Oxyhalogen-sulfur chemistry: Kinetics and mechanism of oxidation of formamidine disulfide by acidic bromate

Article abstract of DOI:10.1039/b305353a

The kinetics and mechanism of the oxidation of formamidine disulfide, FDS, a dimer and major metabolite of thiourea, by bromate have been studied in acidic media. In excess bromate conditions the reaction displays an induction period before formation of bromine. The stoichiometry of the reaction is: 7BrO₃^- + 3[(H₂N(HN=)CS-]₂ + 9H ₂O → 6NH₂CONH₂ + 6SO₄ ^2- + 7Br^- + 12H^- (A). In excess oxidant conditions, however, the bromide formed in reaction A reacts with bromate to give bromine and a final stoichiometry of: 14BrO₃^- + 5[(H₂N(HN=)CS-]₂ + 8H₂O → 10NH ₂CONH₂ + 10SO₄^2- + 7Br₂ + 6H⁺ (B). The direct reaction of bromine and FDS was also studied and its stoichiometry is: 7Br₂ + [(H₂N(HN=)CS-] ₂ + 10H₂O → 2NH₂CONH₂ + 2SO₄^2- + 14Br^- + 18H⁺ (C). The overall rate of reaction A, as measured by the rate of consumption of FDS, is second order in acid concentrations, indicating the dominance of oxyhalogen kinetics which control the formation of the reactive species HBrO₂ and HOBr. The reaction proceeds through an initial cleavage of the S-S bond to give the unstable sulfenic acids which are then rapidly oxidized through the sulfinic and sulfonic acids to give sulfate. The formation of bromine coincides with formation of sulfate because the cleavage of the C-S bond to give sulfate occurs at the sulfonic acid stage only. The mechanism derived is the same as that derived for the bromate-thiourea reaction, suggesting that FDS is an intermediate in the oxidation of thiourea to its oxo-acids as well as to sulfate.

Full text of DOI:10.1039/b305353a

View Article Online / Journal Homepage / Table of Contents for this issue
Oxyhalogen–sulfur chemistry: kinetics and mechanism of oxidation
of formamidine disulfide by acidic bromatey
Nicholas Madhiri,z Rotimi Olojo and Reuben H. Simoyi*
Department of Chemistry, Portland State University, Portland, OR 97207-0751, USA
Received 13th May 2003, Accepted 10th July 2003
First published as an Advance Article on the web 5th September 2003
The kinetics and mechanism of the oxidation of formamidine disulfide, FDS, a dimer and major metabolite
of thiourea, by bromate have been studied in acidic media. In excess bromate conditions the reaction displays
ꢀ
an induction period before formation of bromine. The stoichiometry of the reaction is: 7BrO3
=
+
3[(H2N(HN )CS–]2 + 9H2O ! 6NH2CONH2 + 6SO42ꢀ + 7Brꢀ + 12H+ (A). In excess oxidant conditions,
however, the bromide formed in reaction A reacts with bromate to give bromine and a final stoichiometry of:
=
14BrO3 + 5[(H2N(HN )CS–]2 + 8H2O ! 10NH2CONH2 + 10SO42ꢀ + 7Br2 + 6H+ (B). The direct reaction of
ꢀ
=
bromine and FDS was also studied and its stoichiometry is: 7Br2 + [(H2N(HN )CS–]2 + 10H2O !
2NH2CONH2 + 2SO42ꢀ + 14Brꢀ + 18H+ (C). The overall rate of reaction A, as measured by the rate of
consumption of FDS, is second order in acid concentrations, indicating the dominance of oxyhalogen kinetics
which control the formation of the reactive species HBrO2 and HOBr. The reaction proceeds through an initial
cleavage of the S–S bond to give the unstable sulfenic acids which are then rapidly oxidized through the sulfinic
and sulfonic acids to give sulfate. The formation of bromine coincides with formation of sulfate because the
cleavage of the C–S bond to give sulfate occurs at the sulfonic acid stage only. The mechanism derived is the
same as that derived for the bromate–thiourea reaction, suggesting that FDS is an intermediate in the oxidation
of thiourea to its oxo-acids as well as to sulfate.
behavior in chemistry based on sulfur chemistry has outgrown
Introduction
the basic knowledge of the general dynamical behavior of
sulfur compounds. For example, our inability to derive the
mechanism of the bromate–thiourea reaction rendered the
derivation of the oscillatory mechanism for the bromate–
thiourea reaction impossible.4
Invariably, the oxidation of thiourea gives exotic dynamics.2
Nearly all reactions of thiourea with an oxyhalogen species
give nonlinear kinetics characterized by clock reaction charac-
teristics, autocatalysis and/or autoinhibition.3 The oxidation
of thiourea by acidic bromate, for example, gives oscillatory
behavior in a continuously-stirred tank reactor, CSTR.4 The
reaction of thiourea with acidic iodate is oligooscillatory,
and even the simple iodine–thiourea reaction is extremely com-
plex and autoinhibitory. The chlorite–thiourea reaction has
displayed the most varied complex dynamical behavior that
can only be rivaled by the venerable Belousov–Zhabotinski
reaction.5 The chlorite–thiourea reaction is autocatalytic and
bistable in batch, oscillatory in a CSTR and is the only chemi-
cal system to show homoclinic chaos.6 In unstirred solutions it
generates lateral instabilities that are derived from convective
instabilities.7
Previous studies on the oxidation of thiourea had postulated
a series of 2-electron steps that involved the successive addition
of oxygen to the sulfur center followed by the cleavage of the
C–S bond after the formation of the aminoiminomethane-
sulfonic acid.10 The oxidation of the sulfur center to sulfate
involves an oxidation change of the sulfur center from ꢀ2 to
+6. Many intermediates would occur over this large oxidation
state change. One way of proving the veracity of a proposed
mechanistic pathway is to utilize postulated intermediates in
the mechanism as the starting materials. If the mechanism is
correct, then the reactivity of the intermediate should be con-
sistent with its role in the whole mechanism. In the bromate–
thiourea reaction, the first postulated intermediate used was
the stable and isolable aminoiminomethanesulfinic acid,
AIMSA.10 Kinetics experiments and subsequent simulations
proved that it could be an intermediate in the complete oxida-
tion of thiourea by excess oxidant.
Many studies have pointed out that the propensity of sulfur
to polymerize and form extensive S–S bonds constitutes an
in-built nonlinearity that should be factored into any meaning-
ful evaluation of oxyhalogen–sulfur systems. While several
polysulfides exist, for the bulkier molecules, dimerization is
the maximum degree of polymerization that can be achieved.
Several studies of the oxidation of thiocarbamides had con-
cluded that the formation of the dimeric species can be quan-
titative, and that this could be a possible pathway towards the
metabolic oxidation of most biologically active thiocarb-
amides.11 This assertion was based on the fact that the first
step of the S-oxygenation of an organosulfur compound
Efforts at systematically designing chemical oscillators8 were
greatly enhanced by the introduction of two-component oscil-
lators that involved oxyhalogen–sulfur systems.9 While the
relevant chemistry of oxyhalogens is well known, not much
is known with respect to the sulfur-based set of reactions. It
has been difficult to evaluate the genesis of nonlinearities in
aqueous sulfur chemistry. We recently embarked on a series
of studies aimed at evaluating the kinetics and mechanism of
oxidation of sulfur compounds.2 The major aim of this series
of studies was to evaluate the mechanisms of the several
systematically designed two-component oxyhalogen–sulfur
oscillators. The rate of discovery of nonlinear dynamical
y Part 6 of 10 in the series to honor the memory of Dr. Cordelia
R. Chinake (1965–1998). For Part 5 see ref. 1.
y Present address: Department of Chemistry, West Virginia Univer-
sity, Morgantown, WV 26506-6045.
DOI: 10.1039/b305353a
Phys. Chem. Chem. Phys., 2003, 5, 4149–4156
This journal is # The Owner Societies 2003
4149

View Article Online
(e.g. thiourea) will yield a highly unstable sulfenic acid:12
Methods
Experiments were carried out at 25 ꢂ 1.0 ꢃC. The ionic strength
was maintained at 1.0 M (NaCl) in all experiments. Most of
the reactions were performed on a Hi-Tech Scientific SF-61
DX2 double mixing stopped flow spectrophotometer with an
M300 monochromator and a spectrascan diode array control
unit. The signal from the spectrophotometer was digitized
via an Omega Engineering DAS-50/1 16-bit A/D board inter-
faced to a computer for storage and data analysis. Reaction
progress was followed by monitoring the evolution of Br2 at
390 nm. The direct reactions between bromine and FDS were
also monitored by following the consumption of bromine at
390 nm on the stopped flow spectrophotometer using cells with
a 1 cm path length.
X2ðaqÞ þ ðH2NÞ2C
B
S þ H2O
! H2Nð
B
NHÞCSOH þ 2Hþ þ 2Xꢀ ðR1Þ
In this case X2 could be a halogen or any 2-electron oxidant.
The sulfenic acid being highly unstable, should, in the presence
of further oxidant, be rapidly oxidized to the sulfinic acid:
X2ðaqÞ þ H2Nð NHÞCSOH þ H2O
B
! H2Nð
B
NHÞCSO2 H þ 2Hþ þ 2Xꢀ ðR2Þ
In the absence of further oxidizing agent, however, the sulfenic
acid can disproportionate into a number of thiosulfinates,13
but the major pathway is towards dimerization:14
ðH2NÞ2C
B
S þ H2Nð
B
NHÞCSOH
NHÞCS SCð NHÞNH2 þ H2O ðR3Þ
! H2Nð
B
B
Stoichiometric determinations
Formamidine disulfide exists as a doubly-protonated cation
hydrochloride (with each positive charge delocalized over a
carbon center and two nitrogens) under the conditions used
because it is very unstable in alkaline conditions. However,
we will represent it as a neutral molecule. Reaction (R3), which
Stoichiometric determinations were performed by varying the
amount of bromate while keeping concentrations of FDS con-
stant. The required stoichiometry was determined as the point
just before the reaction solution produced bromine as a final
product.
=
produces the dimer of thiourea, H2N( NH)CS–SC( NH)-
=
Analyses for sulfate, bromate and bromine were also per-
formed: sulfate was analyzed gravimetrically as BaSO4 in
excess bromate conditions. The reactions were allowed to sit
for at least 24 h before addition of barium chloride. Subse-
quently the precipitate was allowed to sit for several hours in
a vacuum desiccator before weiꢀghing. In these gravimetric ana-
lysis experiments; excess BrO3 was first removed iodometri-
cally since it forms slightly insoluble Ba(BrO3)2 precipitate
with BaCl2 , thus distorting the precipitation results. Excess
bromate was acidified and mixed with excess iodide and the
liberated iodine was titrated against standard sodium thiosul-
fate with starch as indicator. Bromine was evaluated from its
absorbance at 390 nm (absorptivity coefficient 142 Mꢀ1 cmꢀ1).
=
NH2 (which is written as [(H2N(HN )CS–]2) is also dominant
in the presence of a weak oxidizing agent which might render
reaction (R2) slow (or unfeasible). In such situations, quantita-
tive formation of the dimeric species might be achieved. A
further slow reaction will then occur with the dimeric species
being oxidized further to the sulfinic and sulfonic acids. In
alkaline conditions, the dimer is very unstable but exists as
the doubly protonated cation in acidic medium. Oxidants such
as hydrogen peroxide and peracetic acid can quantitatively
oxidize thiourea to the dimer, formamidine disulfide, FDS,
in exact stoichiometric equivalents.15 However, with any oxi-
dant, and in any quantity, there will always be a competition
between reactions (R2) and (R3).
Our own research work on the mechanism of the oxidation
of thiourea by bromate had postulated a mechanism that
involved sulfenic, sulfinic and sulfonic acids without involving
FDS. Here we report on a comprehensive kinetics and
mechanistic study on the oxidation of FDS by acidic bromate.
The fact that it was overlooked as a possible intermediate in
the oxidation of thiourea could be explained if the FDS is
labile enough to be easily oxidized to two sulfenic acids:
Results
Stoichiometry
The reaction appeared to deliver two stoichiometries based on
the ratio of the initial concentrations of oxidant to reductant,
R. In excess bromate conditions, R > 3, yellow bromine is
formed as a final product, but in conditions of excess reduc-
tant, no formation of bromine is observed. From our previous
studies, it has been generally acknowledged that formation of
bromine is derived from the reaction of one of the products,
bromide, with excess bromate after all the reducing substrate
has been oxidatively saturated. Hence the true stoichiometry
of the reaction under study is one in which there is just enough
bromate to oxidize FDS with no bromate left to conduct the
bromate–bromide reaction and produce bromine. This ratio
could be obtained by extrapolation: bromate was varied for
fixed set of acid and FDS concentrations for ratios that pro-
duce bromine (excess oxidant). At the end of the reaction the
excess oxidizing power was measured via an iodometric titra-
tion in which excess iodide was added and the liberated iodine
titrated with standard thiosulfate:
X2 þ ½ðH2NðHN ÞCS ꢁ2 þ 2H2O
B
! 2H2Nð
B
NHÞCSOH þ 2Hþ þ 2Xꢀ ðR4Þ
If reaction (R4) is rapid (or in overwhelming excess of X2),
it will be kinetically inconsequential with respect to the rate
of consumption of thiourea. However, this manuscript will
examine if FDS is a viable intermediate in the oxidation of
thiourea.
Experimental
Materials
The following reagents were used without further purification:
sodium bromate, formamidine disulfide dihydrochloride
(FDS), 97% (Aldrich), perchloric acid (70–72%), sodium bro-
mide, bromine, sodium chloride, sodium thiosulfate (Fisher).
All the major reactants were assumed to be of high enough
purity with no need for further standardization. FDS solutions
were prepared just before use and were not kept for more than
48 h. These solutions were stored in dark Winchester bottles
and further protected from light by covering them with alumi-
num foil. Reaction solutions were prepared using singly
distilled water.
BrO3ꢀ þ 9Iꢀ þ 6Hþ ! Brꢀ þ 3I3ꢀ þ 3H2O
ðR5Þ
ðR6Þ
I3ꢀ þ 2S2O3 ! S4O62ꢀ þ 3Iꢀ
2ꢀ
The bromate–bromide reaction will not alter the value of the
thiosulfate titer obtained since it will be merely a rearrange-
ment of oxidizing species. One can then plot the volume of titer
obtained against the amount of bromate used. This plot will
give a straight line. This line is extrapolated to the bromate
volume axis for a value for zero titer and this intercept is the
4150
Phys. Chem. Chem. Phys., 2003, 5, 4149–4156

View Article Online
exact amount of bromate needed to just consume FDS. This
series of titrations was carried out for the bromate–FDS
reaction and the results are shown in Fig. 1 where a fixed
amount of 0.001 M FDS was used. The intercept obtained in
this experiment is exactly 0.00233. The empirical ratio was thus
deduced as 3 : 7. Gravimetric analysis showed that all the sul-
fur in FDS was collected as sulfate in which one mole of FDS
gave two moles of sulfate. Combining these two results gave:
Reaction kinetics
The dynamics of this reaction mirror all other bromate oxida-
tions in which there is a quiescent induction period followed by
a rapid formation of bromine in stoichiometric excess of oxi-
dant. This rapid formation of bromine is also accompanied
by a rapid increase in the redox potential of the reaction mix-
ture from approximately 600 to 1100 mV. A number of para-
meters could be measured for this reaction: induction period,
rate of formation of bromine at the end of the induction per-
iod, and the amount of bromine formed. Fig. 2a shows a series
of absorbance traces obtained from varying initial bromate
concentrations while keeping all other parameters constant.
The traces show a reduction in induction period with increas-
ing bromate concentrations. Fig. 2b shows that there is a linear
relationship between the initial bromate concentrations and
the inverse of the induction period. This would seem to suggest
that bromate is involved in the rate-determining step of the
reaction that leads to the end of the induction period to the
first power. Until the rate of the direct reaction of bromine
and FDS is deduced, one can conjecture that the precursor
reaction to the end of the induction period should be the oxi-
dation of FDS by bromate. This reaction takes place during
the induction period even though no noticeable activity is
observed in redox potential and absorbance at 390 nm. It is
7BrO3ꢀ þ 3½ðH2NðHN
ÞCS ꢁ2 þ 9H2O
B
! 6NH2CONH2 þ 6SO42ꢀ þ 7Brꢀ þ 12Hþ ðR7Þ
In excess bromate conditions, the amount of bromine formed
was observed to be proportional to the initial concentration of
FDS used. This can be justified from the stoichiometry of reac-
tion (R7) in which the bromide needed for the further reaction
with bromate is proportional to initial FDS concentrations. By
utilizing the final absorbance of bromine after prolonged
standing, we could deduce stoichiometrically that one mole
of FDS gave exactly 1.4 moles of aqueous bromine (empiri-
cally a ratio of 5 : 7). Sulfate production based on gravimetric
analysis also showed the 1 : 2 ratio with FDS as found in the
stoichiometry of reaction (R7). By calculating the oxidant
equivalents needed to oxidize the sulfur centers, we could
deduce the following stoichiometry in excess bromate environ-
ments:
14BrO3ꢀ þ 5½ðH2NðHN
ÞCS ꢁ2 þ 8H2O
B
! 10NH2CONH2 þ 10SO42ꢀ þ 7Br2 þ 6Hþ ðR8Þ
This stoichiometry could also be obtained by a linear combina-
tion of 5R7 + 7R9 where R9 is the well known BrO3ꢀ/Brꢀ
reaction:16
BrO3ꢀ þ 5Brꢀ þ 6Hþ ! 3Br2ðaqÞ þ 3H2O
ðR9Þ
The stoichiometry of the direct bromine–FDS reaction was det-
ermined both spectrophotometrically and by direct titration.
Aqueous bromine, in the burette was titrated into a conical
flask with a solution of FDS with freshly-prepared starch.
The end-point is when a persistent blue–black color is observed.
The stoichiometry for this reaction was deduced to be:
7Br2 þ ½ðH2NðHN
B
ÞCS ꢁ2 þ 10H2O
! 2ðH2NÞ2C
B
O þ 2SO42ꢀ þ 14Brꢀ þ 18Hþ ðR10Þ
Fig. 2 (a) Effect of bromate variation on FDS depletion at 390 nm.
In excess bromate the reaction displays an induction period before
formation of bromine. [H+]0 ¼ 0.05 M, [FDS]0 ¼ 0.010 M, [BrO3ꢀ]0 :
(a) 0.035, (b) 0.0375, (c) 0.0400, (d) 0.0425, (e) 0.0450 M. (b) Plot of
inverse of induction time vs. bromate concentration, for the data
shown in Fig. 2a. The induction time is inversely proportional to initial
bromate concentration.
Fig. 1 Stoichiometric determinations by titration. The plot is the
titer of thiosulfate obtained vs. the initialꢀbromate concentrations.
[H+]0 ¼ 0.10 M, [FDS]0 ¼ 0.001 M, [BrO3
]
¼ (a) 0.005,(b) 0.006,
0
(c) 0.010, (d) 0.012 M.
Phys. Chem. Chem. Phys., 2003, 5, 4149–4156
4151

View Article Online
difficult to use redox potentials as an analytical tool, but in
general a single redox couple normally predominates in any
reaction mixture. The final redox potential recorded after the
induction period of 1100 mV appears to be derived from the
Br2(aq)/2Brꢀ couple which has a redox potential of 1.08 V
at standard conditions. Hence the post-induction period activ-
ity is the observation of reaction (R9) with no contribution
from the reducing substrates in the reaction mixture which
would have all been consumed. Although acid is not a reactant
in this mechanism, it delivers a very powerful catalytic effect.
In fact, this reaction will not proceed in pH conditions above
3.0. Fig. 3a shows that acid not only shortens the induction
period, but also increases dramatically the rate of formation
of bromine at the end of the induction period. Fig. 3b shows
that there is a linear relationship between the inverse of the
induction period and the square of the initial acid concentra-
tions. Fig. 4a shows the effect of the reducing substrate,
FDS, on the global dynamics of the reaction. The fact that
lower FDS concentrations give shorter induction periods sug-
gests, once again, that the complete consumption of FDS is
a prerequisite to the end of the induction period. Higher
FDS concentrations give longer induction periods and also
give higher rates of formation of bromine at the end of the
induction period. A plot of this rate of formation of bromine
¼ 0.008 M, [H+]0
¼
ꢀ
Fig. 4 (a) Effect of [FDS] variation. [BrO3
]
(a) 0.0025, (b) 0.0050, (c) 0.0075, (d) 0.0100, (e)
0
0.10 M, [FDS]0
0.0125 M. (b) Plot of initial rate dependence vs. [FDS]0 concentration,
¼
for the data shown in Fig. 4a.
vs. the initial FDS concentration gives a straight line (see
Fig. 4b), with an intercept that nearly coincides with the exact
concentration of FDS required to satiate the stoichiometry of
reaction (R7). This is important in the search of a plausible
mechanism for this reaction. The amount of bromide available
at the start of bromine formation (end of induction period) is
related to the initial concentration of FDS according to the
stoichiometry of reaction (R7); and thus, in our analysis of
the rate of formation of bromine according to reaction (R9),
we can replace the role of bromide by FDS. The direct oxida-
tion of FDS by aqueous bromine is moderately fast (Fig. 5a)
but not close to diffusion-limited as some reactions of bro-
mine with other thiols and thiocarbamides.12,17 The reaction
appears to be mildly autoinhibitory. The Br2–FDS reaction
proceeds with the production of large amounts of acid which
can lower the pH of the reaction solution from approximately
4.5 to 1.2 depending on the initial concentrations [see the stoi-
chiometry of reaction (R10)]. Acid retardation has always been
observed in conditions where electrophilic bromine atoms are
supposed to attack a nucleophilic sulfur center. Previous stu-
dies involving thiourea and bromine showed a rapid initial step
in which bromine is consumed in a diffusion-controlled rate to
form the sulfenic acid which in turn is also rapidly oxidized to
form the sulfonic acid which is then slowly oxidized to form
sulfate and the organic residue.12 Apart from trace e, all the
experiments shown in Fig. 5a are in stoichiometric excess
of reductant which yields a product with no bromine. The
kinetics traces show a rapid initial step in which one mole of
bromine is consumed before a second stage in which there is
a slower consumption of bromine. This initial step is so rapid
that it occurs within the mixing time of the stopped-flow spec-
trophotometer. Fig. 5b shows that this reaction is first order in
Fig. 3 (a) Effect of varying acid concentration on the absorbance
traces observed at 390 nm. Acid simultaneously reduces the induction
time and acceleratꢀes the rate of formation of bromine after the induc-
tion period. [BrO3
]
¼ 0.035 M, [FDS]0 ¼ 0.005 M, [H+]0 ¼ (a) 0.05,
0
(b) 0.055, (c) 0.060, (d) 0.065, (e) 0.080 M. (b) Plot of dependence of
inverse induction time vs. [H+]2, for the data shown in Fig. 3a. The
reaction that consumes FDS has a second order dependence on acid.
4152
Phys. Chem. Chem. Phys., 2003, 5, 4149–4156

View Article Online
Fig. 5 (a) Direct reaction of [FDS] with bromine. [FDS]0 ¼ 5 ꢄ 10ꢀ4
M, [H+] ¼ 0.005 M [Br2]0 ¼ (a) 0.0013, (b) 0.0017, (c) 0.0021, (d),
0.0025, (e) 0.0033 M. (b) Plot of initial rate of depletion of Br2 vs.
the concentration of bromine. The reaction is first order [Br2]0 .
Fig. 6 (a) Effect of Brꢀ on the traces at 390 nm. Bromide has a cat-
alytic effect on the reaction (as evidenced by the decrease in induction
time). Higher bromide concentrations also increased the amount of Br2
formed and the rate of bromine formation. [FDS]0 ¼ 0.01 M,
(b) 2.5 ꢄ 10ꢀ4, (c) 5.0 ꢄ 10ꢀ4 M. (d) A log–log plot of the inverse of
the induction period vs. bromide concentrations. The data used are
derived from Fig. 6a.
bromine concentration. A similar series of experiments showed
that the reaction is also first order in FDS.
Apart from acid, bromide also seems to have a very strong
catalytic effect on the whole reaction even though it is not
a reactant according to the stoichiometries of reactions (R7)
and (R8). Since it is a product in reaction (R7), its involvement
in the acceleration of the reaction renders it as an autocatalyst.
Fig. 6a shows the effect of bromide in which it drastically
reduces the induction period and also increases the rate of for-
mation of bromine after the induction period. Fig. 6a also
shows that this catalytic effect quickly saturates. While there
is a noticeable reduction in induction period in going from
no added bromide to 5 ꢄ 10ꢀ5 and 1.5 ꢄ 10ꢀ4 M; further
increases in initial bromide concentrations do not show an
equally dramatic effect. A log–log plot of bromide concentra-
tions vs. inverse of induction period gives a straight line as
shown in Fig. 6b.
[BrO3ꢀ] ¼ 0.008 M, [H+] ¼ 0.05 M, [Brꢀ]0 added ¼ (a) 5.0 ꢄ 10ꢀ5
,
reaction only serves as a chain-initiation step.
BrO3ꢀ þ Hþ ! HBrO3
ðR11Þ
HBrO3 þ ½ðH2NðHN
B
ÞCS ꢁ2 þ H2O
! 2ðH2NðHN
B
ÞCSOH þ HBrO2 ðR12Þ
=
Both the sulfenic acid, (H2N(HN )CSOH and bromous acid,
HBrO2 , are very unstable and should disproportionate or
=
react further to give the more stable sulfinic acid, H2N(HN )
CSO2H, or bromide, Brꢀ, respectively
Proposed mechanism
ðH2NðHN
B
ÞCSOH þ HBrO2
! ðH2NðHN
The kinetics data summarized in Figs. 2 to 4 suggest a reaction
with a rate law of the form:
B
ÞCSO2H þ HOBr ðR13Þ
ðH2NðHN
B
ÞCSOH þ HOBr
! ðH2NðHN
2
Rate /½BrO3ꢀꢁ½Hþꢁ ½FDSꢁ
and the kinetics data in Fig. 6 also suggest the following
relationship:
ð1Þ
B
ÞCSO2H þ Brꢀ þ Hþ ðR14Þ
The sulfinic acid, aminoiminomethanesulfinic acid, is known
to be very stable and is a well-known metabolite of thiourea.
Addition of reactions R11 + R12 + R13 + R14 gives the over-
all reaction of:
Rate /½Brꢀꢁ
ð2Þ
The nucleophilic nature of the substrate suggests that reaction
with the oxidant is initiated through the initial formation of
bromic acid, which then attacks the S–S bond to commence
the autocatalytic formation of the reactive species. This initial
BrO3ꢀ þ ½ðH2NðHN
B
ÞCS ꢁ2 þ H2O
! 2ðH2NðHN
B
ÞCSO2H þ Brꢀ ðR15Þ
Phys. Chem. Chem. Phys., 2003, 5, 4149–4156
4153

View Article Online
reaction will be bromide-controlled. In highly excess bromate
The bromide formed in reaction (R15) can initiate the most
important oxyhalogen reaction (R16) whose sole purpose in
bromate oxidations is to form the reactive species HBrO2
and HOBr that are responsible for the bulk of the oxidation:18
conditions, R ! 15; then we can assume that the concentrations
of bromate and acid will not have changed much at the point
where bromine production commences; [BrO3]t ꢆ [BrO3]0 ,
[H+]t ꢆ [H+]0 . The only variables would be bromide concen-
trations and other reactive oxybromine species that might still
be in the reaction mixture at the point where all the reductant
has been consumed. Control experiments run with stoichio-
metric amounts of bromide as derived from the stoichiometry
of reaction (R7) gave rates of formation of bromine according
to the expected kinetics of reaction (R9) that were much slower
than those observed in our experiments with FDS. This would
suggest that there is no quantitative formation of bromide
before formation of bromine commences: the true picture
might involve a mixture of HBrO2 , HOBr and Brꢀ with no
reducing species left in the reaction mixture. Without the con-
centrations of all the reacting species known to a certainty, one
cannot evaluate kinetics constants for this process apart from
extrapolations that can be derived from computer simulations.
However, we managed to derive the dependence of the rates of
formation of bromine with respect to the buffered species: bro-
mate and acid via log–log plots. A log–log plot of rate of
formation of bromine and initial acid concentrations gave a
straight line of slope 2.0 and the same log–log plot with respect
to bromate concentrations gave a straight line with a slope
very close to unity. This suggests, again, that reaction (R16)
is the rate-determining step with respect to the formation of
bromine at the end of the induction period.
BrO3ꢀ þ 2Hþ þ Brꢀ ! HBrO2 þ HOBr
ðR16Þ
Both HBrO2 and HOBr will end up as Brꢀ, which means that
overall, there will be an autocatalytic production of bromide if
reactions (R13) and (R14) are both faster than reaction (R12)
because one mole of bromide in reaction (R16) will deliver
two moles of bromide (quadratic autocatalysis). As bromide
increases in concentration, however, HBrO2 would rather
rapidly decompose to hypobromous acid rather than partici-
pate as an oxidant as in reaction (R13):19
HBrO2 þ Brꢀ þ Hþ ! 2HOBr; kf ¼ 2 ꢄ 106; kr ꢅ 0 ðR17Þ
Addition, then, of reactions R11 + R12 + 2R14 + R17 will also
deliver quadratic autocatalysis in bromide:
BrO3ꢀ þ ½ðH2NðHN
B
ÞCS ꢁ2 þ H2O þ Brꢀ
! 2ðH2NðHN
B
ÞCSO2H þ 2Brꢀ ðR18Þ
The autocatalytic production of bromide will render reac-
tion (R16) as the rate-determining step instead of initiation
reaction (R12). The kinetics and mechanism of this step have
been studied in detail, and the rate law governing this step is:
2
Rate ¼ k0½BrO3ꢀꢁ½Brꢀꢁ½Hþꢁ
ð3Þ
Rate law (3) is supported by the data in Figs. 2b and 3b. The
presence of [FDS] in the rate law [see eqn. (1)] is through com-
posite reaction (R15). The rate of formation of bromide in this
reaction will be directly proportional to the initial FDS con-
centration. Thus the bromide concentrations in eqn. (3) will
be proportional to FDS concentrations, and the overall rate
law for the reaction will be given by the rate equation:
The bromine–FDS reaction
The oxidation of FDS by aqueous bromine was straight-
forward with the rate law:
ꢀd½Br2ꢁ=dt ¼ k17½Br2ꢁ½FDSꢁ
The initial step is the formation of the sulfenic acid:
ð5Þ
2
Rate ¼ kapp½BrO3ꢀꢁ½Hþꢁ ½FDSꢁ
ð4Þ
Br2ðaqÞ þ ½ðH2NðHN
B
ÞCS ꢁ2 þ 2H2O
where kapp is the apparent rate constant if we were observing
the rate of consumption of bromate or FDS. The observation
of the rate of depletion of bromate or FDS could not be unam-
biguously observed due to their lack of absorptivity in the UV
region.
! 2 ðH2NðHN
B
ÞCSOH þ 2Brꢀ þ 2Hþ ðR22Þ
Reaction (R22) is a very rapid step with a lower limit bimole-
cular rate constant which we deduced to be k22 ¼ 6 ꢄ 105 Mꢀ1
s
ꢀ1. The next step, though still quite rapid, will involve the for-
mation of the sulfinic acid, and this portion of the reaction
could be captured on a stopped-flow spectrophotometer. We
experimentally evaluated a bimolecular rate constant for the
Further oxidation
In the presence of excess oxidant, we expect the oxo-acids to be
continuously oxidized via S-oxygenation until the C–S bond is
cleaved and sulfate is formed.20 Experiments with barium
chloride as an indicator have shown that the formation of sul-
fate coincides with the formation of bromine. Thus there is a
sequential and orderly oxidation from the sulfenic acid to
the sulfinic, sulfonic, and finally to sulfate. We can safely
assume that the major oxidant in getting FDS to sulfate is
HOBr (if R17 is as fast as it is known to be in literature):
second slower part of the reaction of 25 ꢂ 5 Mꢀ1
s
ꢀ1. As the
reaction proceeds, it produces, as major products, protons
and bromide ions. The mild autoinhibition observed is due
to the formation of the tribromide anion which is a much
poorer electrophile than molecular bromine:21
Br2ðaqÞ þ Brꢀ ! Br3ꢀ; Keq ¼ 17
ðR23Þ
Experiments undertaken of bromine oxidations with added
bromide have shown a measurable retardation of the reaction.
The continuous formation of bromide as the reaction proceeds
pushes the equilibrium of reaction (R23) to the right with con-
tinuously less and less molecular bromide available for reac-
tion. This tribromide equilibrium does not assert itself as
much in the formation of bromine at the end of the induction
period because the bromide formed in reaction (R7) will con-
tinuously be consumed by reaction (R9) to produce bromine.
ðH2NðHN
B
ÞCSO2H þ HOBr
! H2NðHN
B
ÞCSO3H þ Brꢀ þ Hþ ðR19Þ
ðH2NðHN
B
ÞCSO3H þ HOBr þ H2O
! ðH2NÞ2C
B
O þ SO42ꢀ þ Brꢀ þ 3Hþ ðR20Þ
Formation of bromine
Computer simulations
If we assume that there is only one reaction in solution pro-
ducing bromine:
Our inability to experimentally follow the rate of consumption
of FDS made computer simulations studies mandatory in this
mechanism. Our experimental data allowed us to produce a
very concise and abbreviated mechanism that could still ade-
quately describe the observed reaction dynamics. Although
HOBr þ Brꢀ þ Hþ ! Br2ðaqÞ þ H2O
then the rate of formation of bromine will be determined by
the rate of formation of HOBr. This whole section of the
ðR21Þ
4154
Phys. Chem. Chem. Phys., 2003, 5, 4149–4156

View Article Online
we are aware that there are four oxidizing species in the reac-
tion mixture: BrO3ꢀ, HBrO2 , HOBr and Br2(aq) as well as
four reducing species: FDS, the sulfenic, sulfinic and sulfonic
acids, there was no need to produce a mechanism with 16 oxi-
dation–reduction reactions that permute all possible combina-
tions because we could assume HOBr and Br2(aq) as the major
oxidizing species. This could be possible after our assertion
that reaction (R16) was rate-determining and that reaction
(R17) was fast. The full mechanism used for our modeling
studies is shown in Table 1. It consists of three oxyhalogen
reactions (P1–P3) whose kinetic constants have been deduced
from previous studies,19 a rapid protolytic reaction (P4), initia-
tion reactions (P5–P8) and eight oxybromine–sulfur reactions
(P7–P14) in which only aqueous bromine and HOBr have been
utilized as oxidants. Sulfur–sulfur reactions were ignored in
this mechanism because they would be inconsequential in
excess oxidant. Reaction (P4) is known to be diffusion-
controlled. The use of high diffusion-controlled rate constants in
this mechanism greatly increased the stiffness of the generated
ordinary differential equations resulting in simulations that
needed up to 24 h on a Pentium IV 2.4 Ghz computer. The
use of lower than diffusion-controlled rate constants increased
computational efficiency without loss in the accuracy of the
model for as long as the chosen rate constants did not render
reaction (P4) rate-limiting. The stiff set of ordinary differential
equations derived from the kinetics rate laws was integrated by
semi-implicit forth order Runge–Kutta techniques derived
from the CKS software generated by IBM’s Almaden group
as well as the Kintecus software developed by Ianni.22 The
simulations were most sensitive to the kinetics parameters used
for reaction (P5) and the bimolecular rate constant of 400 Mꢀ1
sꢀ1 was based on the best fit to the induction period. Kinetics
parameters used for reactions (P6), (P7), (P8) and (P10) were
Fig. 7 Computer modeling of the reaction mechanism shown in
Fig. 3a, trace e. [FDS]0 ¼ 0.005 M, [BrO3ꢀ] ¼ 0.035 M, [H+] ¼ 0.08 M.
chosen as lower limit values such that these reactions were not
rate-determining. The values for (P12) and (P13) were deduced
from this study. The ‘kink’ which is evident in Fig. 3a in which a
two-stage formation of bromine is observed after the induction
period could only be modeled if kP14 > kP13 . Thus the value of
kP14 was guessed after the experimentally estimated kP13 . Fig. 7
shows that this simple mechanism can easily simulate the induc-
tion period and bromine formation. The data simulated is from
Fig. 3a. The model was able to satisfactorily reproduce acid and
bromate effects in induction periods and rate of formation of
bromine at the end of the induction period.
Table 1 The reduced bromate–FDS mechanism used for computer
simulations
Conclusion
Number Reaction
kf ; kr
The oxidation of the thiourea dimer, FDS, by bromate has a
mechanism that closely mimics the thiourea–bromate mechan-
ism. The only difference lies in the stoichiometry and the fact
that the bromate–FDS reaction is much faster than the corre-
sponding bromate–thiourea reaction. This study clearly shows
that in an oxidative environment, FDS would be one of the
first intermediates formed from thiourea. The physiological
environment also supports the formation of the dimer as the
first step in the oxidative metabolism of thiocarbamides.
P1
BrO3ꢀ + 2H+ + Brꢀ
2.1; 1 ꢄ 104
Ð HBrO2 + HOBr
HBrO2 + Brꢀ + H+ Ð 2HOBr
HOBr + Brꢀ + H+
P2
P3
2 ꢄ 106; 2 ꢄ 10ꢀ5
8.9 ꢄ 108; 110
Ð Br2(aq) + H2O
P4
P5
BrO3ꢀ + H+ Ð HBrO3
1 ꢄ 107; 1 ꢄ 105 a
=
HBrO3 + [H2N(HN )CS–]2 + H2O
400
=
! 2H2N(HN )CSOH + HBrO2
P6
5 ꢄ 104
1 ꢄ 105
1 ꢄ 105
100
=
HBrO2 + [H2N(HN )CS–]2 + H2O
=
! 2H2N(HN )CSOH + HOBr
=
P7
HOBr + [H2N(HN )CS–]2 + H2O
=
! 2H2N(HN )CSOH + Brꢀ + H+
Acknowledgements
=
P8
HOBr + H2N(HN )CSOH
=
! H2N(HN )CSO2H + Brꢀ + H+
We would like to thank Dr. Ian Love of the University of
Zimbabwe for alerting us to the existence of the Kintecus mod-
eling software. This work was supported by grant number
CHE-0137435 from the National Science Foundation.
=
P9
HOBr + H2N(HN )CSO2H
=
! H2N(HN )CSO3H + Brꢀ + H+
=
P10
P11
P12
P13
P14
HOBr + H2N(HN )CSO3H +H2O
=
700
! (H2N)2C O + SO42ꢀ + Brꢀ + 3H+
6 ꢄ 105
2 ꢄ 104
20
=
Br2 + [H2N(HN )CS–]2 + 2H2O
! 2H2N(HN )CSOH + 2Brꢀ + 2H+
=
=
Br2 + H2N(HN )CSOH + H2O
=
References
! H2N(HN )CSO2H + 2Brꢀ + 2H+
=
Br2 + H2N(HN )CSO2H + H2O
=
1
2
3
B. S. Martincigh and R. H. Simoyi, J. Phys. Chem. A, 2002, 106,
482–489.
C. R. Chinake and R. H. Simoyi, J. Phys. Chem., 1993, 97,
11 569.
I. R. Epstein, K. Kustin and R. H. Simoyi, J. Phys. Chem., 1992,
96, 5852.
! H2N(HN )CSO3H + 2Brꢀ + 2H+
=
Br2 + H2N(HN )CSO3H + 2H2O
=
120
! (H2N)2C O + SO42ꢀ + 2Brꢀ + 4H+
a
Though this reaction is diffusion-controlled, these values were lower-
limit kinetics parameters that could be used without rendering this
reaction rate-determining. The use of diffusion-controlled rate con-
stants rendered the mechanism extremely stiff and computationally
slow.
4
5
R. H. Simoyi, J. Phys. Chem., 1986, 90, 2802.
A. L. Lin, M. Bertram, K. Martinez, H. L. Swinney, A. Ardelea
and G. F. Carey, Phys. Rev. Lett., 2000, 84, 4240.
C. J. Doona, R. Blittersdorf and F. W. Schneider, J. Phys. Chem.,
1993, 97, 7258.
6
Phys. Chem. Chem. Phys., 2003, 5, 4149–4156
4155

View Article Online
15 Z. Abednizadeh, M. Gardes-Albert and C. Ferradini, Can.
J. Chem., 1989, 67, 1247.
16 R. M. Noyes, R. J. Field and R. C. Thompson, J. Am. Chem.
Soc., 1971, 93, 7315.
17 I. R. Epstein, K. Kustin and R. H. Simoyi, J. Phys. Chem., 1992,
96, 6326.
7
8
C. R. Chinake and R. H. Simoyi, J. Phys. Chem., 1994, 98, 4012.
I. R. Epstein, K. Kustin, P. De Kepper and M. Orban, Sci. Am.,
1983, 248, 112.
9
M. Orban, P. De Kepper and I. R. Epstein, J. Phys. Chem., 1982,
86, 431.
10 R. H. Simoyi, I. R. Epstein and I. Kustin, J. Phys. Chem., 1994,
98, 551.
11 C. Hwang, A. J. Sinskey and H. F. Lodish, Science, 1992, 257,
1496.
18 R. M. Noyes, J. Am. Chem. Soc., 1980, 102, 4644.
19 I. Szalai, J. Oslonovitch and H. D. Forsterling, J. Phys. Chem. A,
2000, 104, 1495.
12 R. H. Simoyi and I. R. Epstein, J. Phys. Chem., 1987, 91, 5124.
13 S. Oae and T. Tajata, Tet. Lett., 1980, 21, 3213.
14 E. Shouji and D. A. Buttny, J. Phys. Chem. B, 1999, 103,
2239.
20 S. B. Jonnalagadda, C. R. Chinake and R. H. Simoyi, J. Phys.
Chem., 1996, 100, 13 521.
21 S. Plotkin and A. Haim, Inorg. Chim. Acta, 1998, 270, 189.
22 J. C. Ianni, Kintecus, 1995, ver. 3.
4156
Phys. Chem. Chem. Phys., 2003, 5, 4149–4156