Mechanism Involving Hydrogen Sulfite Ions, Chlorite Ions, and Hypochlorous Acid as Key Intermediates of the Autocatalytic Chlorine Dioxide-Thiourea Dioxide Reaction

Article abstract of DOI:10.1002/ejic.201500654

The kinetics of the chlorine dioxide-thiourea dioxide reaction was investigated by monitoring absorbance-time profiles at λ = 360 nm. Under acidic conditions, the primary carbon-containing product is cyanamide, not urea as considered previously for many oxidation reactions of thiourea dioxide. Increase of the rate of the reaction by an increase of pH can be readily explained by the slow pH-dependent formation of a more reactive form of thiourea dioxide (TDO) that is produced steadily and unavoidably as the stock TDO solution ages. We have also found that the absorbance-time profiles of the chlorine dioxide-TDO reaction are sigmoidal with excess TDO. The addition of methionine as a hypochlorous acid scavenging agent inhibits the reaction significantly, whereas the addition of chlorite ions and trace amounts of hydrogen sulfite ions accelerates the decay of chlorine dioxide. On the basis of these experiments, a sixteen-step kinetic model involving hypochlorous acid, chlorite ions, and hydrogen sulfite ions as key intermediates that provide an autocatalytic cycle is proposed to account for the overall kinetic behavior observed, including the slow rearrangement of TDO.

Full text of DOI:10.1002/ejic.201500654

FULL PAPER
DOI:10.1002/ejic.201500654
Mechanism Involving Hydrogen Sulfite Ions, Chlorite Ions,
and Hypochlorous Acid as Key Intermediates of the
Autocatalytic Chlorine Dioxide–Thiourea Dioxide
Reaction
[a]
[b]
[a]
[b]
˝
Ying Hu, Attila K. Horváth,* Sasa Duan, György Cseko,
Sergei V. Makarov,[c] and Qingyu Gao*[a]
Keywords: Autocatalysis / Kinetics / Reaction mechanisms / Reactive intermediates / Oxides
The kinetics of the chlorine dioxide–thiourea dioxide reac-
excess TDO. The addition of methionine as a hypochlorous
tion was investigated by monitoring absorbance–time pro- acid scavenging agent inhibits the reaction significantly,
files at λ = 360 nm. Under acidic conditions, the primary
carbon-containing product is cyanamide, not urea as consid-
ered previously for many oxidation reactions of thiourea di-
oxide. Increase of the rate of the reaction by an increase of
pH can be readily explained by the slow pH-dependent for-
mation of a more reactive form of thiourea dioxide (TDO) that
is produced steadily and unavoidably as the stock TDO solu-
tion ages. We have also found that the absorbance–time pro-
files of the chlorine dioxide–TDO reaction are sigmoidal with
whereas the addition of chlorite ions and trace amounts of
hydrogen sulfite ions accelerates the decay of chlorine diox-
ide. On the basis of these experiments, a sixteen-step kinetic
model involving hypochlorous acid, chlorite ions, and
hydrogen sulfite ions as key intermediates that provide an
autocatalytic cycle is proposed to account for the overall ki-
netic behavior observed, including the slow rearrangement
of TDO.
ent reaction. To pave the way towards detailed mechanistic
insights into the chlorite–TU reaction, knowledge of the
kinetics and mechanism of the title reaction is eagerly anti-
Introduction
The reaction of chlorite ions with thiourea (TU) has at-
tracted the extensive attention of scientists as a new chlorite
oscillator in continuous-flow stirred-tank reactors (CSTRs)
since 1985.[1] Subsequently, several studies indicated that
this reaction displays more common nonlinear phenomena
such as chaos[2] and chemical waves.[3] Although many as-
pects of this fascinating reaction are qualitatively well-
known, no detailed kinetic investigation of this system has
been reported owing to the complexity of the overall reac-
tion, which involves multiple-valence sulfur- and chlorine-
containing intermediates in its subsystems.
2–
cipated. Owing to the formation of SO2H–/SO2 ions from
the rapid decomposition of TDO in alkaline solutions,[5,6]
TDO has been extensively applied in chemistry and chemi-
cal technology as a special and effective reducing agent,[7–10]
especially for the potential chemical reduction of graphene
oxide in recent years.[11,12] Its reducing ability increases with
aging in alkaline solution. The stability of sulfoxylate ions
increases with increasing pH under anaerobic conditions.[13]
Some oxidation reactions of TDO have been controlled in
acidic solutions.[14–16] The oxidations of TDO with chlorite
ions, iodine, and bromine have been performed by Simoyi
et al.[15–18] Regrettably, the tautomerization of TDO, which
seems to occur relatively slowly from (NH2)2CSO2 to
NH2NHCSO2H[19,20] at low pH values, has so far not been
taken into account in these reactions, though the two
tautomers may have quite different reactivity towards chlo-
rine dioxide and other chlorine-containing intermediates.
Chlorine dioxide has long been known as an antimicro-
bial agent, and recent interest has focused on its application
in gaseous form because it seems to be more effective than
it is in aqueous solutions.[21–24] As a radical with an odd
number of valence electrons, ClO2 could react readily with
various sulfur-containing compounds and even display
nonlinear autocatalytic behavior, as in its reactions with
ClO2 as a long-lived chlorine-containing intermediate[3]
and thiourea dioxide (TDO) as a long-lived sulfur-contain-
ing intermediate[4] have already been confirmed in the oxid-
ation of thiourea. Therefore, it is undisputed that the ClO2–
TDO reaction is a subsystem of the chlorite–thiourea par-
[a] School of Chemical Engineering, China University of Mining
and Technology,
Xuzhou 2221111, Jiangsu Province, P. R. China
E-mail: gaoqy@cumt.edu.cn
www.cumt.edu.cn
[b] Department of Inorganic Chemistry, University of Pécs,
Ifjúság u. 6, 7624 Pécs, Hungary
E-mail: horvatha@gamma.ttk.pte.hu
www.pte.hu
[c] State University of Chemistry and Technology,
Sheremetevsky str. 7, 153000 Ivanovo, Russia
Eur. J. Inorg. Chem. 2015, 5011–5020
5011
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
thiourea,[25] thiocyanate ions,[26] and thiosulfate ions.[27,28]
During the autocatalytic process, HOCl and the corre-
sponding reactive sulfur intermediates, thiourea monoxide,
HOSCN, and sulfite ions, serve important governing roles,
as these reactive sulfur intermediates consume chlorine di-
oxide rapidly.
In our previous study,[19] we observed for the first time
that autocatalysis also occurs in the reaction between ClO2
and TDO, though at that time we focused our interest on
an explanation for the aging effect of TDO in acidic me-
dium on the basis of the remarkable reactivity difference
between the tautomeric forms of TDO without a direct in-
sight into the mechanistic details of the given reaction. We
also found that the addition of hydrogen carbonate ions
decreases the length of the induction period of autocataly-
sis; therefore, the otherwise inert hydrogen carbonate ions
may also have an important effect on the autocatalytic
cycle.[19] On the basis of a general interest on the applica-
tion of TDO, we report here a detailed, systematic study
on the kinetics of the ClO2–TDO reaction to obtain better
insights into the mechanism of the fascinating chlorite–
thiourea reaction.
Figure 1. Typical (a) electropherogram and (b) chromatogram dur-
ing the course of the TDO–ClO2 reaction. The electropherogram
and the chromatogram were taken at t = 30 and 13 min, respec-
tively. Initial conditions: pH 2.49, [TDO]0 = 1.03 mm, [ClO2]0
3.12 mm, the detection wavelength was 214 nm.
=
The sulfur balance is perfect within the experimental er-
ror; therefore, sulfite ions should only appear as a short-
lived intermediate under these conditions. These results also
imply that the majority of the initial amount of sulfur forms
sulfate ions; therefore, in addition to TTO, other nitrogen-
and carbon-containing species must form. As many oxid-
ation reactions of TDO with different oxidizing agents were
Results
Identification of Products and Intermediates
Chlorine dioxide is a versatile oxidizing agent because supposed to form urea under acidic conditions[3,16,17] and
its reactions are often accompanied by the formation of a as carbonate and ammonium ions were considered in our
chlorate side-product.[29–31] To check this possibility, we recent article,[19] we performed 13C NMR spectroscopy
performed capillary electrophoresis (CE) measurements to measurements to identify the major carbon- and nitrogen-
detect possible chloride- and sulfur-containing oxyanions.
containing products. The 13C NMR spectrum of the end
As shown in Figure 1a, in addition to chloride ions, chlo- products after the chlorine dioxide completely disappeared
rate ions also form in a significant amount during the from the solution is shown in Figure 2.
course of reaction. Other chlorine-containing products can-
Surprisingly, the 13C NMR spectrum clearly shows that
not be detected under our experimental conditions. It is the carbon-containing end product is not urea or hydrogen
also clarified by Figure 1a that a significant amount of sulf- carbonate but is cyanamide. As we did not find the peak
ate ions are found by the end of the reaction. As thiourea of cyanamide in the 13C NMR spectrum in our previous
trioxide (TTO) is also a conceivable product, HPLC mea- report,[19] this apparent contradiction deserves a conceiv-
surements were performed for the possible identification of able explanation. The main difference during the measure-
this compound. As indicated in Figure 1b, we did indeed ment of the NMR spectrum was the acquisition time. In
find detectable amounts of TTO. The results of the quanti- our previous report, the acquisition time was more than
tative determination of the sulfur-containing products are 18 h, whereas it was only ca. 45 min in the present report.
shown in Table 1.
From these results, we concluded that cyanamide is indeed
Table 1. HPLC determination of TDO, TTO, and sulfate ions with different initial conditions. All concentrations are given in mm. The
pH was adjusted to 2.35 with phosphoric acid/dihydrogenphosphate buffer. The final concentrations were determined after the completion
of the reaction but not earlier than 1 h.
2–
sim
sim
2– sim
Entry
[ClO2]0
[TDO]0
[TDO]ϱ
[TTO]ϱ
[SO4
]
[TDO]ϱ
[TTO]ϱ
[SO4
]
ϱ
ϱ
1
2
3
4
3.00
1.10
2.01
3.12
2.03
3.39
2.03
1.03
0.00
2.01
0.15
0.00
0.14
0.23
0.17
0.10
1.79
1.26
1.63
0.97
0.00
2.31
0.23
0.00
0.02
0.10
0.09
0.00
2.01
0.98
1.71
1.03
Eur. J. Inorg. Chem. 2015, 5011–5020
5012
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
Figure 3. The effect of methionine on the ClO2–TDO reaction. The
initial conditions were as follows: [ClO2]0 = 0.416 mm, [TDO]0
10 mm, [methionine]0 = 30 mm, I = 0.1 m, T = 18 °C, pH 2.35.
=
TDO–chlorite reaction should be very fast and should pro-
duce an intermediate that removes chlorine dioxide ef-
ficiently. Indeed, the reaction between TDO and chlorite
ions is rapid and can be measured conveniently with a
stopped-flow instrument.[18] These experiments indirectly
imply that chlorite ions must play a significant role in the
kinetics of the title reaction. A conceivable explanation of
this experimental fact is that the TDO–chlorite reaction
produces hydrogen sulfite ions as an essential intermediate,
which then remove chlorine dioxide in a rapid reaction.[35]
Figure 2. (A) 13C NMR spectroscopy study of the product of the
ClO2–TDO reaction. The chemical shift of TDO is 177.6 ppm.
[ClO2]0 = 0.06 m, [TDO]0 = 0.25 m, the pH was ca. 2.0. The 13C
NMR spectroscopic data collection was started ca. 90 min later
when the ClO2 was completely reacted. The number of scans was
1024. (B) 13C NMR spectrum of 3.0 m cyanamide. The chemical
shift is 118.0 ppm. (C) 13C NMR spectrum of a 0.15 m urea solu-
tion. The chemical shift is 162.6 ppm.
the primary product of the reaction under our experimental
conditions, but cyanamide appears to be transformed into
ammonium ions and carbon dioxide on a longer timescale
in the presence of excess TDO. As thiourea dioxide may
also serve as an oxidant,[32] this possibility may explain the
difference between our previous and recent reports, but this
phenomenon certainly requires further detailed investi-
gations. Furthermore, the CE and HPLC experiments pre-
sented above also confirm that both chlorite ions and hypo-
chlorous acid can only be short-lived intermediates of the
reaction.
Figure 4. The effect of chlorite ions on the ClO2–TDO reaction.
The initial conditions were as follows: [ClO2]0 = 0.49 mm, [TDO]0
–
= 2.0 mm, [ClO2 ]0 = 0.2 (black), 0.4 (blue), and 0.7 mm (green), I
= 0.1 m, T = 25 °C, pH 2.35.
Effect of Methionine, Chlorite Ions, Sulfite Ions, and pH
If our argument about the influence of chlorite ions on
To demonstrate indirectly that hypochlorous acid is a key the TDO–chlorine dioxide reaction is correct, the addition
intermediate of the title reaction, the reaction was studied of hydrogen sulfite ions should also have a notable effect
at a certain composition in the absence and presence of on the kinetic runs. As is shown in Figure 5, trace amounts
methionine. As Figure 3 demonstrates, methionine de- of added hydrogen sulfite ions accelerate the initial part of
creases significantly the rate of consumption of chlorine di- the reaction; therefore, the fast hydrogen sulfite–chlorine di-
oxide. This straightforwardly suggests that hypochlorous oxide reaction produces either chlorite ions or hypo-
acid is indeed a key intermediate because methionine is a chlorous acid, which have a huge impact on the kinetics.
very efficient HOCl scavenger.[28,33,34]
The pH also has a profound effect on the shape of the
Moreover, the addition of a small amount of chlorite kinetic curves. As Figure 6 depicts, the rate of decay of
ions into the reaction solution has a huge impact on the chlorine dioxide increases significantly with increasing pH.
absorbance measured at the first time point. The chlorine
It is also interesting that the absorbance–time profiles of
dioxide concentration drops immediately within the mixing the TDO–ClO2 reaction are sigmoidal. In our previous pa-
period of the reactants (see Figure 4). As no reaction at all per,[19] we provided experimental evidence that the addition
occurs between chlorite ions and chlorine dioxide, the of hydrogen carbonate ions significantly shortens the induc-
Eur. J. Inorg. Chem. 2015, 5011–5020
5013
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
hypochlorous acid–chlorite reaction.[36] Then, we consid-
ered all of the conceivable reactions among these species
along with their H+- and OH–-catalyzed pathways as an
initial model and systematically removed all of the steps
from the model that had no influence on the quality of the
fit. This method has been successfully applied to several
different chemical systems. As a result of a long but
straightforward elimination process, we finally arrived at
the following compact kinetic model that is able to describe
all of the measured characteristics of the kinetic curves.
–
H3PO4 h H2PO4
TDOǞAIMSA
(E1)
(R1)
(R2)
(R3)
(R4)
(R5)
(R6)
(R7)
Figure 5. The effect of trace amounts of SIV on the ClO2–TDO
reaction. The initial conditions were as follows: [ClO2]0
=
TDO + ClO2 ǞTTO + ClO
AIMSA + ClO2 ǞTTO + ClO
ClO2 + ClOǞCl2O3
Cl2O3 + H2OǞHOCl + ClO3 + H+
AIMSA + Cl2O3 ǞTTO + Cl2O2
Cl2O2 + H2OǞClO3 + Cl– + 2H+
0.416 mm, [TDO]0 = 10.0 mm, [SO32–]0 = 0 (black), 10 (green), and
30 μm (red), I = 0.1 m, T = 18 °C, pH 2.35.
–
–
TTO + Cl2O2 + H2OǞ2HOCl + SO4 + NH2CN + 2H+ (R8)
2–
–
TDO + HOClǞNH2CN + HSO3 + Cl– + 2H+
(R9)
(R10)
(R11)
(R12)
(R13)
(R14)
(R15)
(R16)
TTO + HOClǞNH2CN + SO4 + Cl– + 3H+
2–
–
–
–
HSO3 + ClO2 ǞSO3 + ClO2 + H+
–
2–
–
SO3 + ClO2 + H2OǞSO4 + ClO2 + 2H+
HSO3 + ClO2 ǞSO4 + ClO + H+
–
2–
TDO + ClO2– ǞNH2CN + HSO3 + HOCl
–
HSO3 + ClO2– ǞSO4 + HOCl
–
2–
HSO3 + HOClǞSO4 + Cl– + 2H+
–
2–
The rate coefficients determined by nonlinear simulta-
neous parameter estimation are illustrated in Table 2. The
average deviation was found to be 4.0% by a relative fitting
procedure. Altogether only nine fitted and eight fixed pa-
rameters were used. The quality of the fits for representative
examples are demonstrated in Figures 6, 7, and 8, which
also support that the proposed kinetic model works prop-
erly under our experimental conditions.
Figure 6. Experimental (symbols) and fitted (solid lines) ab-
sorbance–time series in the ClO2–TDO reaction. The initial condi-
tions were as follows: [ClO2]0 = 0.50 mm, [TDO]0 = 10.0 mm, I =
0.1 m, T = 25 °C, pH = 1.87 (black), 2.05 (blue), 2.36 (green), 2.66
(cyan), and 3.07 (red).
tion period through the enhancement of the tautomeric re-
arrangement of TDO, but the experiments performed here
rather support that the interactions of the key intermediates
of the reaction eventually result in the autocatalytic behav-
ior of the system.
Table 2. Fitted and fixed rate coefficients of the proposed model.
If no error is indicated, the given rate coefficient was fixed during
the calculation.
Proposed Kinetic Model
(R1)
k1[TDO]
7.4ϫ10–6 s–1
k1Ј[TDO][H+]–1
k2[TDO][ClO2]
k3[AIMSA][ClO2]
k4[ClO2][ClO]
k5[Cl2O3]
(1.56Ϯ0.05) ϫ10–7 ms–1
On the basis of the experimental results presented, the
following essential species must be considered for the con-
struction of the kinetic model: the reactants (TDO and
ClO2), the products (TTO, sulfate ions, chloride ions,
chlorate ions, and cyanamide), and indirectly identified or
feasible intermediates (such as HOCl, chlorite ions,
hydrogen sulfite ions). Furthermore, in our previous pa-
per,[19] we showed that TDO tends to rearrange slowly into
aminoiminomethanesulfinic acid (AIMSA) under acidic
conditions, and these forms have different reactivity
towards chlorine dioxide. Therefore, AIMSA has to be con-
sidered as a reactant as well. In addition, Cl2O2 was also
considered as a feasible intermediate in the well-known
(R2)
(R3)
(R4)
(R5)
(R6)
(R7)
(R8)
(R9)
(R10)
(R11)
(R12)
(R13)
(R14)
(R15)
(R16)
(2.74Ϯ0.12) ϫ10–2 –1 s–1
m
(2.85Ϯ0.10) ϫ103 m–1 s–1
109 m–1 s–1
Ն1 s–1
k6[AIMSA][Cl2O3][H+]
k7[Cl2O2]
(8.85Ϯ0.60)ϫ108 m–2 s–1
Ն1 s–1
k8[TTO][Cl2O2][H+]–1
k9[TDO][HOCl]
k10[TTO][HOCl]
(66.0Ϯ3.5) s–1
Ն108 m–1 s–1
(2.01Ϯ0.32)ϫ109 m–1 s–1
(17.0Ϯ0.2) m–1 s–1
109 m–1 s–1
–
k11[HSO3 ][ClO2]
–
k12[SO3 ][ClO2]
k13[HSO3 ][ClO2][H+]–1 (6.77Ϯ0.98)ϫ10–2 s–1
–
(1.20Ϯ0.10)ϫ105 m–1 s–1
8.2ϫ 109 m–2 s–1
–
k14[TDO][ClO2 ]
k15[HSO3–][ClO2 ][H+]
–
–
k16[HSO3 ][HOCl]
7.6ϫ 108 m–1 s–1
Eur. J. Inorg. Chem. 2015, 5011–5020
5014
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
AIMSA in the kinetic model may also mean two different
forms of thiourea dioxide, not specifically the dioxide and
the aminoimino form. Our calculations also indicated that
the transformation of two different forms of TDO is ac-
celerated by an increase of pH, which is reflected by the k1Ј
value indicated in Table 2. This result was indirectly con-
firmed by a further calculation that indicated that the
elimination of k1Ј from the kinetic model would lead to an
unacceptably high average deviation of 7.6%. Many other
reasonable possibilities were checked to eliminate k1Ј from
the proposed model with no success, including the hydrox-
ide dependence of the subsequent step of the reaction.
Thus, we concluded that the transformation between the
two forms of TDO in aqueous solution also depends on
pH.
Figure 7. Experimental (symbols) and fitted (solid lines) ab-
sorbance–time series for the ClO2–TDO reaction. The initial condi-
tions were as follows: [ClO2]0 = 0.51 mm, pH 1.87, I = 0.1 m, T =
25 °C, [TDO] = 10.0 (black), 8.0 (blue), 7.0 (green), 5.0 (cyan), 3.0
(red), 1.4 (magenta), 1.0 (brown), 0.7 (grey), 0.5 (purple), and
0.25 mm (yellow).
Step R2 is the initiation of the reaction by a formal oxy-
gen transfer from chlorine dioxide to TDO to form TTO
and a ClO radical. The rate coefficient of this reaction was
found to be k2 = (2.74Ϯ0.12)ϫ10–2 –1 s–1. Although this
m
value seems to be quite low, the elimination of this step
from the model would result in a significant increase in the
average deviation to 5.6%. From this, we concluded that
this reaction is necessary for the adequate description of
our kinetic data.
Step R3 is the other alternative way to initiate the reac-
tion by a direct attack of chlorine dioxide to AIMSA. As
there is no information about the rearrangement of
TTO to aminoiminomethanesulfonic acid (AIMSOA,
NH2NHCSO3H) in aqueous solution, we assume that this
reaction results in the formation of the same product as
Step R2. We calculated k3 to be (2.85Ϯ0.10) ϫ103 m–1 s–1,
that is, five orders of magnitude higher than k2. Therefore,
there is a significant reactivity difference between these
forms towards chlorine dioxide in aqueous solution. This
seems to be analogous to the case of acetone, which is more
sluggish to react with iodine than its enol form.[39]
Figure 8. Experimental (symbols) and fitted (solid lines) ab-
sorbance–time series for the ClO2–TDO reaction. The initial condi-
tions were as follows: [TDO]0 = 0.5 mm, pH 2.36, I = 0.1 m, T =
25 °C, [ClO2] = 1.4 (black), 1.0 (blue), 0.7 (green), 0.3 (cyan), and
0.18 mm (red).
Step R4 is just a rapid radical–radical reaction to pro-
duce Cl2O3, and its rate coefficient has to be close to the
Discussion
diffusion control limit. Therefore, we fixed it to k4
=
As indicated, step E1 is only necessary to take the slight 109 m–1 s–1 during the course of the whole calculation pro-
pH change into account during the course of the reaction. cess.
The ratio of rate coefficients for the rapid forward and re-
Step R5 is one of the possible routes for the rapid re-
verse reaction was fixed during the whole calculation pro- moval of Cl2O3 by hydrolysis. This reaction produces
cess to give the pKa of phosphoric acid to be 2.10 under chlorate ions and the key intermediate hypochlorous acid.
the given experimental conditions.[37]
Its individual rate coefficient cannot be determined from
The initiation of the reaction in our kinetic model starts our experiments; only the k6/k5 ratio could be calculated.
with the relatively slow rearrangement of TDO into Any value of k5 higher than 1 s–1 would lead to the same
AIMSA in aqueous solution.[19] For completeness, it should final result; therefore, we set k5 at the lower limit.
be noted that the calculation indicated that two different
Step R6 is the other pathway for the reaction of Cl2O3
forms of TDO are required to explain our kinetic data, a with AIMSA through a formal oxygen transfer process to
less and a more reactive form with a slow transformation produce Cl2O2 and TTO. As k6/k5 was found to be
process. Therefore, in addition to tautomerization, other (8.85Ϯ0.60) ϫ108 m–2, k5 was set to 1 s–1 to provide the
feasible explanation may exist to describe our kinetic data. actual value of k6. One might also expect that Cl2O3 could
One reasonable possibility is that TDO changes from a less react with TDO. We tried to include this reaction with no
reactive form (oligomeric) to a more reactive form upon success. Thus, we concluded that TDO does not react di-
dissolution of the solid in aqueous solution.[38] Unfortu- rectly with Cl2O3.
nately, there is no way to distinguish between these possibil-
Step R7 is the well-known hydrolysis of Cl2O2 to produce
ities in our kinetic measurements; consequently, TDO and chloride and chlorate ions. As these are short-lived interme-
Eur. J. Inorg. Chem. 2015, 5011–5020
5015
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
diates, the individual rate coefficient k7 cannot be calculated our value of k11 = (17.0Ϯ0.2) m–1 s–1 presented here is in a
from our experiments, and only k8/k7 could be determined. very good agreement with the one obtained from indepen-
We set k7 as 1 s–1 during the calculation process and actu- dent research groups (see below) if one considers that the
ally calculated k8/k7 owing to total correlation of these pa- pKa of hydrogen sulfite is ca. 7.0.[35,41,42] This reflects a
rameters.
value of 1.7ϫ106 m–1 s–1 for the rate coefficient of the sul-
Step R8 is the oxidation of TTO by Cl2O2 to produce the fite–chlorine dioxide reaction.
key intermediate HOCl through a nonelementary reaction.
Step R12 is also a rapid radical–radical reaction to pro-
This was found to be an important step to maintain the duce chlorite ions, the third key intermediate of the title
concentration of HOCl at an appropriate level to produce reaction. The rate coefficient of this step was fixed to
the sigmoidal kinetic curves. This reaction was already sug- 109 m–1 s–1 during the whole calculation process.
gested by Jones et al.[18] with a rate coefficient of
Step R13 is a formal oxygen transfer from chlorine diox-
8ϫ104 m–1 s–1, but we found no opportunity to obtain the ide to a sulfite ion. This means that the sulfite–chlorine di-
individual rate coefficient of this reaction from our mea- oxide reaction occurs through parallel oxygen- and elec-
surements as a result of the total correlation mentioned tron-transfer reactions. Again, if we consider that the pKa
above. We have also found that the rate of this reaction of hydrogen sulfite is 7.0, then we can obtain an apparent
is inversely proportional to [H+], which indicates that the rate coefficient of 6.77ϫ105 m–1 s–1 for the sulfite–chlorine
removal of chlorine dioxide is actually faster at higher pH dioxide reaction; therefore, the overall rate coefficient of the
values. As the kinetic model reported by Jones et al.[18] suf- parallel pathways is 2.4ϫ106 m–1 s–1, which is in sound
fers from other serious problems, which have been enlight- agreement with those found previously.[41–43] The most im-
ened by Stanbury and Figlar,[40] we are inclined not to com- portant conclusion is that we have provided indirect experi-
pare these rate coefficients directly.
mental evidence that the sulfite–chlorine dioxide reaction
Step R9 is one of the reactions responsible for the re- proceeds through parallel pathways.
moval of hypochlorous acid to produce hydrogen sulfite
Step R14 is the reaction between TDO and a chlorite ion
ions as another important intermediate. Evidently this step to form a hydrogen sulfite ion and hypochlorous acid, both
is also a nonelementary reaction, and there is no firm exper- of which are required to maintain the catalytic cycle. The
imental basis to divide it unambiguously into elementary rate coefficient of this reaction cannot be determined from
ones. Hypochlorous acid does not act as a formal oxygen our experiments because k14 and k15 were found to be in
transfer agent; instead, it breaks up the central C–S bond total correlation with each other. As we fixed k15 to
–
to produce HSO3 ions and cyanamide. The former inter- 8.2ϫ109 m–2 s–1, a value of (1.2Ϯ0.1)ϫ105 m–1 s–1 could be
mediate then participates in a chain reaction that results in calculated for the rate coefficient. The result for this reac-
the efficient removal of chlorine dioxide. This pathway will tion is in agreement with an independent result,[18] although
be discussed later. From our measurements, the individual the rate coefficients differ by a factor of 50. This difference
rate coefficient of this reaction cannot be determined, and may stem from the fact that the k15 value used in our fitting
only k10/k9 could be calculated from our experiments. As procedure may not be determined precisely. Indeed, inde-
we expect that the TDO–HOCl reaction should be relatively pendently reported values of this rate coefficient span the
fast, we set k9 to 108 m–1 s–1 to efficiently maintain the cata- range 2.2ϫ108–8.2ϫ109 m–2 s–1 (see below).[43–45] Although
lytic cycle with the controlling and terminating reaction of this may explain the difference quite well, at least one seri-
Step R16 (see below).
ous problem mentioned above questions the validity of the
Step R10 produces directly the major sulfur-containing kinetic model reported by Jones et al.[18] Therefore, we con-
end product, sulfate ions, through the TTO–HOCl reaction. clude that this may also account for the difference.
With k9 fixed to 108 m–1 s–1, we found a somewhat unexpec-
Steps R15 was first studied at approximately neutral pH
tedly high value for k10 of (2.0Ϯ0.3) ϫ109 m–1 s–1. Without and with a huge excess of chlorite ions by Huff-Hartz et
the direct investigation of this reaction, we cannot provide al.,[44] who found k15 to be 5.0ϫ108 m–2 s–1. Frerichs et al.
a reasonable explanation for this value. However, it obtained a k15 value of 2.2ϫ108 m–2 s–1 by simulating the
should be mentioned that the total correlation between the dynamic behavior of the chlorite–sulfite reaction in a con-
parameter sets opens up the possibility that the otherwise tinuous stirred-tank reactor. Under the closest experimental
fixed parameters of those reactions in which the key inter- conditions, a k15 value of 8.2ϫ109 m–2 s–1 was obtained for
mediates are involved (k15, k16) may change slightly. There- this reaction.[43] Therefore, we used this value for k15.
fore, a different k10/k9 value is possible, but it is difficult to
Step R16 was independently studied by Fogelman et
track the problem without direct experimental information. al.,[46] and its rate coefficient was determined to be
Step R11 is an electron-transfer process suggested by 7.6ϫ108 m–1 s–1. We directly adopted this value in our ki-
Merényi et al. as an initiation of the sulfite–chlorine dioxide netic model.
reaction in the presence of a huge chlorite concentration at
The direct reaction of TTO with chlorine dioxide and the
alkaline pH.[41] They found this rate coefficient to be rearrangement of TTO are not included in the proposed
2.6ϫ106 m–1 s–1. However, the rate of the SIV–chlorine diox- kinetic model. Our separate kinetic experiments on this re-
ide reaction decreases with decreasing pH.[42] As the pH action showed that a very slow reaction could be observed
was ca. 2.0 in our study and as sulfite (not hydrogen sulfite) within the timescale of the title reaction under the experi-
is the kinetically active species towards chlorine dioxide,[43] mental conditions studied (Figure 9), and the rate of decay
Eur. J. Inorg. Chem. 2015, 5011–5020
5016
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
of chlorine dioxide does not depend on the age of the solu- shown in Figure 10. The proposed model soundly explains
tion.
the effect of the addition of hydrogen sulfite ions on the
absorbance–time profiles (Figure 10A). The semiquantit-
ative explanation of the effect of methionine requires the
extension of the kinetic model with the rapid direct reaction
between hypochlorous acid and methionine to produce a
kinetically inactive product. In a fair agreement with the
experimental results, we found that the consumption of
chlorine dioxide is delayed significantly as a result of the
efficient removal of hypochlorous acid (Figure 10B). We
also found that the effect of chlorite ions can be semiquanti-
–
tatively described if the TDO–ClO2 reaction is extended
with a process that generates an intermediate or a product
(hydrogen sulfite alone is not enough to account for this
effect) that is able to remove chlorine dioxide efficiently
(Figure 10C). Evidently, Step R14 in the proposed model
is not an elementary reaction, and a detailed independent
investigation is required to elucidate the mechanism of the
Figure 9. Effect of aging on the thiourea trioxide–chlorine dioxide
reaction. The initial conditions were as follows: [TTO]0 = 10.0 mm;
[ClO2]0 = 0.41 mm; T = 25 °C; pH 2.36. The age of the TTO solu-
tion was 135 (black), 1355 (blue), 4135 (green), 9099 (cyan), and chlorite–TDO reaction itself. Nevertheless, these simula-
13650 s (red).
tions further support that the main core of the mechanism
presented here accounts for all of the experimentally ob-
served phenomena. For completeness, it should be men-
tioned that the disappearance of ClO2 indicated in Fig-
ure 10 (A and B) is much faster than that in the correspond-
ing experiments shown in Figures 3 and 5. However, as
indicated in the figure captions, these kinetic curves were
acquired at 18 °C, which is much lower than the tempera-
ture at which the parameters of the kinetic models were
determined. Consequently, the simulated results necessarily
show much faster decay of ClO2. This is consistent with our
From Figure 9, we can predict that the rate coefficient of
the first-order decomposition of TTO is ca. 10–6 s–1 under
the experimental conditions studied; this value is in very
good agreement with the value of 2.5ϫ10–6 s–1 found by
Makarov and co-workers.[47]
The inclusion of Equation (1) in the kinetic model with
a fixed rate coefficient of 2.5ϫ10–6 s–1 does not change the
average deviation, from which we concluded that this pro-
cess may only have a very minor contribution to the kinetics
of the title reaction. Qualitatively, the initial drop in the
absorbance can be explained by the formation of hydrogen
sulfite ions from the slow aging of TTO solution to ef-
ficiently remove ClO2 through the rapid hydrogen sulfite–
chlorine dioxide reaction already included in the proposed
kinetic model. It should be emphasized again that the de-
composition of TTO in acidic solution is very slow and,
therefore, cannot compete with the title reaction.
TTO + H2OǞN-containing products + HSO3 + H+
(1)
–
As the kinetic model was proposed on the basis of the
absorbance–time profiles of chlorine dioxide, it seems cru-
cial to test whether the proposed kinetic model is able to
account for the final concentration distribution of sulfate
ions and TTO, as measured at the end of the reaction by
HPLC. A comparison of the simulated data of Table 1 with
the measured data shows that the proposed kinetic model
predicts soundly the trends in the concentration distribu-
tion of the sulfur-containing end products, though one may
notice that the amount of TTO in each sample is underesti-
mated by approximately 0.1 mm. However, as the error in
the sulfur balance of the measured samples is ca. 0.11 mm,
this agreement also validates our results.
Figure 10. Simulation of the effect of initially added (A) hydrogen
sulfite ions, (B) methionine, and (C) chlorite ions on the TDO–
ClO2 reaction. The initial concentrations are given in Figures 5, 3,
and 4, respectively. For the simulation of the kinetic curves of (B)
and (C), the proposed model was supplemented by the rapid
methionine–hypochlorous acid reaction and a rapid reaction be-
The proposed kinetic model can be validated further by
simulating qualitatively the effect of the addition of
–
tween an intermediate/product of the TDO–ClO2 reaction with
hydrogen sulfite ions, chlorite ions, and methionine, as chlorine dioxide (see text).
Eur. J. Inorg. Chem. 2015, 5011–5020
5017
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
recent report, in which we showed that the half-life of ClO2 disappearance of chlorine dioxide. The kinetic model pro-
increased five to nine times as the temperature decreased posed here indicates that hypochlorous acid and hydrogen
from 25 to 11 °C.[19] From this, we concluded that the dif- sulfite ions are key intermediates for the autocatalytic de-
ferent temperature applied accounts for the half-life differ- crease of chlorine dioxide. The hydrogen sulfite oxidation
ence observed in the measured and calculated curves.
by oxychlorine compounds is responsible for pH autocatal-
Another important issue that should also be discussed is ysis[45] in unbuffered solution and may further enhance the
the possible production of sulfoxylate ions with aging of nonlinear nature of the title reaction. The elucidation of the
TDO, as it is not included in the proposed model. Our pre- mechanism of the chlorine dioxide–thiourea dioxide reac-
vious study[19] evidenced that the addition of formaldehyde tion may also contribute to a better understanding of the
(a well-known sulfoxylate scavenging agent) does not influ- mechanistic background of the oscillatory and chaotic
ence the reaction; consequently, the decomposition of TDO chlorite–thiourea system,[48,49] for which an autocatalytic
to produce sulfoxylate ions can be ruled out.
loop involving the key intermediates hypochlorous acid,
Finally, an explanation of the origin of the sigmoidal hydrogen sulfite ions, and chlorite ions might also play a
shape of the kinetic curves should be discussed. The reac- critical role. More importantly, our results suggest that the
tion starts with the slow direct reactions of TDO and elucidation of the reactions of TDO in aqueous acidic solu-
AIMSA to produce HOCl. The hypochlorous acid then tion should be taken with special circumspection owing to
produces hydrogen sulfite ions from TDO, followed by the the possible reactivity difference of the two forms of TDO
hydrogen sulfite–chlorine dioxide reaction to produce both presented. Therefore, the use of differently aged TDO solu-
chlorite ions and Cl2O3 through Steps R11 and R13. The tions may lead to a misinterpretation of the results.
well-known rapid reaction of chlorite ions with sulfite ions
and TDO under the pH range studied leads again to the
formation of hypochlorous acid. Concurrently, the TDO–
ClO2 reaction partially rebuilds the hydrogen sulfite ion
Experimental Section
–
Material and Buffers: The TDO (Ն98%) was purchased from
Sigma–Aldrich and used in all experiments without further purifi-
cation. The other reagents were of analytical grade, including phos-
phoric acid, potassium dihydrogen phosphate, sulfuric acid, potas-
sium nitrate, tetrabutylammonium hydroxide (TBAOH), and hexa-
dimethrine bromide (HDB). Methanol (chromatographic grade)
was applied in HPLC as a component of the mobile phase. The
solution of the oxidizing agent chlorine dioxide was prepared ac-
cording to the previous literature method[19] and was protected
from light at low temperature. The ClO2 concentration was stan-
dardized spectrophotometrically at λ = 360 nm (ε ≈ 1200 m–1 cm–1).
All working solutions were prepared in Milli-Q water. The pH
value of the reaction system was controlled with phosphoric acid/
dihydrogenphosphate with the pKa of phosphoric acid taken to be
2.1. The ionic strength was always maintained at 0.1 m with a 0.1 m
potassium dihydrogen phosphate buffer component.
concentration to keep the catalytic cycle alive. The autocat-
alytic nature of the system is manifested when Step R8
takes the governing role to increase [HOCl]. However, the
concentration of hypochlorous acid is further regulated by
its reactions with TDO, TTO, and hydrogen sulfite ions;
therefore, under certain experimental conditions, sigmoidal
kinetic curves cannot be observed. The overall effect is also
enhanced by the slow rearrangement of TDO and the re-
markable reactivity difference of the two forms of TDO in
aqueous acidic solution. This result also implies that the
kinetic model of the chlorite–TDO reaction[18] should be
revised thoroughly, not only because of the questionable
significance of the vanishingly slow chlorine dioxide–chlor-
ide reaction realized correctly by Stanbury and Figlar.[41]
On one hand, the slow rearrangement of TDO was not
taken into consideration by Jones et al.,[18] and, most im-
portantly, on the other hand, the autocatalytic loop involv-
ing hypochlorous acid, hydrogen sulfite ions, chlorite ions,
and chlorine dioxide seems to also play a key role in that
reaction. It should also be noted that the present kinetic
model is a single feasibility to describe our kinetic data.
Alternative models may also exist, but we do not see any
known experiments that would contradict our proposed
model at this point. Further refinement is expected after
Methods: Fresh TDO solutions were prepared by dissolving solid
samples in a certain volume of buffer solution for each kinetic mea-
surement. The reaction was initiated by mixing appropriate vol-
umes of TDO and ClO2 solutions in a quartz cuvette with 1 cm
optical path at (25Ϯ0.1) °C, and all kinetic experimental data were
recorded with an S600 diode array spectrophotometer (λ = 200–
400 nm) equipped with a Peltier thermostat and magnetic-stirring
attachments. For all kinetic experiments, it was necessary to ensure
that the age of the TDO solution was always kept the same. All of
the kinetic experiments were initiated by a TDO solution aged for
the elucidation of the kinetics of the subsystems of the title 180 s. The reaction was investigated at five pH values by changing
the initial concentrations of ClO2 and TDO. Several series of reac-
tions were followed by CE and HPLC to clarify the important in-
termediates and the end products of the reactions. The HPLC sepa-
rations were performed with a Dionex chromatographic system
consisting of a Model P680 pump, a Phenomenex Ginimi C18 sep-
aration column (250ϫ4.6 mm, 5 μm), and a UVD 170 UV detec-
tor. The eluent was a mixture of 1 mmol/dm3 tetrabutylammonium
hydroxide (TBAOH) aqueous solution (pH 6.7) and methanol
(95 vol-%) at a flow rate of 0.4 cm3 min–1. Capillary electrophoresis
analysis with indirect ultraviolet detection was performed with a
reaction, but the central part of the model seems to be
firmly established.
Conclusions
As a subsystem of the chlorite–thiourea reaction, the
chlorine dioxide–thiourea dioxide reaction can be used to
distinguish the reactivity of two different forms of thiourea
dioxide and also displays sigmoidal kinetic curves for the Beckman Coulter P/ACE MDQ CE system. The CE running buffer
Eur. J. Inorg. Chem. 2015, 5011–5020
5018
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
was composed of a mixture of 2.0ϫ10–5 mm HDB and 20 mm
KNO3 at pH 7.0 (adjusted with sodium hydroxide). A fused silica
capillary of 57.5 cm (75 μm i.d., 375 μm o.d.) was used. The condi-
tions for the kinetic runs were as follows: separation voltage 15 kV,
capillary temperature 25 °C, cathodic pressure injection for 5 s with
0.5 psi. The carbon-containing products of the ClO2–TDO reaction
were identified by 13C NMR spectroscopy (Bruker AVANCE III
HD 600 MHz) with D2O as a frequency lock. A certain amount of
solid P2O5 was added to the solution to control the pH, and the
reaction was initiated by the addition of ClO2 solution to the TDO
solution. The collection of the 13C NMR spectroscopic data was
started when the color of ClO2 disappeared. In addition, the 13C
NMR spectra of urea and cyanamide were acquired for compari-
son.
[8] S. V. Makarov, Russ. Chem. Rev. 2001, 70, 885–895.
[9] H. Wang, Q. Li, C. Gao, J. Cleaner Prod. 2014, 84, 701–706.
[10] A. A. Cuadri, P. Partal, F. J. Navarro, M. Garcia-Morales, C.
Gallegos, Fuel 2011, 90, 2294–2300.
[11] C. K. Chua, A. Ambrosi, M. Pumera, J. Mater. Chem. 2012,
22, 11054–11061.
[12] Y. Wang, L. Sun, B. Fugetsu, Bull. Chem. Soc. Jpn. 2012, 85,
1339–1344.
[13] S. V. Makarov, R. Silaghi-Dumitrescu, J. Sulfur Chem. 2013,
34, 444–449.
[14] Q. Gao, B. Liu, L. Li, J. Wang, J. Phys. Chem. A 2007, 111,
872–877.
[15] S. V. Makarov, C. Mundoma, J. H. Penn, S. A. Svarovsky,
R. H. Simoyi, J. Phys. Chem. A 1998, 102, 6786–6792.
[16] E. Mambo, R. H. Simoyi, J. Phys. Chem. 1993, 97, 13662–
13667.
Data Treatment: The absorbance at λ = 360 nm, at which only
chlorine dioxide absorbs light, was used for evaluation. Absorbance
values higher than 1.2 were excluded from the data evaluation, be-
cause the absorbance–concentration curves start to deviate from
linearity above this value. The concentration ranges of the reactants
ClO2 and TDO were varied between 0.18–1.4 and 0.25–10.0 mm,
respectively, and the pH was kept between 1.87 and 3.07. For the
HPLC and CE tests, calibration curves for all detectable species
were determined to transform the measured peak areas into con-
centrations after qualitative information was collected. The corre-
lation coefficients were always above 0.999, which indicates that the
relationship between the peak area and the concentration is per-
fectly linear. To obtain the kinetic parameters for the proposed
model, the ZiTa/Chemmech program package[50] was used. Our cri-
terion for an acceptable fit for the relative fitting procedure to ob-
tain reliable kinetic parameters for the model was to achieve 4%,
as the precision of the concentration determination of the volatile
chlorine dioxide is ca. 3–5% during the course of the reaction.
[17] C. R. Chinake, R. H. Simoyi, S. B. Jonnalagadda, J. Phys.
Chem. 1994, 98, 545–550.
[18] J. B. Jones, C. R. Chinake, R. H. Simoyi, J. Phys. Chem. 1995,
99, 1523–1529.
[19] G. Cseko, Y. Hu, Y. Song, T. R. Kégl, Q. Gao, S. V. Makarov,
˝
A. K. Horváth, Eur. J. Inorg. Chem. 2014, 1875–1879.
[20] S. V. Makarov, E. V. Kudrik, Russ. Chem. Bull. Int. Ed. 2001,
50, 203–205.
[21] V. C. H. Wu, A. Rioux, Food Microbiol. 2010, 27, 179–184.
[22] Y. Han, T. L. Selby, K. K. Schultze, P. E. Nelson, R. H. Linton,
J. Food Prot. 2004, 67, 2450–2455.
[23] I. Popa, E. J. Hanson, E. C. D. Todd, A. C. Schilder, E. T.
Ryser, J. Food Prot. 2007, 70, 2084–2088.
[24] Y. Lee, G. Burgess, M. Rubino, R. Auras, J. Food Prot. 2015,
144, 20–28.
[25] G. Rábai, R. T. Wang, K. Kustin, Int. J. Chem. Kinet. 1993,
25, 53–62.
[26] J. N. Figlar, D. M. Stanbury, J. Phys. Chem. A 1999, 103, 5732–
5741.
[27] A. K. Horváth, I. Nagypál, J. Phys. Chem. A 1998, 102, 7267–
7272.
[28] C. Pan, D. M. Stanbury, J. Phys. Chem. A 2014, 118, 6827–
6831.
Acknowledgments
[29] A. K. Horváth, I. Nagypál, I. R. Epstein, J. Phys. Chem. A
2003, 107, 10063–10068.
This work was supported by the National Natural Science Founda-
tion of China (NSFC) (grant number 51221462), by the Natural
Science Foundation of Jiangsu Province (grant number
BK20131111), the Fundamental Research Funds for the Central
Universities (grant number 2015QNA19), the Priority Academic
Program Development of Jiangsu Higher Education Institutions
(grant number JS-2014-37) and by the Hungarian Research Fund
OTKA (grant number K116591). The authors are grateful for the
financial support of Chinese–Hungarian Cooperative (grant
number K-TÉT-CN-1-2012-0030). Financial support by the Social
Renewable Operational Program (SROP-4.2.2.D-15/1/Konv-2015-
0015, Synthesis of supramolecular systems, examination of their
physico-chemical properties and their utilization for separation and
sensor chemistry project) is also gratefully acknowledged.
[30] G. Cseko, A. K. Horváth, J. Phys. Chem. A 2012, 116, 2911–
˝
2919.
[31] L. Xu, G. Cseko, A. Petz, A. K. Horváth, J. Phys. Chem. A
˝
2014, 118, 1293–1299.
[32] W. Gao, W. Qi, J. Lai, L. Qi, S. Majeed, G. Xu, Chem. Com-
mun. 2015, 51, 1620–1623.
[33] X. L. Armesto, L. M. Canle, M. I. Fernandez, M. V. Garcia,
J. A. Santabella, Tetrahedron 2000, 56, 1103–1109.
[34] D. I. Pattison, M. J. Davies, Chem. Res. Toxicol. 2001, 14,
1453–1464.
[35] A. K. Horváth, I. Nagypál, J. Phys. Chem. A 2006, 110, 4753–
4758.
[36] G. Peintler, I. Nagypál, I. R. Epstein, J. Phys. Chem. 1990, 94,
2954–2958.
[37] IUPAC Stability Constant Database, Royal Society of Chemis-
try, London, 1990–1997.
[38] J. Shao, X. Liu, P. Chen, Q. Wu, X. Zheng, K. Pei, J. Phys.
Chem. A 2014, 118, 3168–3174.
[39] M. J. Pilling, P. W. Seakins, Reaction Kinetics, Oxford Univer-
sity Press, Oxford, UK, 1995.
[40] D. M. Stanbury, J. N. Figlar, Coord. Chem. Rev. 1999, 187, 223–
[1] M. Alamgir, I. R. Epstein, Int. J. Chem. Kinet. 1985, 17, 429–
439.
[2] C. J. Doona, R. Blittersdorf, F. W. Schneider, J. Phys. Chem.
1993, 97, 7258–7263.
[3] C. R. Chinake, R. H. Simoyi, J. Phys. Chem. 1994, 98, 4012–
4019.
232.
[4] I. R. Epstein, K. Kustin, R. H. Simoyi, J. Phys. Chem. 1992,
96, 5852–5856.
[5] Z. Kis, S. V. Makarov, R. Silaghi-Dumitrescu, J. Sulfur Chem.
2010, 31, 27–39.
[6] S. A. Svarovsky, R. H. Simoyi, S. V. Makarov, J. Phys. Chem.
B 2001, 105, 12634–12643.
[7] S. V. Makarov, A. K. Horváth, R. Silaghi-Dumitrescu, Q. Gao,
Chem. Eur. J. 2014, 20, 14164–14176.
[41]
G. Merényi, J. Lind, X. Shen, J. Phys. Chem. 1988, 92, 134–
137.
K. Suzuki, G. Gordon, Inorg. Chem. 1978, 17, 3115–3118.
A. K. Horváth, I. Nagypál, I. R. Epstein, Inorg. Chem. 2006,
45, 9877–9883.
[42]
[43]
[44]
K. E. Huff-Hartz, J. S. Nicoson, L. Wang, D. W. Margerum,
Inorg. Chem. 2003, 42, 78–87.
Eur. J. Inorg. Chem. 2015, 5011–5020
5019
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
[45]
[48] R. Rushing, R. C. Thomson, Q. Gao, J. Phys. Chem. A 2000,
104, 11561–11565.
G. A. Frerichs, T. M. Mlnarik, R. J. Grun, R. C. Thomson, J.
Phys. Chem. A 2001, 105, 829–837.
[49] Q. Gao, J. Wang, Chem. Phys. Lett. 2004, 391, 349–353.
[50] ZiTa/Chemmech: A Comprehensive Program Package for Fitting
Parameters of Chemical Reaction Mechanism, versions 2.1–5.0,
G. Peintler, 1989–2001
[46] K. D. Fogelman, D. M. Walker, D. W. Margerum, Inorg. Chem.
1989, 28, 986–993.
[47] S. V. Makarov, C. Mundoma, J. H. Penn, J. L. Petersen, S. A.
Svarovsky, R. H. Simoyi, Inorg. Chim. Acta 1999, 286, 149–
154.
Received: June 25, 2015
Published Online: October 5, 2015
Eur. J. Inorg. Chem. 2015, 5011–5020
5020
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim