
(42) Nicoson, J. S.; Margerum, D. W. Inorg. Chem. 2002, 41, 342−
347.
́
(43) Horvath, A. K.; Nagypal, I.; Peintler, G.; Epstein, I. R.; Kustin,
K. J. Phys. Chem. A 2003, 107, 6966−6973.
(44) Kieffer, R. G.; Gordon, G. Inorg. Chem. 1968, 7, 235−239.
(45) Kieffer, R. G.; Gordon, G. Inorg. Chem. 1968, 7, 239−244.
(46) Svenson, D.; Kadla, J. F.; Chang, H.; Jameel, H. Can. J. Chem.
2002, 80, 761−766.
(47) Lindgren, B. O.; Svahn, C. M.; Widmark, G. Acta Chem. Scand.
1965, 19, 7−13.
(48) Rosenblatt, D. H.; Hull, L. A.; Luca, D. C. D.; Davis, G. T.;
Weglein, R. C.; Williams, H. K. R. J. Am. Chem. Soc. 1967, 89, 1158−
1163.
(49) Hull, L. A.; Davis, G. T.; Rosenblatt, D. H.; Williams, H. K. R.;
Weglein, R. C. J. Am. Chem. Soc. 1967, 89, 1163−1170.
(50) Hull, L. A.; Giordano, W. P.; Rosenblatt, D. H.; Davis, G. T.;
Mann, C. K.; Milliken, S. B. J. Phys. Chem. 1969, 78, 2147.
(51) Davis, G. T.; Demek, M. M.; Rosenblatt, D. H. J. Am. Chem. Soc.
1972, 94, 3321−3325.
(52) Burrows, E. P.; Rosenblatt, D. H. J. Org. Chem. 1982, 47, 892−
893.
(53) Chen, A. S. C.; Larson, R. A.; Snoeyink, V. L. Environ. Sci.
(84) Simmie, J. M.; Black, G.; Curran, H. J.; Hinde, J. P. J. Phys.
Chem. A 2008, 112, 5010−5016.
(85) Lemmon, E.; McLinden, M.; Friend, D.; Linstrom, P.; Mallard,
W. NIST Chemistry WebBook. In NIST Standard Reference Database
Number 69, National Institute of Standards and Technology:
Washington, DC, 2011.
(86) Kojima, M.; Sakuragi, H.; Tokumaru, K. Bull. Chem. Soc. Jpn.
1985, 58, 521−524.
(87) Garrison, J. M.; Ostovic, D.; Bruice, T. C. J. Am. Chem. Soc.
1989, 111, 4960−4966.
(88) Marcus, R. Discuss. Faraday Soc. 1960, 29, 21−31.
(89) Marcus, R.; Sutin, N. Biochim. Biophys. Acta 1985, 811, 265−
322.
(90) Nelsen, S. F.; Pladziewicz, J. R. Acc. Chem. Res. 2002, 35, 247−
254.
(91) Kowert, B. A.; Marcoux, L.; Bard, A. J. J. Am. Chem. Soc. 1972,
94, 5538−5550.
(92) Wu, Y.-D.; Wong, C.-L.; Chan, K. W.; Ji, G.-Z.; Jiang, X.-K. J.
Org. Chem. 1996, 61, 746−750.
Technol. 1982, 16, 268−273.
(54) Meyer, M. M.; Kass, S. R. J. Phys. Chem. A 2010, 114, 4086−
4092.
(93) Fiessinger, F.; Richard, Y.; Montiel, A.; Musquere, P. Sci. Total
Environ. 1981, 18, 245−261.
(55) Alfassi, Z. B.; Huie, R. E.; Neta, P. J. Phys. Chem. 1986, 90,
(94) Rice, R. G.; Gomez-Taylor, M. Environ. Health Perspect. 1986,
69, 31−44.
4156−4158.
(56) Ganiev, I. M.; Ganieva, E. S.; Kabal’nova, N. N. Russ. Chem. Bull.
2004, 53, 2281−2284.
(95) Li, J. W.; Yu, Z.; Cai, X.; Gao, M.; Chao, F. Water Res. 1996, 30,
2371−2376.
(57) Koelewijn, P. Recl. Trav. Chim. Pays-Bas 1972, 91, 759−779.
(58) Twigg, G. H. Chem. Eng. Sci. 1954, 3, 5−16.
(59) Brill, W. F. J. Am. Chem. Soc. 1963, 85, 141−145.
(60) Furman, C. S.; Margerum, D. W. Inorg. Chem. 1998, 37, 4321−
4327.
(96) Chang, C.; Hsieh, Y.; Lin, Y.; Hu, P.; Liu, C.; Wang, K.
Chemosphere 2001, 44.
(61) Masschelein, W. J. J.Am. Water Works Assoc. 1984, 76, 70−76.
(62) Albrich, J. M.; Hurst, J. K. FEBS Lett. 1982, 144, 157−161.
(63) Cady, G. Inorganic Syntheses; McGraw-Hill Company: New
York, 1957.
(64) Lehtimaa, T.; Kuitunen, S.; Tarvo, V.; Vuorinen, T. Ind. Eng.
Chem. Res. 2010, 49, 2688−2693.
(65) Gear, C. W. Commun. ACM 1971, 14, 176−179.
(66) Nelder, J. A.; Mead, R. Comput. J. 1965, 7, 308−313.
(67) Marquardt, D. W. J. Soc. Ind. Appl. Math. 1963, 11, 431−441.
(68) Bevington, P. R.; Robinson, D. K. Data Reduction and Error
Analysis for the Physical Science, 3rd ed.; McGraw-Hill: 2003.
(69) Ni, Y.; Kubes, G.; van Heiningen, A. J. Pulp Pap. Sci. 1994, 20,
J103−J106.
(70) Taube, H.; Dodgen, H. J. Am. Chem. Soc. 1949, 71, 3330−3336.
(71) Emmenegger, F.; Gordon, G. Inorg. Chem. 1967, 6, 633−635.
(72) Aieta, E. M.; Roberts, P. X. Environ. Sci. Technol. 1986, 20, 50−
55.
(73) Tang, T. F.; Gordon, G. Environ. Sci. Technol. 1984, 18, 212−
216.
(74) Lide, D. R. CRC Handbook of Chemistry and Physics, 71st ed.;
CRC Press: Boca Raton, FL, 1990−1991; pp 8−35.
(75) Alfassi, Z. B.; Huie, R. E.; Mosseri, S.; Neta, P. Radiat. Phys.
Chem. 1988, 32, 85−88.
(76) Groves, J. T.; Myers, R. S. J. Am. Chem. Soc. 1983, 105, 5791−
5796.
(77) Collman, J. P.; Kodadek, T.; Brauman, J. I. J. Am. Chem. Soc.
1986, 108, 2588−2594.
(78) Sivey, J. D.; Roberts, A. L. Environ. Sci. Technol. 2012, 46.
6727
dx.doi.org/10.1021/ic500512e | Inorg. Chem. 2014, 53, 6715−6727