Bisulfite-driven autocatalysis in the bromate-thiosulfate reaction in a slightly acidic medium

Article abstract of DOI:10.1021/ic302019k

The thiosulfate-bromate reaction has been studied by high-performance liquid chromatography, monitoring the concentrations of thiosulfate and tetrathionate simultaneously. It is found that concentration-time curves of both species display a sigmoidal shape in a slightly acidic, well-buffered medium. Unlike the previously reported complex reaction systems involving bromate, this nonlinear dynamical behavior originates from neither proton nor bromine(III) autocatalysis under our experimental conditions. We demonstrated that sulfur(IV) species significantly accelerates the reaction; therefore, it acts as an autocatalyst. To the best of our knowledge, no reaction system has yet been reported among the pH-driven oxysulfur-oxyhalogen systems, where sulfur(IV) has such a remarkable role. On the basis of the simultaneous evaluation of [S ₂O₃^2-] and [S₄O₆ ^2-] time series, an eight-step kinetic model is proposed to account for the experimental observations. The model employed here may serve as a solid starting point to extend it for other oxysulfur-oxyhalogen systems where such a seemingly general phenomenon may become observable.

Full text of DOI:10.1021/ic302019k

Communication
pubs.acs.org/IC
Bisulfite-Driven Autocatalysis in the Bromate−Thiosulfate Reaction
in a Slightly Acidic Medium
Zhen Wang, Qingyu Gao,* Changwei Pan, Yuemin Zhao, and Attila K. Horvat
†
,†
†
†
h*,‡,§
́
†
‡
§
College of Chemical Engineering, China University of Mining and Technology, Xuzhou 221116, People's Republic of China
Department of Inorganic Chemistry, University of Pec
Janos Szentagothai Research Center, Ifjusag ut
S Supporting Information
́ ́
́ ́ ́
s, Ifjusag ja 6, H-7624 Pecs, Hungary
ut
́
́
́
́
́
ja 20, H-7624 Pec
́
s, Hungary
*
The thiosulfate−bromate reaction was followed by high-
performance liquid chromatography (HPLC); the details are
given in the Supporting Information (SI). Figure 1 shows that the
ABSTRACT: The thiosulfate−bromate reaction has been
studied by high-performance liquid chromatography,
monitoring the concentrations of thiosulfate and tetrathi-
onate simultaneously. It is found that concentration−time
curves of both species display a sigmoidal shape in a
slightly acidic, well-buffered medium. Unlike the previously
reported complex reaction systems involving bromate, this
nonlinear dynamical behavior originates from neither
proton nor bromine(III) autocatalysis under our exper-
imental conditions. We demonstrated that sulfur(IV)
species significantly accelerates the reaction; therefore, it
acts as an autocatalyst. To the best of our knowledge, no
reaction system has yet been reported among the pH-
driven oxysulfur−oxyhalogen systems, where sulfur(IV)
has such a remarkable role. On the basis of the
Figure 1. Measured (dots) and calculated (lines) concentration−time
−
2−
2
−
2−
curves at [BrO ] = 10.0 mM and pH = 4.75. [S O ] /mM = 0.2
simultaneous evaluation of [S O3 ] and [S O ] time
3 0 2 3 0
2
4
6
(black), 0.3 (blue), 0.5 (green), 0.7 (cyan), 1.0 (red), 1.4 (magenta), and
series, an eight-step kinetic model is proposed to account
for the experimental observations. The model employed
here may serve as a solid starting point to extend it for
other oxysulfur−oxyhalogen systems where such a
seemingly general phenomenon may become observable.
2
.0 (yellow). Filled and empty symbols denote the concentrations of
2−
2−
2−
S2O3 and S O6 , respectively. Note that the time for the [S O ] is
4
4
6
shifted by 10000 s along the x axis for better visibility.
formation of tetrathionate as well as the consumption of
thiosulfate follows a sigmoidal-shaped curve, meaning that
some sort of autocatalysis is involved. It should be emphasized
that the appearance of a sigmoidal-shaped concentration−time
curve of a product alone does not necessarily mean autocatalysis.
A striking counterexample (often seen in common textbooks but
not emphasized clearly) is a simple first-order consecutive
process via a long-lived intermediate. In that case, the formation
of a product proceeds through an apparent induction period
because its formation is delayed, but no autocatalysis is involved.
If, however, the sigmoidal-shaped concentration−time curves of
the reactant and product are simultaneously observed, the
presence of autocatalysis is clearly suggested. The phenomenon
of autocatalysis can easily be proven experimentally by the
addition of an autocatalyst to the reacting solution, which results
in a shortening of the induction period. Because the thiosulfate−
bromate reaction may have two limiting stoichiometries
xyhalogen and oxysulfur compounds have been exten-
O
sively used to study oscillations and pattern formations in
1 2,3 4
recent years. In the case of the catalyzed and uncatalyzed
bromate-driven system, positive feedback arises from bromine-
III) autocatalysis, but this phenomenon generally occurs at
strongly acidic conditions, where the pH is often less than 1.0. In
(
5
the case of the pH-driven [BrO (x = 2, 3) oxysulfur substrate ]
x
+
systems, the nonlinearity, however, is usually explained by H
autocatalysis. Among these systems, the possibility of the
autocatalytic feature being maintained by substances other than
protons has so far avoided any significant attention.
In this Communication, we prove that the batch thiosulfate−
bromate reaction in a slightly acidic, buffered medium (pH =
4
.25−5.0) is autocatalytic with respect to sulfur(IV) species. The
fact that this species is not a final product of the reaction,
meaning that the autocatalysis is driven by a long-lived
intermediate, gives a further unique feature of this system. This
phenomenon was discovered meanwhile by investigating the
detailed kinetics of the thiosulfate−bromate reaction, but we
believe it is worth reporting because the autocatalytic role of
sulfur(IV) species may be more general in oxyhalogen−oxysulfur
chemistry.
6
S2O32 + BrO3− + 6H → 3S O + Br + 3H O
−
+
2−
−
(1)
(2)
4
6
2
S2O3 + 4BrO3− + 3H O → 6SO4 + 4Br + 6H+
2−
2−
−
3
2
Received: September 17, 2012
Published: November 1, 2012
©
2012 American Chemical Society
12062
dx.doi.org/10.1021/ic302019k | Inorg. Chem. 2012, 51, 12062−12064

Inorganic Chemistry
Communication
tetrathionate and especially bromide ions might be good
candidates as autocatalysts. Figure 2, however, clearly demon-
Figure 3. Effect of sulfite ion on the measured (symbols) and calculated
2
−
(
[
solid lines) concentration−time curves at [S O ]0 = 0.5 mM,
2
3
−
2
−
BrO ] = 10.0 mM, and pH = 5.0. [SO ]0/mM = 0.0 (black), 0.005
3
0
3
(
blue), 0.01 (green), 0.02 (cyan), and 0.04 (red). Filled and empty
2
−
2−
symbols denote [S O ] and [S O6 ], respectively. Note that the time
for the [S O6 ] is shifted by 20000 s along the x axis for better visibility.
2
3
4
2
−
4
nature is more pronounced at high pHs and low reactant
concentrations. (b) The positive feedback cycle is maintained by
steps R3 and R6 producing HSO3− autocatalytically, resulting in
an increase of its concentration. Although at first sight this would
suggest that bromite is also involved in the autocatalytic loop,
one should bear in mind that bromite is consumed
instantaneously (see the rate coefficients of R2 and R3) as long
as thiosulfate is present; therefore, its concentration is controlled
−
2−
Figure 2. Effect of Br (A) and S O (B) on the measured (symbols)
4
6
2
−
and calculated (lines) concentration−time curves at [S O ]0 = 0.5
mM, [BrO ] = 10.0 mM, and pH = 5.0. (A) [Br ] /mM = 0.0 (black),
.0 (blue), and 2.0 (green). (B) [S O ] /mM = 0.0 (black), 0.25
blue), and 0.5 (green). Filled and empty symbols stand for [S O
−10
2
3
−
−
at such a low level (around 10 M) that it cannot act as an
3
0
0
2
−
autocatalyst. The observation that the addition of bromite only
decreases the initial concentration of thiosulfate (see Figure S4 in
the SI) also supports this statement. The concentration of
1
(
4 6 0
2
−
]
2
3
2−
and [S O6 ], respectively.
4
bisulfite, however, continuously increases in the induction period
5
(
it may even reach the 5 × 10− M level under our experimental
strates that neither of them is responsible for autocatalysis under
our experimental conditions. Hypobromite and bromite ions
have also been probed to act as possible autocatalysts, but in both
cases, we found that a trace amount of these species results in a
conditions) because its removal via step R7 is slower than its
production via step R3 and via the indirect route of step R2
followed by step R5. After thiosulfate is completely consumed,
excess bromate diminishes bisulfite quickly via steps R6−R8. In
addition to the clear qualitative picture, it should also be noted
that kR6 is in very sound agreement with the value reported
2−
sudden drop of the [S O ] and a simultaneously rapid rise of
2
3
S4O62 ] (see the SI). This fact can be easily explained by rapid
−
[
8
reactions between thiosulfate and bromite as well as between
thiosulfate and hypobromite. Because bromine also reacts very
rapidly with thiosulfate and tetrathionate, this indicates that
previously in an unbuffered medium, which gives strong support
of our kinetic model. (c) The formation of tetrathionate is usually
6
explained by the rapid step R4 in different oxidation reactions of
9
none of the conceivable bromine-containing species is
responsible for the autocatalytic effect.
thiosulfate. We found the rate coefficient of this reaction to be in
total correlation with k , meaning that we could calculate only
R5
A further systematic search among the possible intermediates
showed that the addition of a trace amount of sulfite significantly
shortened the induction period, as shown in Figure 3. These
experiments clearly prove that (bi)sulfite is the autocatalyst
under our experimental conditions. We also noticed that, in the
absence of initially added sulfite, the induction period is more
pronounced at lower thiosulfate (see Figure 1) and bromate
concentrations (see the SI) but a decrease of the pH significantly
shortens the induction period. To account for all of the
experimental observations, we employed the following kinetic
the ratio of k /k from our experiments. The essence of step R5
is also well understood; this helps to provide an additional source
R5 R4
of bisulfite besides step R3. Because the rate of step R5 is
+
proportional to [H ], it also provides a straightforward
explanation of the pH dependence of the kinetic curves in
which the autocatalytic nature drives the reaction. (d) Lee and
10
Lister reported that thiosulfate reacts rapidly with bromite in
two parallel pathways, leading to the formation of tetrathionate
and sulfate. This observation is also fulfilled by our kinetic model
because the bromite−thiosulfate reaction has to be fast enough
to compete with the rapid removal of bromite via step R7. An
opposite case would mean that (bi)sulfite cannot be accumulated
to such an extent to be the autocatalyst. The only difference
between our study and Lee and Lister’s report is that the formal
kinetic order of thiosulfate is greater than 1 under our
experimental conditions. This fact was confirmed by nonlinear
7
model by program package ZiTa. Rate coefficients determined
by nonlinear simultaneous parameter estimation are illustrated in
Table 1. The average deviation was found to be 2.8% by a relative
fitting procedure. Altogether only five fitted parameters were
used, and the rest of the parameters were directly taken from
previous reports. The SI contains all of the fitted kinetic data
besides Figures 1−3.
The following facts have to be emphasized to support the
kinetic model. (a) The rate equation of step R1 controls the
induction period of the reaction, indicating that the autocatalytic
parameter estimation because if we ignore k ′ from the model,
R2
the average deviation would increase to 4.5%, indicating
2−
systematic deviations at higher [S O3 ]. The difference may
2
be explained by quite different experimental conditions. (e) The
1
2063
dx.doi.org/10.1021/ic302019k | Inorg. Chem. 2012, 51, 12062−12064

Inorganic Chemistry
Communication
a
Table 1. Fitted and Fixed Rate Coefficients of the Proposed Model
reaction
no.
chemical equation
rate equation
parameter
2
R1
S2O3 + BrO3− + H+ → S O OH + HBrO2
2−
−
2−
−
+
59.4 ± 4.5 M− s−1
k [S O ][BrO3 ][H ]
R1 2 3
2
3
8
−1
R2
S2O3 + BrO − → 2S O OH + Br−
2−
−
2−
−
2− 2
−
kR2 = 10 M s−1, k ′/k = 1900 ± 600
2
k [S O ][BrO2 ] + k ′[S O ] [BrO2
]
R2
R2
2
2
3
R2
2
3
R2
2
3
−1
M
R3
R4
R5
R6
R7
R8
S2O3 + BrO − + H O → 2HSO + Br−
2−
−
2−
−
k /k = 1.42 ± 0.22
R3 R2
k [S O ][BrO2
]
2
2
3
R3
2
3
>10 M s−1
2
−1
S2O3 + S O OH + H+ → S O + H O
2−
−
2−
2−
−
kR4[S O ][S O OH ]
2
3
4
6
2
2
3
2
3
BrO3 + S O OH + H O → 2HSO + BrO2− + H+
−
−
−
k [S O OH ][BrO ][H ]
R5 2 3 3
−
−
+
k /k = 51.2 ± 7.1 M−1
R5 R4
2
3
2
3
HSO3 + BrO − → SO42− + BrO − + H+
−
−
−
+
7460 ± 980 M−2 s−1
k [HSO ][BrO ][H ]
3
2
R6
3
3
−
−
3 × 10 M s−1
7
−1
HSO3 + BrO2− → SO42− + HOBr
−
kR7[HSO3 ][BrO2
]
−
10 M s−1
9
−1
HSO3− + HOBr → SO42− + Br− + 2H+
kR8[HSO3 ][HOBr]
a
No error indicates that the given value was fixed during the fitting procedure.
1
1
rate coefficient k was reported by Hartz et al., and it was fixed
AUTHOR INFORMATION
Corresponding Author
E-mail: gaoqy@cumt.edu.cn (Q.G.), horvatha@gamma.ttk.pte.
hu (A.K.H.).
Notes
R7
■
*
during the calculation process. It should also be noted that,
although bromite is the major bromine(III) species under our
experimental conditions, bromous acid (pK = 3.43) might also
be able to open up other pathways at different experimental
conditions, especially at higher bromate excess and lower pHs
throughout the well-known equilibrium
12
a
The authors declare no competing financial interest.
−
+
ACKNOWLEDGMENTS
HBrO + BrO + H ↔ Br O + H O
(3)
■
2
3
2
4
2
This work was supported, in part, by Grants 21073232 and
5092002 from the National Natural Science Foundation of
China, Grant BK2011006 from the Jiangsu Natural Science
Foundation, PAPD, and the Hungarian Research Fund (OTKA
Grant CK78553).
followed by subsequent reactions of Br O such as in eq 4:
2
4
S2O32 + Br O + H O → S O OH + 2BrO + H+
−
−
−
(4)
2
4
2
2
3
2
Under our experimental conditions, however, the roles of eqs 3
and 4 in autocatalysis can be ruled out (see the SI, Figure S4), and
this makes another significant difference between the present
REFERENCES
■
3
(
1) (a) An Introduction to Nonlinear Chemical Dynamics, Oscillations,
Waves, Patterns, and Chaos, 1st ed.; Epstein, I. R., Pojman, J. A., Eds.;
Oxford University Press: New York, 1998. (b) Horvat
DeKepper, P. Science 2009, 324, 772−775. (c) Jimenez, Z.; Marts, B.;
Steinbock, O. Phys. Rev. Lett. 2009, 102, 244101. (d) Bansagi, T., Jr.;
reaction and other bromate-driven systems exhibiting nonlinear
dynamics. The same can also be said about the role of HOBr. To
be completely consistent with our experiments, we tentatively
propose eq 5 to be included with a reasonably large rate
coefficient to explain the perturbation effect of hypobromite
initially added to the reacting solution.
́
h, J.; Szalai, I.;
́
́
Vanag, V. K.; Epstein, I. R. Science 2011, 331, 1309−1312.
(
(
8
2) Zhabotinsky, A. M. Biofizika 1964, 9, 36.
3) Field, R. J.; Koros, E.; Noyes, R. J. Am. Chem. Soc. 1972, 94, 8649−
̈
̈
S2O32 + HOBr → S O OH + Br
−
−
−
2
3
(5)
664.
4) (a) Orban
Kurin-Cso
(
́
, M.; Koros, E. Nature 1978, 273, 371−372. Orban, M.;
̈
̈
́
However, eq 5 was found to be unnecessary for the description of
the kinetic curves, except for the ones in which the effect of the
initially added hypobromite was investigated. Therefore, we
would rather not include it in the final model. The reason can also
be understood easily because after initiation there is already
enough bisulfite to prevent the accumulation of HOBr and, as
was pointed out previously, the bisulfite concentration starts to
decrease only after thiosulfate is completely removed. As a result,
there is no need for eq 5 to remove hypobromous acid unless it is
intentionally added to study its effect.
Of course, the proposed model may further be improved
especially at higher thiosulfate concentrations, but it is per se
capable of explaining all of the most inevitable experimental
findings, providing a solid starting point for future studies to
explore the more intimate details of this promising system.
̈
rgei, K.; Zhabotinsky, A. M.; Epstein, I. R. J. Am. Chem. Soc.
998, 120, 1146−1150.
5) (a) Edblom, E. C.; Luo, Y.; Orban
Phys. Chem. 1989, 93, 2722−2727. (b) Szan
Chem. A 2005, 109, 5398−5402. (c) Orban, M.; Epstein, I. R. J. Phys.
Chem. 1995, 99, 2358−2362.
(6) Varga, D.; Horvath, A. K. J. Phys. Chem. A 2009, 113, 9988−9996.
7) Peintler, G. ZiTa: a comprehensive program package for fitting
parameters of chemical reaction mechanism, version 5.0; Attila Jozsef
1
(
́
, M.; Kustin, K.; Epstein, I. R. J.
to, T.; Rabai, G. J. Phys.
́
́
́
́
́
(
́
University: Szeged, Hungary, 1989−1998.
(
(
2
2
(
8) Szirovicza, L.; Boga, E. Int. J. Chem. Kinet. 1989, 30, 869−874.
9) (a) Voslar, M.; Matejka, P.; Schreiber, I. Inorg. Chem. 2006, 45,
, G.; Horvath, A. K. Inorg. Chem.
824−2834. (b) Rauscher, E.; Cseko
̈
́
011, 50, 5793−5802.
10) Lee, C. L.; Lister, M. W. Can. J. Chem. 1979, 57, 1524−1530.
(11) Hartz, K. E. H.; Nicoson, J. S.; Wang, L.; Margerum, D. W. Inorg.
Chem. 2003, 42, 78−87.
(
1
12) Faria, R. B.; Epstein, I. R.; Kustin, K. J. Phys. Chem. 1994, 98,
ASSOCIATED CONTENT
■
363−1367.
*
S
Supporting Information
Details of the kinetic experiments and additional measured and
calculated kinetic curves. This material is available free of charge
via the Internet at http://pubs.acs.org.
1
2064
dx.doi.org/10.1021/ic302019k | Inorg. Chem. 2012, 51, 12062−12064