Mechanism of Chlorine Dioxide and Chlorate Ion Formation from the Reaction of Hypobromous Acid and Chlorite Ion

Article abstract of DOI:10.1021/ic980262q

The rate of oxidation of ClO₂^- by HOBr is first-order in each reactant and is general-acid-assisted in the presence of phosphate or carbonate buffers. The products are ClO₂ and ClO₃^-, where the relative yield depends on the concentration ratio of ClO₂^-/OH^-. The kinetic dependence indicates the presence of a steady-state intermediate, HOBrOClO^- (or HOBrClO₂^-), that undergoes general-acid-assisted reactions to generate a metastable intermediate, BrOClO (or BrClO₂). This intermediate reacts very rapidly by two competing pathways: in one path ClO₂^- reacts to form 2ClO₂ and Br^-, and in the other path OH^- (or H₂O) reacts to form ClO₃^- and Br^-. Competition between these pathways determines the yield of ClO₂ but does not affect the rate of loss of HOBr. The reactions are followed by the formation of ClO₂ in the presence of excess ClO₂^-. The rate expression for the loss of HOBr is k₁[ClO₂^-][HOBr]Σ(k _HA[HA])/(k_-1 + Σ(k_HA[HA])), where k₁ (for the formation of the intermediate) is 97 M^-1 s^-1 and k_HA/k_-1 (M^-1) values, which depend on the acid (HA) strength, are 3.1 × 10⁵ for H₃O⁺, 8.3 for H₂PO₄^-, and 0.064 for HCO₃^- (25.0 °C, μ = 1.0 M). Reactions between HOBr and ClO₂^- are much faster than those between HOCl and ClO₂^-.

Full text of DOI:10.1021/ic980262q

Inorg. Chem. 1998, 37, 4321-4327
4321
Mechanism of Chlorine Dioxide and Chlorate Ion Formation from the Reaction of
Hypobromous Acid and Chlorite Ion
Christopher S. Furman and Dale W. Margerum*
Department of Chemistry, Purdue University, West Lafayette, Indiana 47907-1393
ReceiVed March 9, 1998
The rate of oxidation of ClO2- by HOBr is first-order in each reactant and is general-acid-assisted in the presence
-
of phosphate or carbonate buffers. The products are ClO2 and ClO3 , where the relative yield depends on the
-
-
concentration ratio of ClO2 /OH . The kinetic dependence indicates the presence of a steady-state intermediate,
-
-
HOBrOClO (or HOBrClO2 ), that undergoes general-acid-assisted reactions to generate a metastable intermediate,
-
BrOClO (or BrClO2). This intermediate reacts very rapidly by two competing pathways: in one path ClO2
reacts to form 2ClO2 and Br , and in the other path OH (or H2O) reacts to form ClO3 and Br . Competition
between these pathways determines the yield of ClO2 but does not affect the rate of loss of HOBr. The reactions
are followed by the formation of ClO2 in the presence of excess ClO2 . The rate expression for the loss of HOBr
is k1[ClO2 ][HOBr]∑(kHA[HA])/(k-1 + ∑(kHA[HA])), where k1 (for the formation of the intermediate) is 97 M
-
-
-
-
-
-
-1
-
1
-1
5
+
-
s
and kHA/k-1 (M ) values, which depend on the acid (HA) strength, are 3.1 × 10 for H3O , 8.3 for H2PO4 ,
-
-
and 0.064 for HCO3 (25.0 °C, µ ) 1.0 M). Reactions between HOBr and ClO2 are much faster than those
-
between HOCl and ClO2 .
Introduction
the steps leading up to the formation of the metastable
intermediate are regarded as rate-limiting. Studies by Gordon
Intermediates of XClO2 (where X ) Cl, Br, or I) have been
proposed as transient species in chlorite ion reactions with
molecular chlorine, bromine, or iodine, as well as with hypo-
2
-4,7
8
and co-workers
and by Peintler et al. provide additional
-
kinetic information about the reactions of ClO2 with HOCl
over a range of pH and reactant ratios. High concentrations of
phosphate or acetate buffer were used in these experiments, but
kinetic effects of the buffer were not considered.
1
-16
halous acids (HOX, where X ) Cl or Br).
Taube and
1
Dodgen were the first investigators to propose a metastable
intermediate, Cl2O2, in the reactions of ClO2- with Cl2 or HOCl.
Their tracer experiments with radioactive chlorine gave unla-
15
Valdes-Aguilera et al. studied the reactions of bromine with
-
-
2-
ClO2 and HClO2 in three buffer systems (HSO4 /SO4
,
-
-
beled ClO2 and ClO3 as products along with labeled Cl . They
concluded that the main pathway to products involved an
intermediate in which the chlorine atoms were distinct from
one another. An unsymmetrical intermediate
-
-
ClCH2COOH/ClCH2COO , and CH3COOH/CH3COO ) and
reported that ClO2 was the exclusive oxidation product (no
-
ClO3 was found). They proposed BrClO2 as an intermediate
-
O
O
(eq 5) and, in contrast to the behavior of Cl2O2, they propose
(
Cl-Cl- or Cl-O-Cl-O) was proposed with the mecha-
-
that the BrClO2 reaction with ClO2 (eq 6) is the rate-
nistic steps in eqs 1-4.
Cl * + ClO2 f Cl*ClO + Cl*
determining step.
-
2
-
-
-
Br + ClO a BrClO + Br
(5)
(6)
(1)
(2)
2
2
2
2
-
-
+
-
BrClO + ClO f 2ClO + Br
H + HOCl* + ClO2 f Cl*ClO + H O
2
2
2
2
2
A complex pH dependence was observed, but kinetic effects
of the buffers were not considered. They also indicated that
HOBr was much less reactive than Br2 in its reactions with
fast
2
Cl*ClO2 8 Cl * + 2ClO2
(3)
(4)
2
-
ClO2 .
fast
-
-
+
Cl*ClO + H O 8 Cl* + ClO3 + 2H
2
2
(
(
(
7) Gordon, G.; Tachiyashiki, S. EnViron. Sci. Technol. 1991, 25, 468-
474.
The metastable intermediate, Cl2O2, can decompose either by
a second-order reaction (eq 3) to give chlorine dioxide or by a
first-order reaction with water (eq 4) to yield the chlorate ion.
Equations 3 and 4 are considered to be fast reactions, whereas
8) Peintler, G.; Nagypal, I.; Epstein, I. R. J. Phys. Chem. 1990, 94, 2954-
2958.
9) Fabian, I.; Gordon, G. Inorg. Chem. 1992, 31, 2144-2150.
(
10) Grant, J. L.; DeKepper, P.; Epstein, I. R.; Kustin, K.; Orban, M. Inorg.
Chem. 1982, 21, 2192-2196.
(
11) Aieta, E. M.; Roberts, P. V. EnViron. Sci. Technol. 1986, 20, 50-55.
(
(
(
1) Taube, H.; Dodgen, H. J. Am. Chem. Soc. 1949, 71, 3330-3336.
2) Emmenegger, F.; Gordon, H. Inorg. Chem. 1967, 6, 633-635.
3) Gordon, G.; Kieffer, R. G.; Rosenblatt, D. H. In Progress in Inorganic
Chemistry; Lippard, S. J., Ed.; Wiley-Interscience: New York, 1972;
Vol. 15, pp 201-286.
4) Tang, T.; Gordon, G. EnViron. Sci. Technol. 1984, 18, 212-216.
5) Rabai, G.; Beck, M. T. Inorg. Chem. 1987, 26, 1195-1199.
6) Rabai, G.; Orban, M. J. Phys. Chem. 1993, 97, 5935-5959.
(12) Ni, Y.; Kubes, G. J.; Van Heiningen, A. R. P. J. Pulp Pap. Sci. 1993,
19, J1-J6.
(13) De Meeus, J.; Sigalla, J. J. Chim. Phys.-Chim. Biol. 1966, 63, 453-
459.
(14) Epstein, I. R.; Kustin, K. J. Phys. Chem.1985, 89, 2275-2282.
(15) Valdez-Aguilera, O.; Boyd, D. W.; Epstein, I. R.; Kustin, K. J. Phys.
Chem. 1986, 90, 6696-6702.
(
(
(
(16) Emerson, D. W. Ind. Eng. Chem. Res. 1993, 32, 1228-1234.
S0020-1669(98)00262-6 CCC: $15.00 © 1998 American Chemical Society
Published on Web 07/21/1998

4322 Inorganic Chemistry, Vol. 37, No. 17, 1998
Furman and Margerum
In wastewater treatment, chlorine is a widely used disinfec-
and ClO2-. The present work addresses the kinetics and
mechanism of this reaction.
tant.17 Chlorine hydrolysis (eq 7) is relatively rapid with a
-
1
18
hydrolysis rate constant of 22.3 s at 25.0 °C.
Experimental Section
+
-
Cl (aq) + H O a HOCl + H + Cl
(7)
Reagents. Solutions were made with deionized, distilled
water. Working solutions of NaOCl (ClO3 -free) were prepared
2
2
-
from a stock solution obtained by slowly bubbling Cl2(g)
through stirred solutions of NaOH (∼0.1 M) maintained at 0-4
°C. The NaOCl solutions were standardized spectrophotometri-
cally. The molar absorptivity for NaOCl was determined by
A typical range of bromide ion concentrations in groundwater
is 0.01-3 mg/L.19 Chlorine reacts extremely rapidly with Br
-
9
-1 -1
(
the rate constant is 7.7 × 10 M s ) to give BrCl(aq), which
5 -1 20
hydrolyzes rapidly (k > 10 s ) to give HOBr. The reaction
-
two methods. Iodometric titrimetric methods gave ꢀ ) 362
between HOCl and Br also is proposed to proceed through
292
-
1
-1 29
(
2 M cm . In a second method, a solution of NaNO2
BrCl, followed by rapid hydrolysis to HOBr. The rate constant
3
-1 -1 21
was added in excess to the NaOCl, and the mixture was adjusted
to pH 5 with acetic acid. After complete reaction, the solution
for the overall reaction (eq 8) is 1.55 × 10 M s .
-
-
-
was diluted and analyzed for NO3 by capillary electrophoresis
HOCl + Br f HOBr + Cl
(8)
-
(
CE) methods. The [NO3 ] determined was used to calculate
-
-1
-1
the [OCl ]. This gave 362 ( 5 M cm as the molar
absorptivity for OCl at 292 nm, in agreement with the
Mixtures of HOCl and HOBr can react to generate chlorate
ion (eq 9) or bromate ion (eq 10).22
-
iodometric method.
Commercially available NaClO2 was recrystallized by using
HOBr
-
+
-
3
0
3HOCl
8 ClO3 + 3H + 2Cl
(9)
a procedure modified from previous reports. The carbonate
content of the commercial solid (measured by CE) was removed
by precipitation with a solution of BaCl2. (Some Ba(ClO2)2
coprecipitated.) The remaining supernatant liquid was collected
and cooled in an ice bath. The NaClO2 was recrystallized in
the temperature range from -5 to -15 °C, collected by vacuum
filtration, and washed with 100% acetone. A final recrystalli-
zation step was performed on the solid from a 75:25 acetone/
water mixture. The solid was washed, and argon gas was used
to evaporate residual acetone from the final product. The
NaClO2 was stored over P2O5 in vacuo and kept in the dark.
The purity was determined by standard iodometric titrimetry,
and the product was confirmed to be free of other anionic
impurities by capillary electrophoresis. The purified NaClO2
-
-
+
2
HOCl + HOBr f BrO3 + 2Cl + 3H
(10)
Bromate ion is a carcinogen and nephrotoxin.23-28 The
Environmental Protection Agency has proposed that the maxi-
mum contaminant level (MCL) in drinking water should be less
than 0.01 mg/L, with a maximum contaminant level goal
MCLG) of no detectable BrO3 . Data on the health effects
associated with exposure to ClO3 are not complete, so chlorate
-
19
(
-
ion will not be regulated as part of the Disinfectants and
Disinfection Byproduct Proposed Rule.19 However, it is a
candidate for future regulation. On the other hand, the proposed
MCL for chlorite ion (ClO2 ) is 1.0 mg/L. Reactions between
-
-
1
-1
has a molar absorptivity of ꢀ ) 154.0 ( 0.7 M cm at 260
nm. Sodium chlorite solutions were prepared from the recrys-
tallized solid (99.98%).
mixtures of HOCl and HOBr initially produce chlorite ion and
bromide ion (eq 11).22
+
HOCl + HOBr f ClO2- + 2H Br-
Chlorine dioxide was synthesized by a process described in
a review paper by Masschelein. Acetic anhydride was added
(11)
31
to a solution of NaClO2, and argon gas was used to strip off
ClO2 from the main reaction tower. A second tower contained
a 5% solution of NaClO2 and served as a scrubber to remove
any possible Cl2 that may have formed in the process. The
ClO2 was trapped in a third gas collection tower which contained
H2O at 0-4 °C. Solutions of ClO2 were freshly diluted prior
to use. Small amounts of n-pentane were added to the solutions
to cover the surface and retard the evaporation of ClO2, as
If HOCl is in higher concentration than HOBr (which is typically
the case in water chlorination), the Br- generated is rapidly
converted back to HOBr (eq 8). As a consequence, the next
stage in the halogen redox process is the reaction between HOBr
(
17) Maxted, J. R. Proceedings Wastewater Disinfection AlterantiVes,
Design, Operation and Effectiveness Workshop Water Pollution
Control Federation, Washington, DC, 1983.
3
2
(
(
18) Wang, T. X.; Margerum, D. W. Inorg. Chem. 1994, 33, 1050-1055.
19) USEPA. National Primary Drinking Water Regulations; Disinfectants
and Disinfection Byproducts; Proposed Rule; Fed. Regist. 1994, 59,
recommended by Lengyel and co-workers. The molar ab-
sorptivity was found to be 1230 ( 10 M-1 cm-1 at 359 nm by
using standard iodometric titrimetry.
3
8668-38829.
-
-
(
20) Wang, T. X.; Kelley, M. D.; Cooper, J. N.; Beckwith, R. C.; Margerum,
Solutions of BrO3 and ClO3 were prepared from their
respective sodium salts obtained from Aldrich. Reagent purity
D. W. Inorg. Chem. 1994, 33, 5872-5878.
(
(
(
21) Kumar, K.; Margerum, D. W. Inorg. Chem. 1987, 26, 2706-2711.
22) Lewin, M.; Avrahami, M. J. Am. Chem. Soc. 1955, 77, 4491-4498.
23) Japanese Ministry of Health and Welfare, Cancer Research Report,
-
4+
was confirmed by standard iodometric (for BrO3 ) and Ce /
Fe (for ClO3 ) titrations.
2
+
-
1
976, pp 744-745 (Aug 1977).
Solutions of bromide-free HOBr were made by the reaction
(
(
(
24) Kurokawa, Y.; Hayashi, Y.; Maekawa, A.; Takahashi, M.; Kokubo,
T.; Odashima, S. J. Natl. Cancer Inst. 1983, 71, 965-982.
25) Kurokawa, Y.; Maekawa, A.; Takahashi, M.; Hayashi, Y. EnViron.
Health Perspect. 1990, 87, 309-335.
-
of HOCl with Br in a 1:1 mole ratio. An aliquot of this
solution was made basic, and the solution was standardized
spectrophotometrically based on the molar absorptivity of
26) Gosselin, R. E.; Smith, R. P.; Hodge, H. C.; Braddock, J. E. In Clinical
Toxicology of Commercial Products; Tracy, T. M., Ed.; Williams &
Wilkins Inc.: Baltimore, MD, 1984; pp II 3-4, 110-112, III 74-
-1
-1
33
NaOBr, ꢀ ) 332 M cm at 329 nm.
Under reaction
77.
(29) Beckwith, R. C.; Margerum, D. W. Unpublished results.
(
27) Budavari, S. The Merck Index; Merck & Co., Inc.: Rahway, NJ, 1989;
pp 1212-1213.
(30) Fabian, I.; Gordon, G. Inorg. Chem. 1992, 31, 3785-3787.
(31) Masschelein, W. J. J. Am. Water Works Assoc. 1984, 76, 70-76.
(32) Lengyel, I.; Li, J.; Epstein, I. R. J. Phys. Chem. 1992, 96, 7032-
(33) 7T 0r o3 y7 ,. R. C.; Margerum, D. W. Inorg. Chem. 1991, 30, 3538-3543.
(28) Loebl, S. The Nurse’s Drug Handbook; John Wiley & Sons Inc.: New
York, 1989; pp 1164, 1166.

ClO2 and ClO3- Formation from HOBr and ClO2-
Inorganic Chemistry, Vol. 37, No. 17, 1998 4323
Table 1. Observed Rate Constants and Fraction of ClO Produced
conditions where an impurity of BrO2- (4-5%) was problematic
in our analyses, an alternate preparation of HOBr was used
where liquid Br2 was added to cold solutions of NaOH. This
2
-
2
a
from HOBr Reaction with Excess ClO
[ClO2-], M
p[H+]
kobsd, s-1
fraction ClO2
-
solution was relatively free of BrO2 (less than 0.01 µM). The
0
0
.001 344
.001 973
6.68
6.67
6.68
6.66
6.68
6.66
6.68
6.66
6.67
6.66
6.67
6.67
6.67
0.0281(6)
0.042(3)
0.0582(4)
0.075(2)
0.0850(8)
0.106(5)
0.121(5)
0.162(5)
0.21(2)
0.22(1)
0.49(1)
0.74(1)
0.97(1)
0.178
0.258
0.345
0.359
0.395
0.441
0.450
0.499
0.542
0.646
0.746
0.812
0.876
-
Br generated by this alternate preparation was removed by
34
reaction with AgOH.
0.002 653
0
0
0
0
0
.003 294
.003 759
.004 604
.005 313
.006 578
A 1.0 M NaOH solution was prepared by diluting a saturated
NaOH solution into He-sparged water. It was stored in
polypropylene bottles under Ascarite II to protect from CO2
absorption and standardized titrimetrically against primary
standard KHP to a phenolphthalein end point. Stock HClO4
solutions were standardized with a NaOH solution. These
solutions were used for pH adjustments where necessary.
Phosphate buffer solutions were prepared from NaH2PO4 and
Na2HPO4. Carbonate buffer solutions were prepared from
NaHCO3 and NaOH. The ionic strength (µ) for all experiments,
except for those where capillary electrophoresis was involved,
was controlled with analytical reagent grade NaClO4 that was
recrystallized from water, redissolved, and standardized gravi-
metrically.
0.007 877
0.010 51
0
0
0
.021 00
.031 50
.041 91
a
Conditions: [Br(I)] ) 0.0525 mM, 0.10 M [PO
NaClO , 25.0 °C. The numbers in parentheses are standard deviations.
The fraction of ClO is defined as the concentration of ClO found
divided by twice the [HOBr]
4 T
] , µ ) 1.0 M
4
2
2
i
.
to form N-chloroethylenediamine that does not interfere with
Methodology and Instrumentation. An Orion model 720A
digital pH meter equipped with a Corning combination electrode
-
the analysis of ClO3 by CE. There also was no interference
from the excess ethylenediamine, as it is a water-soluble
compound that is present as a cation under the conditions of
the separation.
(model 2513) was used in all pH measurements. The electrode
was calibrated through titrations of standard HClO4 with
standard NaOH to correct the measured pH values to p[H ] at
+
Chlorate ion recovery from the reaction between HOBr and
2
5.0 ( 0.1 °C and µ ) 1.0 M. Analysis of the titration data by
-
excess ClO2 was determined by CE. Prior to the analysis, the
35
the method of Gran was followed by a linear regression of
sample solution was placed into a gas collection tower and
degassed with He to remove any dissolved ClO2.
+
+
+
measured pH and p[H ] values. We use p[H ] ) -log [H ]
because values are expressed in molarities rather than in
activities (pH ) -log aH).
Spectrophotometric measurements of the reaction between
-
HOBr and ClO2 were performed on a Perkin-Elmer Lambda
A Hewlett-Packard 3DCapillary Electrophoresis System
equipped with a negative power supply, a variable-wavelength
UV detector, and Hewlett-Packard extended light path capillaries
9
UV/vis/near-IR spectrophotometer used in conjunction with
PECSS (Perkin-Elmer computerized spectroscopy software).
Faster rates of reaction were followed with either an Applied
PhotoPhysics stopped-flow spectrophotometer interfaced to a
Acorn RISC PC using kinetic software (v 4.25) or with a
Dionex-Durrum model D-110 stopped-flow instrument inter-
faced to a Zenith 151 CPU with a Metrabyte DASH-16 A/D
interface card. All reactions were run under pseudo-first-order
(i.d. ) 75 µm, Leff ) 72 cm, Ltot ) 80.5 cm) was used for the
product identification of anions. The capillary was contained
in a Peltier temperature-controlled cartridge maintained at 15.0
(
0.1 °C. A pressure sample injection mode was used. Data
3D
acquisition was performed with the HP CE Chemstation and
software. The run buffer consisted of 0.10 M boric acid and
-
conditions with ClO2 in large excess relative to the initial
5
(
.0 mM sodium borate at pH 8.0. Naphthalenedisulfonic acid
disodium salt from Kodak) (NDS) was incorporated into the
buffer for the indirect detection of anions. Diethylenetriamine
Aldrich) was added to the buffer to reverse the electroosmotic
[HOBr]. The rates of the reaction were monitored at 359 nm
for the appearance of ClO2. Reactions were followed for at
least 4-5 half-lives. Reactions were thermostated at 25.0 (
(
0
.2 °C and were maintained at an ionic strength (µ) of 1.0 M.
flow to allow for the detection of the anions in the samples. As
the separated ions migrate past the detector window, they are
measured as negative peaks relative to the high baseline of the
NDS. By reversing the signal and the reference wavelengths
on the diode array detector of the Hewlett-Packard 3 Capillary
Electrophoresis System, a positive signal was obtained.
Hypochlorite stock solutions were tested for chlorate ion after
treatment of an aliquot of the hypochlorite solution with
The rate constants reported are the results obtained from an
average of 5-10 traces. The observed pseudo-first-order rate
constant is defined by eq 12, where [HOBr]T ) [HOBr] +
-
[
OBr ].
D
-
d[HOBr]T
) kobsd[HOBr]T
(12)
dt
3
6-39
analytical reagent grade (neat liquid) ethylenediamine.
This
The integrated form in terms of the observed chlorine dioxide
is given by eq 13. The fraction of HOBr that is converted to
ClO2 varies with the reaction conditions, but these terms cancel
out in eq 13.
solution was subsequently injected onto the CE system to obtain
-
any signal due to ClO3 . Ethylenediamine was used in 25-
fold excess relative to the original hypochlorite concentration
(
34) Noszticzius, Z.; Noszticzius, E.; Schelly, Z. A. J. Phys. Chem. 1983,
87, 510-524.
[ClO2]∞
(
(
35) Rossotti, F. J. Rossotti, H. J. J. Chem. Educ. 1965, 42, 375-378.
36) Bolyard, M.; Fair, P. S.; Hautman, D. P. EnViron. Sci. Technol. 1992,
ln
) kobsd
t
(13)
(
)
[
ClO ] - [ClO ]
2 ∞ 2 t
26, 1663-1665.
(
(
(
37) Hautman, D. P.; Bolyard, M. J. Am. Water Works Assoc. 1992, 84,
8-93.
38) Bolyard, M.; Fair, P. S.; Hautman, D. P. J. Am. Water Works Assoc.
Results and Discussion
8
Products. The chlorine dioxide formed from the reaction
1993, 85, 81-88.
-
between HOBr and excess ClO2 is used to monitor the kinetics.
39) Gordon, G.; Adam, L. C.; Bubnis, B. P.; Hoyt, B.; Gillette, S. J.;
Wilczac, A. J. Am. Water Works Assoc. 1993, 85, 89-97.
Table 1 gives the yield of ClO2 in terms of the fraction ClO2

4324 Inorganic Chemistry, Vol. 37, No. 17, 1998
Furman and Margerum
-
Table 2. Chlorate Ion Found from the HOBr and ClO
as Determined by CE
2
Reaction
a
p[H+]
[ClO
-] found, M
ClO
-
recovered, (%)
b
3
3
-
-
-
-
5
5
4
4
6
6
7
7
.10
.56
.06
.51
8.2 × 10
9.3 × 10
1.0 × 10
1.2 × 10
54
61
66
79
a
-
Reaction conditions: [ClO
.02 M phosphate buffer (total), and 25.0 °C. Percent is relative to
HOBr]
2
] ) 3.761 mM, [HOBr] ) 0.151 mM,
o
b
0
[
o
.
produced (f ) [ClO2]found/2[HOBr]i) based on the stoichiometry
-
in eq 14. The other product is ClO3 (eq 15).
HOBr + 2ClO2- f 2ClO + Br + OH
-
-
(14)
(15)
2
Figure 1. Chlorite ion dependence of the observed first-order rate
-
constant for the reaction of HOBr with excess [ClO
2
]. Conditions:
, µ ) 1.0
), 25.0 °C. The second-order rate constant determined from
HOBr + ClO2 f ClO - + Br + H
-
-
+
+
[
HOBr]
M (NaClO
the slope is 23 M
o
4 T
) 0.0525 mM, p[H ] 6.67 ( 0.01, in 0.1 M [PO ]
3
4
-
1
-1
s
.
The yield of ClO2 increases with increase of ClO2- concentra-
tion and with acidity. In the presence of excess ClO2 , the ClO2
product is relatively stable and does not decay over the time
intervals of the measurements. The relative yields of ClO2 and
ClO3 depend on competing kinetic pathways, but we will show
that this occurs after the rate-determining steps. Therefore, one
or more intermediate species must be present. Chlorate ion was
confirmed as a reaction product by capillary ion electrophoresis
measurements. The ClO3- yield increases with p[H ], as shown
in Table 2.
-
intermediate. We propose that BrOClO is extremely reactive.
-
-
It can react rapidly with another ClO2 to give 2ClO and Br ,
2
-
or, alternatively, it can react rapidly with OH or H O to give
ClO3 and Br . The ratio of products depends on the ratio of
k3[ClO2 ]/k4[OH ], but neither k3 nor k4 is in the rate expression
2
-
-
-
-
-
because these reactions occur after the rate-determining forma-
-
tion of BrOClO. With HOBrOClO as a steady-state species,
+
the rate expression for the loss of HOBr is given by eq 19,
where HA is any acid.
Kinetics. The observed first-order rate constant is directly
-
-
d[HOBr]T
k [ClO ][HOBr](
∑(k
HA[HA]))
proportional to the concentration of excess chlorite ion (Figure
1
2
)
(19)
-
1
). Despite the fact that two ClO2 are needed per HOBr for
dt
k-1
+
(kHA[HA])
∑
-
ClO2 formation (eq 14), and despite large variations in the yield
of ClO2 as conditions change, the reaction is always first order
+
-
In our studies, HA is H3O , H2PO4 , HCO3 , or H2O. The
expression for kobsd (eq 20) has a first-order dependence in
ClO2 ], does not depend on the k3 or k4 rate constants, but is
highly dependent on the concentrations of acidic buffer species
in accord with our observations.
-
in [ClO2 ]. The kobsd values increase with increased concentra-
+
-
-
tions of [H ], [H2PO4 ], and [HCO3 ]. A mechanism is needed
that accounts for this acid assistance, the first-order dependence
-
[
-
in both [HOBr] and [ClO2 ], and the relative yields of ClO2
-
and ClO3 .
+
Proposed Mechanism. In the proposed mechanism (eqs 16-
[H ]
-
-
k [ClO ]
∑(k
HA[HA])
1
8), a weak complex is formed between HOBr and ClO2 .
1
2
+
HOBr
[
]
[
H ] + Ka
kobsd
)
(20)
k
-
1
k-1
+
(kHA[HA])
-
-
∑
HOBr + ClO2 { k1 } HOBrOClO
(16)
(17)
-
In the proposed mechanism, OBr is not a reactive species. The
is 8.59 under our conditions, so correction
for the [HOBr]/[HOBr]T ratio is needed at higher pH. Equation
HOBr
40
kHA
value of pKa
-
-
HA + HOBrOClO
8 BrOClO + H O A
2
2
0 includes this correction.
k3
The yield of ClO2 (eq 21) depends on the competition
-
-
+
ClO2- 8 2ClO + Br
between the BrOClO reaction with ClO2 and its reaction with
2
fast
-
BrOClO
(18)
H2O or OH (eq 18).
k4
-
-
-
+
{
+ OH fast8 ClO3 + Br + H
-
2
k [ClO ]
3
2
f ) ClO fraction )
(21)
This requires an expansion of the number of valence electrons
around the bromine atom from 8 to 10, in a manner similar to
the bonding in Br3 or in HOBrI . The intermediate given
2
2
H O
OH
-
-
k4 + k4 [OH ] + 2k [ClO ]
3
2
-
- 33
-
Figure 2 shows the dependence of the ClO2 yield on the ClO2
-
in eq 16 indicates O-Br-O bonding in HOBrOClO , but
+
concentration at p[H ] ) 6.67 ( 0.01. Figure 3 shows the
O-Br-Cl bonding is also possible, which would correspond
+
-
p[H ] dependence of the ClO2 yield when the ClO2 concentra-
-
to a HOBrClO2 intermediate. Intermediates of this type have
H O
tion is constant (3.761 mM). These data give values for k4
2
/
3
been proposed by Gordon et al. for the reactions of HOCl with
-
4
OH
3
k3 ) 1.84((0.04) × 10 M and k4 /k3 ) 6.83((0.05) × 10 .
-
-
ClO2 , where HO-Cl-ClO2 precedes the formation of Cl-
ClO2. In our mechanism, HOBrOClO is a steady-state
OH
H O
7
-1
-
The ratio of k4 /k4 , so the OH
2
is 3.71 × 10
M
-
intermediate that undergoes general-acid-assisted reactions to
eliminate water and form BrOClO (or BrClO2) as a metastable
(
40) Gerritsen, C. M.; Gazda, M.; Margerum, D. W. Inorg. Chem. 1993,
32, 5739-5748.

ClO2 and ClO3- Formation from HOBr and ClO2-
Inorganic Chemistry, Vol. 37, No. 17, 1998 4325
-]. The fraction of ClO
is
+
-]
, µ ) 1.0 M
Figure 2. Yield of ClO
defined as the concentration of ClO
[
2
vs excess [ClO
found divided by twice the initial
HOBr]. The dotted line is the least-squares fit obtained from eq 21.
2
2
Figure 3. Yield of ClO
) 3.761 mM, [HOBr]
2
as a function of p[H ]. Conditions: [ClO
2
i
2
o
) 0.0525 mM, 0.10 M [PO
). The dotted line is the least-squares fit based on eq 21.
4
]
T
(NaClO
4
+
Conditions: [HOBr]
o
4 T
) 0.0525 mM, p[H ] 6.67 ( 0.01, 0.10 M [PO ] ,
µ ) 1.0 M (NaClO
4
), 25.0 °C.
+
path dominates above p[H ] 7. In contrast to the strong buffer
dependence found for the rate constants in eq 17, the yield of
ClO2 is not affected by changes in HPO42 or CO3 concentra-
-
2-
tions.
Acid and Buffer Dependence upon kobsd. The dependence
+
of kobsd from p[H ] 5.00 to 9.01 in 0.10 M phosphate or
carbonate buffer can be accounted for in terms of the reaction
-
mechanism (eqs 16-18) and eq 20. Because HOBrOClO is
a steady-state intermediate, we can evaluate the ratio of kHA/
k-1 rate constant, but not the individual rate constants for its
acid-assisted conversion to BrOClO. The resulting expressions
for kobsd in phosphate buffer is given by eq 22 and in carbonate
buffer by eq 23.
-
Figure 4. Effect of H
2
PO
4
concentration on the observed first-order
-
rate constant. Conditions: [ClO
2
] ) 3.761 mM, [HOBr]
mM, p[H ] 5.91-5.92, or 6.39-6.45 in 0.10 M [PO , µ ) 1.0 M
(NaClO ). The dashed line is the nonlinear least-squares fit based on
o
) 0.0525
kobsd
)
+
4 T
]
+
+
k
k [H ]
H PO
[H ]
4
H
2
4
-
-
k1
+
[H PO ] [ClO ]
eq 22. The y-intercepts (O) are the calculated rate constants for the
(
2
4
)
2
+
HOBr
k-1
k-1
[
[H ] + Ka
]
term without phosphate.
+
k
k [H ]
H PO
H
2
4
-
1
+
+
[H PO ]
2
4
k-1
k-1
(22)
kobsd
)
+
+
k
k
HCO3
H O
k [H ]
[H ]
2
H
-
-
k1
+
+
[HCO ] [ClO ]
(
3
)
2
HOBr
+
k-1
k-1
k-1
[
]
[H ] + Ka
+
k
k
H O
k [H ]
HCO
2
H
3
-
1
+
+
+
[HCO ]
3
k-1
k-1
k-1
(23)
The kH O/k-1 term is negligible for the phosphate buffer region
2
and, therefore, is not given in eq 22. The correction term for
the OBr concentration is small for most of the phosphate data
but becomes appreciable for the carbonate data. The depen-
-
-
Figure 5. Effect of the HCO
3
concentration on the observed first-
-
order rate constant. Conditions: [ClO
.0525 mM, p[H ] 8.49-8.52 in 0.10 M [CO
The dashed line is the least-squares fit based on eq 23. The y-intercept
2
] ) 3.761 mM, [HOBr]
o
)
4
+
0
3
]
T
, µ ) 1.0 M (NaClO ).
-
+
dence of kobsd on the [H2PO4 ] concentration at p[H ] 5.92 and
-
6
.42 is given in Figure 4. The [HCO3 ] dependence is shown
in Figure 5, and the p[H ] dependence is given in Figure 6. All
these data were used together in a least-squares fit of the rate
(0) is the calcuated rate constant based on the k /k term.
H
-1
+
+
p[H ] 8.51 ( 0.02 permit us to estimate that kH O/k-1 is
2
-
1
-1
5
-4
constants that gave k1 ) 97(6) M s , kH/k-1 ) 3.1(5) × 10
approximately 2 × 10 . The dashed line in Figures 4 and 5
-1
-1
-1
M , kH PO /k-1 ) 8.3(6) M , and kHCO /k-1 ) 0.064(4) M .
show the fit of eqs 22 and 23 to these rate constants. Three
calculated lines are given in Figure 6 in order to show the
contributions of each acid as the pH changes. The dashed line
for the phosphate-buffered reactions is obtained from eq 22.
2
4
3
It was not possible to obtain a reliable value for kH O/k-1 because
2
the contribution of this term to the reaction rate is very small.
However, the intercept in Figure 5 and the values for kH/k-1 at

4326 Inorganic Chemistry, Vol. 37, No. 17, 1998
Furman and Margerum
Table 3. Summary of Rate and Equilibrium Constantsa
) 96 ( 6 M s-1, pK
-
1
HOBr
) 8.59
b
k
1
k
4
k
4
a
H2O
-4
/k
3
) (1.84 ( 0.04) × 10
M
OH/k
) (6.83 ( 0.05) × 10
pK HA
3
3
1
HA
a
k
HA/k-1, M-
O+
PO
-1.72c
(3.1 ( 0.5) × 10
5
H
H
HCO
3
-
4
-
6.26d
2
8.3 ((0.6)
d
3
9.48e
0.064 ((0.004)
-6 f
H
2
O
15.74
∼3.5 × 10
a
Conditions: µ ) 1.0 M (NaClO
4
), 25.0 °C, pK
w
) 13.60.
e
b
reference 40. c -log(55.5). d Determined for reaction conditions. The
value is from -log(K
f
H O
2
w
/55.5). Value calculated from (kHA
/k-1)/55.5.
Figure 6. Dependence on the observed first-order rate constant as a
+
function of solution p[H ]. Conditions: [HOBr]
[
o
) 0.0525 mM,
] ) 3.761 mM, in 0.10 M total buffer (], phosphate or O,
carbonate), µ ) 1.0 M (NaClO ), 25.0 °C. The calculated nonlinear
least-squares fits based on eq 22 with (- - - H and H ) and (‚‚‚)
-
ClO
2
4
+
-
2
PO
based on eq 23, and for H and H O terms only
4
+
-
+
H , H
2
O, and HCO
3
2
(-‚-).
Scheme 1. Proposed General-Acid-Assisted Path To
Generate BrOClO (or BrClO2)
Figure 7. Brønsted plot for general-acid assistance of the reaction
-
between the steady-state intermediate, HOBrOClO , and HA, where
K
a
is the ionization constant of HA. The slope of the line (R) equals
0
.59 ( 0.01 (eq 24).
data points. Rate constants for the acid-catalyzed decomposition
-
of the complex, HOBrOClO , increase with increasing acid
strength.
1
XClO2 Intermediates. Taube and Dodgen addressed the
connectivity of the intermediate Cl2O2. Their results showed
The dotted line for the carbonate buffered reactions is obtained
that Cl2O2 was not symmetrical (i.e., it is neither Cl-O-O-
-O)
from eq 23. The dashed-dot line shows the contribution of
Cl nor O-Cl-Cl-O) but could be either Y-shaped (Cl-Cl-O
or chain-like (Cl-O-Cl-O). Although many authors imply
Y-shaped structures for Cl O and BrClO , the connectivity is
+
-
-
H3O and H2O if H2PO4 and HCO3 are not present.
Brønsted-Pedersen Relationship. The buffer dependence
upon kobsd is a result of general-acid assistance by HA in the
transfer of a proton combined with H2O-Br bond cleavage in
the step leading to BrOClO formation (eq 17 and Scheme 1).
The ratio of rate constants (kHA/k-1) that we resolved from the
2
2
2
not known in aqueous solution. We have written the metastable
intermediate as Br-O-Cl-O because of some related studies
-
42
between HOCl and BrO . However, the present work cannot
2
-O.
distinguish between Br-O-Cl-O and Br-Cl-O On the
other hand, chloryl chloride (Cl-ClO2) has been prepared in both
noble gas matrixes and in the gas phase, and evidence was found
for a Y-shaped intermediate. Photolysis of matrix-isolated
ClClO2 resulted in the isomers ClOClO and ClOOCl. In the
gas phase, ClClO2 decomposes into ClO2 and Cl2. At room
temperature and a partial pressure of 1 mbar, the half-life of
+
p[H ] and buffer dependence studies show a Brønsted relation-
4
1
ship (eq 24),
4
3
kHA
log
) log G + R log(K q/p)
(24)
a
a
(
)
k p
-1
4
3
ClClO2 is 1 min. The isomer, ClOClO, has been prepared in
where p is the number of equivalent proton sites on the acid
form of the buffer, HA, q is the number of equivalent basic
sites on the conjugate base, A , GA is the Brønsted proportion-
4
4
noble gas matrixes by Jacobs and co-workers. Molecules of
ClOClO, BrOClO, and IOClO have been isolated and character-
-
4
5
ized in argon matrixes. These molecules have been found to
rearrange to XClO2 (X ) Cl, Br, or I) upon exposure to visible
radiation, but ClClO2 and BrClO2 can be isomerized back to
ality constant, and R is the Brønsted coefficient. The R value
can range from 0 to 1 and reflects the degree of proton transfer
in the transition state. Table 3 summarizes the values for kHA/
4
5
HA
the chain form by near-UV radiation.
k-1 and Ka for each acid, HA. The kH O/k-1 value is divided
2
by the molarity of water (55.5 M) to convert it to the same
units as the other acids. Figure 7 is a Brønsted plot with a
slope (R) of 0.59 ( 0.1 based on the H3O , H2PO4 , and HCO3
(
42) Perrone, T. F.; Margerum, D. W. To be submitted for publication.
+
-
-
(43) Muller, H. S. P.; Willner, H. Inorg. Chem. 1992, 31, 2527-2534.
(44) Jacobs, J.; Kronberg, M.; Muller, H. S. P.; Willner, H. J. Am. Chem.
Soc. 1994, 116, 1106-1114.
(45) Johnsson, K.; Engdahl, A.; Kolm, J.; Nieminen, J.; Nelander, B. J.
Phys. Chem. 1995, 99, 3902-3904.
(
41) Bell, R. P. The Proton in Chemistry, 2nd ed.; Cornell University:
Ithaca, NY, 1973; p 198.

ClO2 and ClO3- Formation from HOBr and ClO2-
Inorganic Chemistry, Vol. 37, No. 17, 1998 4327
Recently, Guha and Francisco46 have calculated geometries,
vibrational spectra, and relative energies of BrOClO, BrClO2,
ClBrO2, ClOBrO, and other XBrO2 isomers in the gas phase.
These calculations show that BrClO2 is more stable than
BrOClO by only 3.7 kcal mol . Solvation effects are not
known, and in aqueous solution the relative energies of these
two forms could shift significantly. The above studies show
that either BrOClO or BrClO2 is a reasonable intermediate for
the proposed mechanism.
Scheme 2. Proposed Reactions of BrOClO To Give ClO2
and ClO3
-
-
1
Reaction Pathways. The general-acid-assisted dependence
that we have observed and assigned to eq 17 in the mechanism
requires that proton transfer occur during the process of
generating BrOClO (or BrClO2). If proton transfer occurred
-
in a prior equilibrium to convert HOBrOClO to H2OBrOClO
before the rate-determining step, the reaction rate would depend
+
only on the H concentration, as opposed to the experimental
dependence on general acid (HA) concentrations. Therefore,
we suggest the transition state shown in Scheme 1, where the
oxygen-bromine bond breaks as proton transfer occurs. The
Brønsted value of 0.59 indicates that a significant degree of
proton transfer occurs in the transition state, leading to BrOClO
(or BrClO2) formation.
Scheme 2 shows the proposed pathways for the conversion
of BrOClO to products. Under our conditions, the reactions of
-
-
BrOClO with ClO2 and with OH occur after the rate-
determining step. Hence, k3 and k4 do not contribute to the
-
H2O
OH
-
kobsd values, but the ratios of k3[ClO2 ]/(k4
+ k4 [OH ])
-
determine the yields of ClO2 compared to ClO3 . As seen in
Scheme 2, OH attack (k4 path) at the chlorine atom of BrOClO
will lead to the elimination of Br and the formation of ClO3 .
Valdez-Aguilera et al.15 studied the Br2 reaction with ClO2
-
two ClO2 molecules. Alternatively, ClO2- attack on bromine
-
-
-
in BrOClO could give a (OClOBrOClO) intermediate that
-
-
breaks up to form ClO + Br + ClO .
2
2
7
,8
and reported a BrClO2 intermediate that reacts rapidly with
The observed rate constants for ClO2 formation in the
-
3
-1 -1
-
ClO2 with a rate constant of 2.94 × 10 M s . This value
reactions of HOCl with ClO2 appear to be 1-2 orders of
-
-
meets our requirement that k3[ClO2 ] is much larger than
magnitude smaller than those for HOBr with ClO2 under
k3∑kHA[HA]/(1 + ∑kHA[HA]). A self-exchange rate constant
similar conditions. The buffer dependencies for the HOCl
reactions have not been resolved, so it is difficult to make exact
comparisons.
-
47
4
-1 -1
for ClO2/ClO2 has been evaluated to be 3.3 × 10 M s .
It seems unlikely that the less favorable electron-transfer reaction
-
would occur rapidly between BrClO2 and ClO2 to give ClO2
-
-
-
Acknowledgment. This work was supported by Environ-
mental Protection Agency Grant 821245 and by National
Science Foundation Grant CHE-96-22683.
and BrClO2 (followed by breakup of BrClO2 to Br and
ClO2). We propose a Br(OClO)2- intermediate that dissociates
-
-
into Br + 2ClO2. The intermediate might be formed by ClO2
attack on chlorine in BrOClO, leading to Br elimination and
-
Supporting Information Available: Listings of kinetic data and
2
yield of ClO (5 pages). Ordering information is given on any current
masthead page.
(
(
46) Guha, S.; Francisco, J. S. J. Phys. Chem. A 1997, 101, 5347-5359.
47) Awad, H. H.; Stanbury, D. M. J. Am. Chem. Soc. 1993, 115, 3636-
3642.
IC980262Q