Oxyhalogen sulfur chemistry: Kinetics and mechanism of the oxidation of a Bunte salt 2-aminoethanethiolsulfuric acid by chlorite

Article abstract of DOI:10.1039/b106167b

Spectrophotometric and ¹H NMR methods have been used to study the kinetics and mechanism of oxidation of the Bunte salt, 2-aminoethanethiol sulfuric acid, H₂NCH₂CH₂S-SO₃H (AETSA) by chlorite in mildly acidic media. The reaction is characterized by a long quiescent induction period followed by rapid and autocatalytic production of chlorine dioxide. The formation of chlorine dioxide is much more pronounced in stoichiometric excess of chlorite. The stoichiometry of the reaction in excess chlorite just before formation of chlorine dioxide was determined to be: 2ClO₂^- + H₂NCH₂CH₂S-SO₃H → ClNHCH₂CH₂SO₃H + SO₄^-2 + Cl^- + H⁺, while in excess AETSA the stoichiometry was: 3ClO₂^- + 2H₂NCH₂CH₂S-SO₃H + 2H₂0 → 2NH₂CH₂CH₂SO₃H + 2SO₄^-2 + 3Cl^- + 4H⁺. Although the products in excess chlorite also included pure taurine and dichlorotaurine, monochlorotaurine was the dominant species at pH 1-3. This Bunte salt showed a facile S-S bond cleavage after a single S-oxygenation step on the inner sulfur atom. The sulfoxide is quite stable but there was no experimental evidence for the existence of the sulfone-sulfonic acid. Sulfate production was almost quantitative for the oxidation of only one of the sulfur atoms. Further reaction of the taurine occurred only on the nitrogen atom with no cleavage of the C-S bond. A 21-reaction kinetics scheme model gave reasonable agreement with experiment.

Full text of DOI:10.1039/b106167b

View Article Online / Journal Homepage / Table of Contents for this issue
Oxyhalogen–sulfur chemistry: Kinetics and mechanism of the
oxidation of a Bunte salt 2-aminoethanethiolsulfuric acid by chlorite¤
Cordelia R. Chinake,” Claudius Mundoma,° Rotimi Olojo, Tabitha Chigwada and
Reuben H. Simoyi*
Center for Nonlinear Science and the Department of Chemistry, W est V irginia University,
Morgantown, W V 26506-6045, USA
Received 11th July 2001, Accepted 22nd August 2001
First published as an Advance Article on the web 29th October 2001
Spectrophotometric and 1H NMR methods have been used to study the kinetics and mechanism of oxidation
of the Bunte salt, 2-aminoethanethiol sulfuric acid, H NCH CH S-SO H (AETSA) by chlorite in mildly acidic
2
2
2
3
media. The reaction is characterized by a long quiescent induction period followed by rapid and autocatalytic
production of chlorine dioxide. The formation of chlorine dioxide is much more pronounced in stoichiometric
excess of chlorite. The stoichiometry of the reaction in excess chlorite just before formation of chlorine dioxide
was determined to be: 2ClO ~ ] H NCH CH SÈSO H ] ClNHCH CH SO H ] SO ~2 ] Cl~ ] H`,
2
2
2
2
3
2
2
3
4
while in excess AETSA the stoichiometry was: 3ClO ~ ] 2H NCH CH SÈSO H ] 2H O ]
2
2
2
2
3
2
2NH CH CH SO H ] 2SO ~2 ] 3Cl~ ] 4H`. Although the products in excess chlorite also included pure
2
2
2
3
4
taurine and dichlorotaurine, monochlorotaurine was the dominant species at pH 1È3. This Bunte salt showed
a facile SÈS bond cleavage after a single S-oxygenation step on the inner sulfur atom. The sulfoxide is quite
stable but there was no experimental evidence for the existence of the sulfone-sulfonic acid. Sulfate production
was almost quantitative for the oxidation of only one of the sulfur atoms. Further reaction of the taurine
occurred only on the nitrogen atom with no cleavage of the CÈS bond. A 21-reaction kinetics scheme model
gave reasonable agreement with experiment.
would mean the deactivation of the standard reactive oxygen
Introduction
species before they destroy DNA molecules. One distinct
problem with this assertion is the inertness of taurine. It is so
inert that it is considered to be the metabolic end-product of
most thiols, especially cysteine and cysteamine.7 We have
studied the reactivity of taurine towards the oxidizing agents
chlorite and bromate.8 In both cases, taurine was either unre-
active or very slow to react. It deÐnitely did not appear to be
a very good candidate to act as an antioxidant. One line of
thought is that its rate of reaction is not important since it
exists in high mM concentrations in the body. It can be more
e†ective on the basis of pure mass action kinetics by its shear
quantity over all other reagents in the body.
We felt, however, that since taurine is a metabolic sink for
sulfur-bearing amino acids, that a more e†ective way of study-
ing its e†ect is to study its metabolic precursors. One could
then surmise that the abundance of taurine can be directly
correlated to the viability of the physiological reactions
forming it. The Ðrst metabolic precursor we studied was hypo-
taurine.9 Hypotaurine is oxidized easily and rapidly to
produce taurine. Taurine is not further oxidized at the sulfur
center, and in the presence of halogens and oxyhalogen com-
pounds, the nitrogen center can be oxidized further to give
nitrogenÈhalogen derivatives of taurine.
The next taurine precursor we have studied is a Bunte salt,
2-aminoethanethiol sulfuric acid (AETSA). AETSA is easily
produced in nearly 100% yields from the reaction of bisulÐte
with the oxidized form of cysteamine, cystamine.10 AETSA
was a very interesting precursor to us because it is widely used
as an antiulcerogenic. The use of AETSA reduced the ulcer-
ogenic e†ects of phenylbutazone or oxyphenylbutazone when
they are used as antiinÑammatory agents.11 This is a thera-
peutic e†ect that has been linked with taurine itself. Indus-
trially, ferrous complexes of SH-containing ligands are widely
Taurine is the most abundant sulfur-containing amino acid in
the human body. It often exists in greater concentration than
all the other amino acids combined.1,2 By virtue of its abun-
dance, it is surmised that it should have numerous physiologi-
cal roles.3 Taurine is highly concentrated in neurologically
excitable tissues such as the brain and retinal membranes. For
most physiologically-active compounds, it has been easy to
rationalize their roles mechanistically. With taurine, this has
not been the case. Many researchers have implicated taurine
in many physiological roles without adequate mechanistic
basis for these assertions. In general, the study of taurine is
based on a cause and e†ect: any clinical state brought about
by a deÐciency in taurine gives taurine a role in that state.
Examples include the established role of taurine in
osmoregulation in marine invertebrates, inhibition of nerve
impulses in invertebrate and vertebrate nerve tissue, regula-
tion of excitability in tissues, and elimination of cholestrol in
marine invertebrates.4 It is also reported that taurine can kill,
in vitro, Shistosoma mansoni. It is also widely accepted that
taurine can inhibit calcium-activated respiration in small
mammalian liver mitochondria.5
In the presence of such highly-reactive antioxidants such as
vitamin C and several thiols, taurine has been tabbed as one
of the most important antioxidants in the human body.6 This
¤ Part 4 (of 10) in the series of publications dedicated to the memory
of Dr Cordelia R. Chinake, 1965È1998. Part 3: Int. J. Chem. Kinet.,
submitted.
” Deceased.
° Present address: Department of Chemistry, Florida State Uni-
versity, Tallahassee, FL 32306-4390, USA. CM is a United Negro
College Fund PÐzer Postdoctoral Fellow.
DOI: 10.1039/b106167b
Phys. Chem. Chem. Phys., 2001, 3, 4957È4964
This journal is ( The Owner Societies 2001
4957

View Article Online
used for the removal of NO from Ñue gases that contain
In excess AETSA, the stoichiometry is deduced to be:
3ClO ~ ] 2H NCH CH SÈSO H ] 2H O ]
SO . In those NO-strippers that used cysteamine,
2
H NCH CH SH, the cysteamine was converted to the Bunte
2
2
2
2
3
2
2
2
2
salt, AETSA. Hydrolysis of AETSA, however, regenerated
2NH CH CH SO H ] 2SO 2~ ] 3Cl~ ] 4H` (R3)
2
2
2
3
4
cysteamine:12
Qualitative tests for SO 2~ and Cl~ were positive. Stoichiom-
4
H NCH CH SÈSO H ] H O ]
etry (R2) was especially correct for reactions in mildly acidic
2
2
2
3
2
media. The stoichiometry obtained was usually between that
for (R2) and (R3). NMR data showed evidence for the forma-
tion of dichlorotaurine which could be formed by the dispro-
portionation of monochlorotaurine:17
H` ] H NCH CH SH ] HSO ~ (R1)
2
2
2
4
As a result of this important industrial process, several
researchers have worked on the hydrolysis of AETSA.
In this paper we report on a series of experiments on the
oxidation mechanism of AETSA by the oxyhalogen ion, chlo-
rite. The cysteamine formed in reaction (R1) can, in the pres-
ence of an oxidizing agent, produce either hypotaurine or
taurine itself. Most of our previous work on the oxidation of
an organic sulfur compound showed successive oxygen atom
addition on the sulfur center followed by cleavage of the CÈS
bond to give sulfate and organic residue.13 The use of even the
most powerful oxidizing agents on taurine does not lead to
cleavage of the CÈS bond.
2ClNHCH CH SO H H
2
2
3
NH CH CH SO H ] Cl NCH CH SO H (R4)
2
2
2
3
2
2
2
3
This equilibrium (R4) was shifted to the right in highly acidic
environments due to the protonation of the amino group in
taurine:
NH CH CH SO H ] H` H [H NCH CH SO H]`
2
2
2
3
3
2
2
3
The major product from the initial oxidation of AETSA,
however, was identiÐed, using UV and NMR spectroscopy, to
be chlorotaurine, ClNHCH CH SO H. The NMR spectrum
2
2
3
Experimental
was obtained after incubating reaction mixtures for periods of
36 h. Fig. 1 shows a series of NMR spectra used to determine
the nature of the products from the reaction. Spectrum (a)
shows the spectrum of AETSA in acidic solution. It is a multi-
plet comprising two overlapping triplets at ca. d \ 3.26 which
is characteristic of the coupling of the adjacent methylene
protons. Spectrum (b) shows the spectrum after addition of
chlorite but before completion of the reaction. This shows the
emergence of two triplets downÐeld of the previous multiplet.
These triplets can only be formed from a sulfoxide in which
the inner sulfur atom picks up an oxygen atom:
Materials
The following analytical grade chemicals were used without
further puriÐcation: perchloric acid, 72%, sodium chloride
(Fisher) and D O (Aldrich). Stock solutions were prepared
2
using doubly-distilled water. 2-aminoethanethiolsulfuric acid
(Aldrich) was recrystallized from water before use. Sodium
chlorite (Aldrich) was recrystallized from its technical grade,
which is ca. 80% pure. Recrystallization was carried out from
a waterÈethanol mixture to bring the assay value to 98%. The
recrystallized chlorite was standardized iodometrically by
adding acidiÐed potassium iodide and titrating the liberated
iodine against sodium thiosulfate using freshly prepared
H NCH CH SÈSO H ] [O] % H NCH CH S(O)SO H
2
2
2
3
2
2
2
3
The sulfoxide renders the inner sulfur center asymmetric,18
with the adjacent hydrogen atoms diastereotopic, thus indu-
cing the two new set of triplets downÐeld:
starch as indicator.14 ClO was prepared by oxidizing pot-
2
assium chlorate in a sulfuric acidÈoxalic acid mixture and
then stored in perchloric acid at 4 ¡C.15 Standardization was
carried out spectrophotometrically by using its absorption
coefficient 1265 M~1 cm~1 at 360 nm.16
In spectrum (c) the doublet peaks grow, showing the stabil-
ity of the sulfoxide as well as its increase in concentration. At
Methods
Stoichiometric determinations were carried out by mixing
various ratios of chlorite and AETSA in stoppered volumetric
Ñasks and scanning them spectrophotometrically for forma-
tion of ClO over a period of 2 days. 1H NMR spectra of
2
AETSA and its oxidation products were obtained on a JEOL
GX 270 spectrometer using D O as solvent. Kinetic experi-
2
ments were carried out at 25.0 ^ 0.5 ¡C and the ionic strength
was maintained at 1.0 M using NaCl. Traces were obtained by
following the appearance of ClO at 360 nm. Since the reac-
2
tions are sufficiently slow, a conventional Perkin Elmer
Lambda 2S UVÈVis spectrophotometer was used to follow
the reaction. Some of the chlorine dioxide experiments were
carried out on a Hi-Tech ScientiÐc double-mixing stopped-
Ñow spectrophotometer. Qualitative analysis of the sulfate
was by precipitation as BaSO .
4
Results
Stoichiometry
Fig. 1 (a) NMR spectrum of AETSA in acid. (b) Spectrum of reac-
The stoichiometry of the reaction of AETSA with excess chlo-
rite in perchloric acid was determined to be:
tion products of AETSAÈClO ~ reaction in acid showing appearance
2
of new peaks after 5 min reaction time. (c) NMR spectrum of
AETSAÈClO ~ taken after 24 h under condition of excess AETSA. (d)
2ClO ~ ] H NCH CH SÈSO H ]
2
2
2
2
2
3
Spectrum of same reaction under conditions of excess ClO ~ after a
2
ClNHCH CH SO H ] SO 2~ ] Cl~ ] H` (R2)
24 h incubation period. (e) NMR spectrum of pure taurine.
2
2
3
4
4958
Phys. Chem. Chem. Phys., 2001, 3, 4957È4964

View Article Online
Fig. 2 E†ect of chlorite on the absorbance at 360 nm. Concentration
of chlorite is directly proportional to the amount of ClO formed and
inversely proportional to the induction period at Ðxed AETSA con-
2
Fig. 4 Absorbance traces at 360 nm showing e†ect of acid on the
oxidation of AETSA by chlorite. Induction time for the reaction is
inversely dependent on [H`] . [AETSA] \ 1.0 ] 10~3 M,
centration: [AETSA] \ 0.001 M, [H`] \ 0.05 M, I
\ 1.0 M,
o
o
(NaCl)
[ClO ~] \ (a) 7.0 ] 10~3, (b) 6.0 ] 10~3, (c) 5.0 ] 10~3 and (d)
o
o
2
[ClO ~] \ 4.0 ] 10~3 M, I
\ 1.0 M, [H`] \ (a) 0.08, (b) 0.06,
4.0 ] 10~3 M.
2
(NaCl)
o
(c) 0.05, (d) 0.04 and (e) 0.025 M.
this point, the former AETSA multiplet bifurcates into the
characteristic triplet peaks of taurine. Upon oxidation by
excess chlorite, the multiplex resolved into two broad peaks at
d \ 3.21È3.28 and d \ 3.37È3.45, associated with a series of
compounds which all contain the two sets of methylene
protons. These triplets are not as resolved as those for pure
taurine shown in spectrum (e). This apparent lack of
resolution is due to the mixing of the three taurine-based pro-
ducts: taurine, chlorotaurine and dichlorotaurine. The substi-
tution of chlorine atoms for the hydrogen atoms at the
Reaction dynamics
The reaction could only be followed by NMR and UVÈVis
absorbance measurements. SpeciÐc electrodes for chloride and
sulfate were quickly poisoned by the sulfur generated from the
AETSA and deposited on the electrode surfaces. The high
ionic strengths used precluded the use of conductivity mea-
surements. In conditions of excess chlorite, the reaction could
be followed by observing the formation of chlorine dioxide.
The chlorine dioxide is formed after some induction period
that is determined by the initial concentrations of the
reagents, especially the ratio of reductant (AETSA) to oxidant.
The formation of chlorine dioxide is very rapid and autocat-
alytic after the expiration of the induction period. Fig. 2 shows
that at Ðxed AETSA concentrations, increasing chlorite con-
centrations decreased the induction period and increased the
rate of formation of chlorine dioxide. The sharp rate of forma-
tion of chlorine dioxide after the induction period is very sig-
niÐcant and allows us to separate the reaction into two
sections: an initial section that involves the consumption of
AETSA, and a second, which involves an extraneous oxyhalo-
gen reaction that forms chlorine dioxide. It would appear,
from the absorbance data, that the rates of these two sections
are so di†erent in magnitude that they can be treated as two
separate processes. Thus the induction period observed would
be related to the rate of reaction that consumes AETSA in an
inverse-type of dependence; i.e. a shorter induction period
would suggest that AETSA had been rapidly consumed. Fig. 3
shows the plot of the induction time vs. the inverse concentra-
tion of chlorite. The fact that this plot gives a straight line
implies that the reaction that depletes AETSA is Ðrst order in
chlorite.8 Fig. 3 was also important in conÐrming our stoichi-
ometry. Whenever there is excess chlorite over that needed to
just attain stoichiometry (R2), then the excess chlorite after
a
b
nitrogen center introduces
a more electron-withdrawing
group, but the e†ect is not as strong since the chlorine atom
attaches to the nitrogen center as a positive cation. The
protons attached to nitrogen and oxygen atoms, however,
cannot be detected due to rapid exchange with solvent.
consumption of AETSA will react to form ClO , and one can
2
measure a Ðnite induction period. If however, there is less
chlorite to satisfy stoichiometry (R2), then there will be no
formation of ClO and the induction period will be inÐnity.
2
However, the x-intercept of the graph gives the inverse of the
Fig. 3 Plot of induction time vs. 1/[ClO ~] using stoichiometric
2
amount of chlorite just needed to satisfy stoichiometry (R2).
In this reaction the intercept concentration for 0.001 M
AETSA was 0.0022 M ClO ~, thus conÐrming stoichiometry
excess of chlorite. [AETSA] \ 0.001 M, [H`] \ 0.05 M, I
\
o
o
(NaCl)
1.0 M, [ClO ~] \ (a) 7.0 ] 10~3, (b) 6.0 ] 10~3, (c) 5.0 ] 10~3 and
2
(d) 4.0 ] 10~3 M.
2
Phys. Chem. Chem. Phys., 2001, 3, 4957È4964
4959

View Article Online
Fig. 7 Absorbance traces of AETSA oxidation by chlorine dioxide,
ClO in acidic solution. [ClO ] \ 6.61 ] 10~4 M; [H`] \ 0.10 M;
Fig. 5 Induction time vs. the inverse of the initial acid concentra-
2
2
I
\ 1.0 M [AETSA] \ (a) 3.0 ] 10~4, (b) 4.0 ] 10~4, (c)
tions [H`] . The induction period is more sensitive to acid at lower
(NaCl)
o
5.0 ] 10~4, (d) 6.0 ] 10~4 and (e) 7.5 ] 10~4 M.
concentrations. [AETSA] \ 1.0 ] 10~3, [ClO ~] \ 4.0 ] 10~3 M,
o
2
I
\ 1.0 M, [H`] \ (a) 0.08, (b) 0.06, (c) 0.05, (d) 0.04 and (e)
(NaCl)
o
0.025 M.
the induction time does not change but the rate of formation
of chlorine dioxide increases with higher initial AETSA con-
centrations.
(R2). The stoichiometry was not exactly 2 : 1 because of the
expected formation of dichlorotaurine. However, the y-
intercept does not have any stoichiometric signiÐcance.
The e†ect of acid was more complex than that of chlorite
(Fig. 4). It appears the e†ect of acid is more pronounced at
low acid than at high acid concentrations. Linearity in the
plot of inverse acid concentration vs. induction period is lost
as the acid concentrations increased (Fig. 5). Fig. 4 shows that
although acid shortens the induction period, the rate of pro-
duction of chlorine dioxide is retarded by high acid concentra-
tions.
The direct reaction of chlorine dioxide with AETSA deter-
mines the induction time for the reaction. If this reaction is
slow, we would expect oligo-oscillatory behavior19 in the con-
centration of chlorine dioxide. If, however, the reaction is
faster than the rate of formation of chlorine dioxide, then a
sharp induction period will be observed. Fig. 7 shows a very
complex reaction of chlorine dioxide with AETSA. In excess
AETSA, there is a rapid and smooth consumption of chlorine
dioxide as can be seen in traces (c)È(e). In conditions where
[ClO ] ^ [AETSA] we observe an initial depletion of chlo-
Experiments performed with [ClO ~] P [AETSA]
2
0
0
2 0
0
showed that higher AETSA concentrations increased the
induction period. This is to be expected if the total consump-
tion of AETSA is a prerequisite to the formation of chlorine
dioxide (Fig. 6). In conditions where [ClO ~] A [AETSA] ,
rine dioxide followed by re-formation of chlorine dioxide.
Mechanism
2
0
0
From the data shown in Fig. 7, it would appear that the total
oxidation of AETSA is a prerequisite for the formation of
chlorine dioxide. NMR data suggest a step-wise oxidation
that passes through the sulfoxide.
H NÈCH ÈCH ÈSÈSO H ] H O ]
2
2
2
3
2
H NÈCH ÈCH ÈS(O)ÈSO H ] 2H` ] 2e~ (R5)
2
2
2
3
The oxidant could be ClO ~ or any of its decomposition pro-
ducts: HOCl and Cl . The acid dependence data are so
complex that they have to be addressed Ðrst. Chlorous acid
2
2
has a pK of about 2.020 and so will exist predominantly as
a
the protonated form at pH \ 2:
ClO ~ ] H` H HClO ;
K
(R6)
2
2
a1
The AETSA should also be protonated, but this will be pre-
dominantly on the amino group:
H NÈCH ÈCH ÈSÈSO H ] H` H
2
2
2
3
[H NÈCH ÈCH ÈSÈSO H]` (AETSAÈH`) (R7)
3
2
2
3
We can assume complete protonation of the AETSA in the
conditions utilized in our experiments, and wherever we indi-
cate the structure of AETSA, it will be in the protonated form.
There, will, however, be a mixture of HClO and ClO ~ in
Fig. 6 E†ect of varying [AETSA] on formation of chlorine dioxide
o
at Ðxed chlorite concentrations. Traces show that AETSA is com-
pletely depleted before the formation of ClO commences:
2
[ClO ~] \ 3.8 ] 10~3 M,
I
\ 1.0 M, [H`] \ 0.30 M,
2
2
2
(NaCl)
o
solution, with both species reacting with AETSA, but at di†er-
ent rates.
[AETSA] \ (a) 2.0 ] 10~3, (b) 1.5 ] 10~3, (c) 1.0 ] 10~3 and (d)
o
5.0 ] 10~4 M.
4960
Phys. Chem. Chem. Phys., 2001, 3, 4957È4964

View Article Online
Consumption of AETSA
H NÈCH ÈCH ÈSÈSO H ] ClO ~ ] H` ]
2
2
2
3
2
AETSA has a weak absorption peak at 241 nm, but, fortu-
nately, this peak is isolated (see Fig. 8(a)). We can use this
peak and our basic knowledge of oxyhalogen chemistry
together with the data on induction periods to evaluate the
mechanistic details of consumption of AETSA by chlorite.
Fig. 8(b) shows the complex dependence of the consumption
of AETSA on acid concentration. Although acid seems to
catalyze the reaction, this e†ect quickly tails o† as the acid
concentration is increased to below pH 2.0 (see Fig. 8(c)).
Reactions (R8) and (R9) represent the initial steps for the oxi-
dation of AETSA. The HOCl formed in reaction (R8) and (R9)
H NÈCH ÈCH ÈS(O)ÈSO H ] HOCl;
k
k
(R8)
(R9)
2
2
2
3
1
H NÈCH ÈCH ÈSÈSO H ] HClO ]
2
2
2
3
2
H NÈCH ÈCH ÈS(O)ÈSO H ] HOCl;
2
2
2
3
2
It would be safe to assume that reactions (R8) and (R9) are
rate-determining and the rate of consumption of AETSA can
be written as:
Rate \ k [AETSA][ClO ~][H`] ] k [AETSA][HClO ] (1)
1
2
2
2
Reaction (R8) is not actually a termolecular process. It
involves the initial attack of chlorite on AETSA followed by a
very rapid protolytic process that abstracts a proton to give
electroneutrality. The values of k and k will determine the
reacts very rapidly with any remaining ClO ~ in solution to
2
produce ClO :21
2
1
2
e†ect of acid. Concentrations of ClO ~ and HClO are
2ClO ~ ] HOCl ] H` ] 2ClO ] Cl~ ] H O (R10)
2
2
2
2
2
related by the following mass balance equation that tracks all
chlorine(III) species:
If the reaction of ClO with AETSA is slow, then there will
be an immediate accumulation of yellow ClO at the begin-
ning of the reaction, and there will be no induction period
observed. Since the reaction displays an induction period, it
would appear that the ClO formed in (R10) is immediately
mopped up by AETSA (and its derivatives that have not been
oxidatively saturated). HOCl can also oxidize both AETSA
and its derivatives.
2
[Cl(III)] \ [ClO ] ] [HClO ]
(2)
2
T
2
2
The equilibrium in reaction (R6) allows us to write the follow-
ing relation also:
2
[HClO ] \ k~1[ClO ~][H`]
(3)
We can use eqn. (2) and (3) to write [HClO ] and [ClO ~] in
2
a1
2
2
2
terms of the total chlorine(III) species:
HOCl ] H NÈCH ÈCH ÈSÈSO H ]
[Cl(III)]
2
2
2
3
T
[ClO ~] \
(4)
2
1 ] K~1[H`]
H NÈCH ÈCH ÈS(O)ÈSO H ] Cl~ (fast) (R11)
a1
2
2
2
3
and
HOCl ] H NÈCH ÈCH ÈS(O)ÈSO H ]
2
2
2
3
K~1[H`][Cl(III)]
a1
T
[HClO ~] \
(5)
H NÈCH ÈCH ÈS(O )ÈSO H ] Cl~ ] H` (R12)
2
1 ] K~1[H`]
2
2
2
2
3
a1
It is highly unlikely that the sulfone formed in reaction
(R12) is stable, and it should be easily oxidized to sulfate and
taurine.
We can use eqn. (4) and (5) to write eqn. (1) in a form that
involves only measurable quantities:
Rate \ [AETSA][Cl(III)] [H`]
T
HOCl ] H NÈCH ÈCH ÈS(O )ÈSO H ] H O ]
G
k
k K~1
a1
H
2
2
2
2
3
2
1
2
]
]
(6)
1 ] K~1[H`] 1 ] K~1[H`]
H NÈCH ÈCH ÈSO H ] SO 2~ ] Cl~ ] H` (R13)
a1
a1
2
2
2
3
4
In the limit of very low acid concentrations, 1 ] K~1[H`] B
1 and eqn. (6) simpliÐes to:
a1
Though ClO ~ can also perform the oxidations listed in
2
reactions (R11)È(R13), the bulk of these will be borne by
Rate \ [AETSA][Cl(III)] [H`]Mk ] k K~1N
(7)
HOCl. This is due to the rapid and autocatalytic formation of
T
1
2 a1
HOCl from ClO ~. These autocatalytic steps involve the well-
Thus the data will show a Ðrst order dependence with acid at
low acid concentrations only. As the acid concentration is
increased, the terms in the denominator in eqn. (6) become
viable and the reaction will show saturation. This is what was
observed in our experiments.
Eqn. (7) also shows that the reaction will display Ðrst order
kinetics with respect to chlorite, irrespective of the strength of
the acid. The same applies to AETSA dependences.
2
known asymmetric intermediate Cl O :22
2
2
ClO ~ ] HOCl ] H` ] Cl O ] H O
(R14)
(R15)
2
2
2
2
Cl O ] 2H` ] Red ] 2HOCl ] Red2`
2
2
The “RedÏ represents any 2-electron reductant in solution:
AETSA, the sulfoxide and the sulfone. Combining reactions
Fig. 8 (a) UV absorption spectrum of AETSA showing a peak at 241 nm. There was no interference from the oxyhalogen species at this
wavelength. (b) Absorbance traces showing consumption of AETSA by chlorite at varying [H`]. [ClO ~] \ 0.005 M; [AETSA] \ 0.005 M,
2
I
\ 1.0 M and [H`] \ (a) 0.20, (b) 0.10, (c) 0.05 and (d) 0.025 M. (c) Plot of initial rate vs. [H`] showing a hint of saturation kinetics as acid
(NaCl)
concentrations are increased.
Phys. Chem. Chem. Phys., 2001, 3, 4957È4964
4961

View Article Online
(R14) and (R15) shows the (quadratic) autocatalytic formation
of HOCl:
either HOCl or Cl :
2
2Cl(III) ] Cl(I) ] 2Cl(IV) ] Cl(ÈI)
(R21)
(R22)
ClO ~ ] HOCl ] 3H` ] Red ] 2HOCl ] H O ] Red2`
and
2
2
(R16)
Cl(III) ] Cl(0) ] Cl(IV) ] Cl(ÈI)
With such a rapid formation of HOCl, all oxidations by
chlorite will be insigniÐcant. Since the rate of formation of
HOCl is rate-determining, then reactions (R8) and (R9) will
still be rate-determining. The induction period will end when
It is reasonable to assume that reaction (R10) (in the form of
(R21), above), would be the dominant route for the production
of ClO . Reaction (R10) is a composite that involves the
2
initial reaction of ClO ~ with HOCl (as in reaction (R14)).
2
reaction (R10) produces more ClO than can be consumed by
The protonated form of ClO ~; chlorous acid, HClO , is a
poorer nucleophile than the unprotonated form (HOCl is an
electrophile), and hence at high acid concentration, reaction
(R10) would become marginal in favor of the slower reaction
2
2
2
AETSA, the sulfoxide and the sulfone.
Reaction of chlorine dioxide with AETSA
of HOCl with HClO :
2
Chlorine dioxide is an odd-electron species, and in most of its
oxidations, the initial step involves the addition of an electron
to produce chlorite:
2HClO ] HOCl ] H` ] 2ClO ] Cl~ ] H` ] H O
2
2
2
(R10b)
or simply there would be less free ClO ~ available to e†ect
ClO ] e~ ] ClO ~
(R17)
2
2
2
reaction R10.
The chlorite then carries out most of the oxidations. Reaction
(R17), however, will not be fast. Further reduction of ClO ~
Quantitative formation of ClO . By incubating reaction
2
will produce HOCl which will give back ClO according to
2
2
solutions for extended periods of time (7 days or more), one
reaction (R10). This will not be very helpful with respect to the
notices that there is a slower rate of formation of chlorine
dioxide that continues way after the AETSA has been con-
sumed. Since we utilized highly acidic media, we expect dis-
proportionation of chlorite to chlorine dioxide to predominate
over the other pathway that gives chlorate (in basic
environments). Both stoichiometries (R2) and (R3) give Cl~
ions as products. In the presence of acid, the following slow
reaction will occur:24
reduction of AETSA, and one would expect a very slow reac-
tion which is basically controlled by the free radical kinetics of
reaction (R17). Fortunately, HOCl reaction with AETSA is
fast enough to compete favorably with (R10) to produce Cl~
(also produced in reactions (R11)È(R13) which can directly
attack ClO in acidic environments to produce more reactive
2
species ClO ~ and HOCl:23
2
ClO ~ ] Cl~ ] 2H` H 2HOCl
(R23)
2ClO ] Cl~ ] H O ] 2ClO ~ ] HOCl ] H` (R18)
2
2
2
2
The production of HOCl will immediately trigger reaction
(R10), forming chlorine dioxide. Since reaction (R18) is slow
on the timescale of the oxidation of AETSA, it will not inter-
fere with the kinetics of the reaction we are studying. It will,
(R18) is not a single step process, and would involve the initial
formation of some intermediate powerful oxidizing species
such as Cl O .
2
2
Cl O ~ ] ClO H ClO ~ ] Cl O
(R19)
(R14)
however, be responsible for the continued increase in ClO
concentration long after the oxidation of AETSA is complete.
2
2
2
2
2 2
2
Cl O ] H O H ClO ~ ] HOCl ] H`
By eliminating HOCl, one can add (R23) ] 2(R18) to obtain a
2
2
2
2
well-known reaction for the decomposition of ClO ~ in acidic
The formation of these powerful oxidizing species would then
initiate the oxidation, thereby by-passing bottleneck estab-
lished by reaction (R17). (R18) is the reverse of (R10). Its direc-
tion is determined by the HOCl concentration. At high HOCl
concentration reaction (R10) will dominate.
The data in Fig. 7 show what appears to be a violation of
the second law of thermodynamics by the delivery of a non-
monotonic decrease in the position of the lowest free energy
with respect to chlorine dioxide concentration (see traces (a)
2
media:21
5ClO ~ ] 4H` ] 4ClO ] 2H O ] Cl~
(R24)
2
2
2
Ultimately, the Ðnal chlorine dioxide formed depends on the
excess chlorite concentrations according to reaction stoichi-
ometry (R24). In mildly acidic environments, stoichiometry
(R24) can take several days to establish.
Computer simulations
and (b)). The observed “cuspÏ in the decay of ClO can only be
2
produced by a combination of two or more nonlinear kinetics
The full mechanism, which combines the chloriteÈAETSA
reaction and the chlorine dioxideÈAETSA reaction, was com-
piled into the 21 reactions shown in Table 1. There are 12
oxychlorineÈsulfur reactions in which the chlorine center is
reduced, 4 pure oxyhalogen reactions, 3 sulfurÈsulfur dispro-
portionation reactions and 2 acidÈbase equilibria. To model
this reaction network the stochastic algorithm contained in
the chemical kinetics simulator (CKS) software package devel-
oped by IBMÏs Almaden Research Center was used. This algo-
rithm was as accurate as the semi-implicit fourth order
RungeÈKutta scheme developed by Kaps and Rentrop. The
CKS package, however, required fewer computer resources.
Reactions (M1) and (M21) are rapid protolytic reactions
whose kinetics parameters were inconsequential for the simu-
lations as long as they were fast (and not rate-determining).
The ratios of the forward and reverse reactions were chosen
pathways. These could be in the form of autocatalysis or
autoinhibition. Quadratic autocatalysis is displayed in the
production of HOCl, and this fuels its rapid increase in which
it is not all consumed by the time the reducing substrate,
AETSA has been completely consumed. From the above reac-
tion network, it is not difficult to explain the observed increase
in ClO at the end of the reaction in Fig. 7. The system can
2
proceed according to (R18) until all the AETSA has been con-
sumed, leaving behind ClO ~, Cl~ and some HOCl. These
2
will recombine in the absence of reductant to re-form ClO .
2
This reaction is necessary for stoichiometric consistency.
Formation of ClO
2
Most rates of oxyhalogen reactions are enhanced by acid. In
this particular reaction system we observe a retardation of
ClO formation with acid (see Fig. 4). ClO is formed by the
such that they maintained the standard pK s of the respective
a
weak acids. The parameters for the oxychlorine reactions,
2
2
oxidation of chlorite. This oxidation can be carried out by
(M11), (M12), (M17) and (M18) were derived from a com-
4962 Phys. Chem. Chem. Phys., 2001, 3, 4957È4964

View Article Online
Table 1 Full mechanism for the chlorite-AETSA reaction. Rate constants used are shown after the semi-colon. For a reversible reaction, the
reverse rate constant is shown after a comma. Except where water is involved, the order of the reaction is derived from the molecularity
No.
Reactantsa
Productsa
M1
M2
M3
M4
M5
M6
M7
M8
ClO ~ ] H`
H
]
]
]
]
]
]
]
]
H
H
H
]
]
]
]
H
H
H
H
H
HClO ; 1 ] 109, 1.91 ] 107
2
2
ClO ~ ] RSÈSO H
RS(O)SO H ] ~OCl; 5.80
2
3
3
HClO ] RSÈSO H
RS(O)SO H ] HOCl; 3.20
2
3
3
ClO ] RS(O)SO H
2~
RS(O )SO H ] ~OCl; 45
3
2
3
ClO ] RS(O )SO H ] H O
RSO H ] SO 2~ ] H` ] HOCl; 5 ] 103
2~
2
3
2
3
4
HOCl ] RSÈSO H
RS(O)SO H ] Cl~ ] H`; 43.8
3
3
HOCl ] RS(O)SO H
RS(O )SO H ] Cl~ ] H`; 5 ] 102
3
2
3
HOCl ] RS(O )SO H ] H O
RSO H ] SO 2~ ] Cl~ ] 3H`; 1 ] 103
2
HOCl ] RSO H
3
3
2
3
4
M9
ClHNCH CH SO H ] H O; 5 ] 106
2
2
3
2
M10
M11
M12
M13
M14
M15
M16
M17
M18
M19
M20
M21
2ClHNCH CH SO H
RSO H ] Cl NCH CH SO H; 5 ] 107, 1 ] 106
2
2
3
3
2
2
2
3
ClO ~ ] HOCl ] H`
Cl O ] H O; 1.01 ] 106, 1 ] 10~1
2
2
2
2
Cl O ] ClO ~
2ClO ] Cl~; 1.5 ] 103, 5.5 ] 106
2
2
2
2
Cl O ] RSÈSO H ] H O
RS(O)SO H ] 2HOCl; 15.5
2
2
3
2
3
Cl O ] RS(O)SO H ] H O
RS(O )SO H ] 2HOCl; 17.5
2
2
3
2
2
3
Cl O ] RS(O )SO H ] 2H O
RSO H ] SO 2~ ] 2HOCl ] 2H`; 5 ] 103
2
2
2
3
2
3
4
2ClO ] RSÈSO H ] H O
2ClO ~ ] RS(O)SO H ] 2H`; 5 ] 106
2
3
2
2
3
ClO ] Cl~
Cl O ~; 4 ] 106, 5 ] 105
2
2
2
Cl O ~ ] ClO
ClO ~ ] Cl O ; 5 ] 103, 5 ] 102
2
2
2
2
2 2
2RS(O)SO H
RSÈSO H ] RS(O )SO H; 1 ] 106, 3 ] 107
3
3
2
3
RS(O)SO H ] RS(O )SO H ] H O
RSÈSO H ] RSO H ] SO 2~ ] 2H`; 50, 0.01
3
2
3
2
3
3
4
~OCl ] H`
HOCl; 1 ] 109, 5 ] 105
a R \ H NCH CH È, RSO H \ taurine.
2
2
2
3
bination of literature values and our own estimates.25h27 It
turned out that the most important parameter in determining
the clock-nature of the overall reaction was the reverse rate
initial guess value for each process. For example, NMR data
showed the evolution of a stable sulfoxide, and thus the
further oxidation of the sulfoxide to the sulfone was assumed
to be relatively slow compared to the further oxidation of the
sulfone. Reaction (M10) was, in most cases, inconsequential in
the oxidation of AETSA, but was included in very highly
acidic environments due to the formation of the protonated
taurine molecule that tended to pull the equilibrium of (M10)
to the right. Reactions (M19) and (M20) are dispro-
portionation reactions whose sole role is to ensure stoichio-
metric consistency after complete consumption of either the
oxidant or AETSA. Thus it was not at all important what rate
constants were given to these reactions as long as they were
fast enough to establish equilibrium within the time frame of
the chloriteÈAETSA reaction. High values of the kinetics
parameters made the simulations sti† and slow, while too low
values introduced Ñuctuations in the AETSA and oxychlorine
species which we had not observed experimentally. All reac-
tions involving oxidation by the intermediate, Cl O were
constant for (M12), k
. Several authors have worked on the
m12
kinetics of reaction (M11).24 Reaction (M11) was not treated
computationally as a termolecular reaction. The acid concen-
tration was high enough for us to assume bu†ering, and so the
rate constant k \ 1.01 ] 106 M~2 s~1 reported in the liter-
11
ature was replaced by a pseudo-rate constant, k@ \ k
11
11
[H`] where [H`] was assumed constant throughout the
reaction lifetime.
0
0
The oxychlorineÈsulfur reactions were derived from a pure
permutation of 3 oxidants (ClO ~, HOCl and Cl O ) with
2
2 2
four reductants (AETSA, the sulfoxide, the sulfone and
taurine). Experimental data then aided us in determining an
2
2
deliberately slowed so that all the observed autocatalysis can
be derived from the production of HOCl (reactions (M12)È
(M15)). The autocatalysis was then clearly evident if HOCl
oxidations were faster than those by either ClO ~ or Cl O
2
2 2
itself.
As expected, the most important kinetics parameter in this
whole reaction scheme was k (forward rate constant for reac-
2
tion (M2)). The value chosen for our simulation; 5.8 M~2 s~1,
was chosen by estimating the initial rate of depletion of
AETSA shown in Fig. 8(a) and then working around that esti-
mate for the best Ðt. The selected value of k , after estab-
3
lishment of k , was useful only for determining the degree of
2
acid retardation in the formation of chlorine dioxide after the
induction period. Otherwise one could shut down (M3) and
rely on the equilibrium in reaction (M1) to handle the acid
retardation.
There was reasonable agreement between our model (Table
1) and the experimental data. The chlorine dioxideÈAETSA
reaction could also be modeled by the same mechanism by
assuming an initial concentration of chlorite of zero. Fig. 9
shows our Ðt to a typical trace with excess oxidant. Though
simulations of the chlorine dioxideÈAETSA reaction gave
Fig. 9 Results of modeling reaction mechanism shown in Table 1 for
the conditions [ClO ~] \ 0.004 M; [AETSA] \ 0.001 M;
good agreement for the consumption of ClO , the model,
2
2
0
0
however, gave a zero order-type of trace for the re-formation
[H`] \ 0.08 M. Data points generated by the model are shown as
0
square symbols.
of ClO (as opposed to the “cuspÏ observed in Fig. 7).
2
Phys. Chem. Chem. Phys., 2001, 3, 4957È4964
4963

View Article Online
7
L. Young and G. A. Maw, Metabolism of Sulfur Compounds,
Methuen, London, 1958.
Conclusion
8
9
C. R. Chinake and R. H. Simoyi, J. Phys. Chem., 1997, 101, 1207.
B. S. Martincigh, C. Mundoma and R. H. Simoyi, J. Phys. Chem.,
1996, 100, .
Our kinetics study has shown that AETSA is a good precur-
sor for taurine in acidic oxidizing environments, and that
some of its physiological e†ects might actually be derived
from taurine itself, (and its derivatives) which is a stable end-
product of the S-oxidation of AETSA. The SÈS bond is very
important in physiological mechanisms and its oxidation
mechanisms are a Ðrst step towards characterizing its physio-
logical e†ects.
10 H. Distler, Angew. Chem., Int. Ed. Engl., 1967, 6, 544.
11 A. W. Tyrrell, T etrahedron L ett., 1985, 26, 1753.
12 J. M. Anderson and N. E. Pawlowski, J. Am. Chem. Soc., 1966,
88, 5245.
13 R. H. Simoyi and I. R. Epstein, J. Phys. Chem., 1987, 91, 5124.
14 A. J. Indelli, J. Phys. Chem., 1964, 68, 3027.
15 Handbook of Preparative Organic Chemistry, ed. G. Brauer, Aca-
demic Press, New York, 1963, vol. 1, p. 301.
16 I. Lengyel, Gy. Rabai and I. R. Epstein, J. Am. Chem. Soc., 1990,
112, 9104.
Acknowledgements
This work was supported by grant number CHE 9632592
from the National Science Foundation. Some of the writing
up of this work was done while RHS was on sabbatical leave
at the University of Cape Town. RHS is grateful to the UCT
Chemistry Department for hosting him.
17 Y. Y. Lin, C. E. Wright, M. Zagorski and K. Nakanishi, Biochem.
Biophys. Acta, 1988, 969, 242.
18 C. Mundoma and R. H. Simoyi, J. Chem. Soc., Faraday T rans.,
1997, 93, 1543.
19 Gy. Rabai and M. T. Beck, J. Chem. Soc., Dalton T rans., 1985,
1669.
20 P. W. Atkins and J. A. Beran, General Chemistry, ScientiÐc Amer-
ican Books, New York, 2nd edn., 1990, p. 550.
21 Z. Jia, D. W. Margerum and J. S. Francisco, Inorg. Chem., 2000,
39, 2614.
22 H. Taube and H. Dodgen, J. Am. Chem. Soc., 1949, 71, 3330.
23 Gy. Rabai and M. Orban, J. Phys. Chem., 1993, 97, 5935.
24 G. Peintler, I. Nagypal and I. R. Epstein, J. Phys. Chem., 1990,
94, 2954.
25 J. N. Figlar and D. M. Stanbury, J. Phys. Chem. A, 1999, 103,
5732.
References
1
2
R. W. Chesney, Adv. Pediatr., 1985, 32, 1.
A. M. Pruski, A. Fiala-Medioni and J. C. Colomines, C. R. Acad.
Sci. III, 1997, 320, 791.
3
4
J. G. Jacobsen and L. H. Smith, Physiol. Rev., 1968, 48, 424.
S. T. Test, M. B. Lampert, P. J. Ossanna, J. G. Thoene and S.
Weiss, J. Clin. Invest., 1984, 74, 1341.
P. Dolara, P. Marino and F. Bu†oni, Biochem. Pharmacol., 1973,
22, 2085.
5
6
26 I. Lengyel, L. Gyorgyi and I. R. Epstein, J. Phys. Chem., 1995, 99,
C. E. Wright, T. T. Lin, Y. Y. Lin, J. A. Sturman and G. E. Gaull,
Prog. Clin. Biol. Res., 1985, 179, 137.
12804.
27 F. Emmenegger and G. Gordon, Inorg. Chem., 1967, 6, 633.
4964
Phys. Chem. Chem. Phys., 2001, 3, 4957È4964