A new method for measuring carbonate radical reactivity toward pesticides

Article abstract of DOI:10.1002/etc.5620190605

The carbonate radical is produced from the scavenging of the hydroxyl radical by carbonate/bicarbonate and is strongly electrophilic toward electron-rich compounds such as aromatic amine and sulfur-containing compounds. Carbonate radical reactivity (CO

Full text of DOI:10.1002/etc.5620190605

Environmental Toxicology and Chemistry, Vol. 19, No. 6, pp. 1501–1507, 2000
2000 SETAC
Printed in the USA
᭧
0730-7268/00 $9.00
ϩ .00
A NEW METHOD FOR MEASURING CARBONATE RADICAL REACTIVITY
TOWARD PESTICIDES
J
IPING
HUANG and SCOTT A. MABURY*
80 St. George Street, Department of Chemistry, University of Toronto, Toronto, Ontario M5S 3H6, Canada
(
Received 19 July 1999; Accepted 6 October 1999)
Abstract—The carbonate radical is produced from the scavenging of the hydroxyl radical by carbonate/bicarbonate and is strongly
electrophilic toward electron-rich compounds such as aromatic amine and sulfur-containing compounds. Carbonate radical reactivity
• Ϫ
(CO3 ) was measured by setting up a competition between a chemical of known reactivity and a compound with unknown reactivity
• Ϫ
toward CO3 . The carbonate radical was produced using a novel method whereby KOONO (potassium peroxonitrite) was dissolved
•
Ϫ
• Ϫ
• Ϫ
into 0.01 NaHCO3 solution initially producing a HO , which was rapidly scavenged by HCO3 , yielding CO3 . CO3 production
• Ϫ
was monitored using a stopped-flow spectrometer at the absorption band for CO3 of 600 nm. Results using this competition
kinetic method indicated the second-order rate constants for substituted anilines corresponded to reference values with a relative
ϩ
error less than 10%. Using
␴
values, the rate constants of substituted anilines at pH 9.0 correlated well (R2
ϭ
0.98) with the
para
Hammett equation yielding ␳ ϭ Ϫ0.89. To measure the reactivity of pesticides toward the carbonate radical,
p-nitroaniline (PNA)
was selected as the standard competitor. Of the nine pesticides tested, fenthion was the most reactive, followed by phorate, with
rate constants above 107/M/s. Fluometuron and atrazine were of intermediate reactivity, and methyl parathion was one of the least
reactive.
Keywords—Carbonate radical
Hydroxyl radical
Competition kinetics
Potassium peroxonitrite
Pesticides
INTRODUCTION
duced by flash photolysis of a sodium persulfate solution to
form carbonate radicals [7]. Additionally, the carbonate radical
has been generated through photolytic decomposition of the
inorganic complex Co(NH3) 4CO3 via charge transfer to metal
[2].
The persistence of chemical pollutants in natural waters is
determined by the ability of these waters to cleanse themselves
of xenobiotics. When sunlight falls on natural waters, it drives
a complex series of reactions that produce the oxidants re-
sponsible for the self-cleansing capacity of the water. Hy-
droxyl, hydroperoxyl/superoxide, organoperoxyl, and carbon-
ate radicals as well as singlet oxygen, aqueous electrons, and
triple states contribute to systematic cleansing by reaction with
aquatic pollutants, which they ultimately degrade [1]. How-
ever, little attention has been paid to the carbonate radical
ϩ
The carbonate radical is a powerful oxidant with a one-
electron reduction potential of 1.59 volts (V) at pH 12 [8] and
is capable of the one-electron oxidation of several amino acids
and proteins [3,9,]. In pulse radiolysis and flash photolysis
experiments, the carbonate radical has been shown to react
rapidly with electron-rich aromatic compounds such as anilines
and phenols, primarily as an electron acceptor [3]. Second-
• Ϫ
(CO3 ) reactions with organic pesticides, especially the mea-
surement of reaction rate constants.
The carbonate radical has been identified as a transient
species with a broad optical absorption in the visible range,
order rate constants are low (
compounds but higher (
106–107/M/s) for sulfur compounds
k Ͻ
105/M/s) for simple aliphatic
k
ϭ
such as methionine [10]. Aromatic compounds show variation
in the reactivity depending on the nature of the substituent on
the aromatic ring. The reaction rate constant is high for indole
with
␭
at 600 nm. It is possible to monitor the formation
max
• Ϫ
and disappearance of CO3 in the 500 to 700 nm range through
flash photolysis, which can be used to determine the rate con-
and its derivatives such as tryptophan (k Ͼ
108/M/s) [11]. The
• Ϫ
rate constants have been measured for the reaction of carbonate
radicals, generated in the flash photolysis of aqueous solution
stants of CO3 [2]. The carbonate radical is in equilibrium
with its conjugate acid (bicarbonate radical) with pKa 9.6 [3]
or pKa 7.9 [4]; however, no change in optical and electrore-
flectance spectra have been observed. At low ionic strength,
there is little dependence of the rate constant on pH, suggesting
that the two forms of the radical have similar intrinsic reac-
tivities [5].
There are several ways to produce the carbonate radical as
a secondary radical. The most popular method is through the
reaction of hydroxyl radicals with either carbonate or bicar-
bonate ions. Photolysis of H2O2 forms hydroxyl radicals, which
then react with carbonate anions to produce carbonate radicals
ϩ
of Co(NH3) 4CO3 , with a series of substituted benzene deriv-
atives. The rate constants at pH 7, ranging from 3
ϫ
103/M/
s for benzene to 1.8
ϫ
109/M/s for N,N-dimethylaniline, cor-
ϩ
related well with the Hammett equation using
␴
values for
para
electrophilic substitution, yielding ␳ ϭ Ϫ3.6 [3]. The absolute
value of is higher than that of the hydroxyl radical toward
␳
benzene and its derivatives, with ␳ ϭ Ϫ0.41 [12], which in-
dicates the lower reactivity and higher selectivity of the car-
bonate radical. The rate constants toward
p-substituted anilines
have also been determined by the flash photolysis technique,
• Ϫ
ϩ
para
[6]. Carbonate anions can also react with SO4 radicals pro-
which yields ␳ ϭ Ϫ1.0 when using
value as observed for the reaction of the carbonate radical
with substituted phenols [14].
Under environmental conditions, carbonate radicals may
␴
values [13], the same
␳
* To whom correspondence may be addressed
(smabury@alchemy.chem.utoronto.ca).
1501

1502
Environ. Toxicol. Chem. 19, 2000
J. Huang and S.A. Mabury
Fig. 1. Chemical structures of pesticides used in this investigation.
react rapidly with aromatic pesticides such as monuron or the
carcinogen benzidine, thus influencing their fate in the envi-
ronment [6]. Carbonate radicals may oxidize sulfur-containing
pesticides, thus limiting their persistence; it may also play an
important role in the conversion of organophosphate insecti-
cides to the more toxic oxon derivative in natural waters [11].
Finally, it may contribute to the cycle of naturally occurring
sulfur compounds in natural waters.
bonate with second-order rate constants of 4.2
ϫ
108/M/s and
ϫ
107/M/s to yield CO3 [11]. The advantages of KOONO
• Ϫ
1.5
•
as a HO source include the stability of the solid solution and
• Ϫ
the lack of light required for initiation of CO3 production.
The peroxinitrite can be quantified by dissolution in 0.1 M
NaOH using ⑀302 ϭ 1,670/cm/M [15]; this quantitation permits
reproducible amounts of carbonate radicals to be generated in
subsequent experiments.
Methods utilizing ultraviolet light may complicate kinetic
analysis for those compounds that are also photolabile due to
concomitant photo-degradation. King et al. [15] reported a
solid-state hydroxyl source, KOONO (potassium peroxonitrite,
which was generated by photolysis of KNO3 through a 254
nm ultraviolet lamp), that, on directly dissolving in water,
generated hydroxyl radicals [15]. The addition of KOONO to
By utilizing competition kinetics with Fenton’s reaction for
the production of hydroxyl radicals, the relative reaction rates
of 28 constituents of a PCB mixture were measured [16]. Ma-
bury and Crosby also used competition kinetics to estimate
the reactivity of 10 organic pesticides toward hydroxyl radicals
with p-nitroso-N,N-dimethylaniline as the competitor [17]. A
new competition kinetic method for measuring the reactivity
of pesticides toward the carbonate radical has been developed
in this study by using KOONO as a specific source of hydroxyl
radicals to yield the carbonate radicals.
a solution buffered at pH 7.0 forms pernitrous acid (pKa
ϭ
6.8). The pernitrous acid formed undergoes homolytic fission
•
•
to form HO and NO2 as
Our objective was to develop a method to generate the
carbonate radical without light and measure the reactivity of
the carbonate radical toward a number of commonly used sul-
fur-containing and aromatic pesticides (Fig. 1) that have wide
Ϫ
ϩ
•
•
ONOO
ϩ
H
→
ONOOH
→
NO2
ϩ
HO
(1)
By adding KOONO solution to carbonate or bicarbonate,
the HO subsequently reacts with excess carbonate or bicar-
•

New method for measuring carbonate radical reactivity
Table 1. HPLC conditions for competitor analysis
Detection
Environ. Toxicol. Chem. 19, 2000
1503
pound in order to determine its reactivity toward carbonate
radical.
If the competition were simply between the probe and stan-
dard competitor, it follows that
wavelength
(nm)
Mobile phase
(H2O : CH3CN)
Retention time
(min)
Competitora
So
S
k
p[probe]
DMA
PCA
PNA
254
240
380
60:40
50:50
45:55
7.9
6.1
5.5
ϭ
1
ϩ
(2)
k
c[competitor]
where So represents the rate of loss of the competitor with no
probe, represents the rate of loss when the pesticides are
a DMA
nitroanaline.
ϭ N,N-dimethylanaline; PCA ϭ p-chloroaniline; PNA ϭ p-
S
present in the system, and kc and kp are the rate constants at
which carbonate radicals react with the competitor and pes-
ticide, respectively. If the system is simply a competition be-
tween the competitor and probe, then a plot of So/S versus
(probe)/(competitor) produces a straight line with a slope of
use in agriculture. We hypothesized that the carbonate radical
would contribute to limiting their overall persistence in natural
waters.
kp/kc. The validity of the method overall can be tested by the
linearity of this relationship. An absolute value can be cal-
culated for kp based on the second-order rate constant for the
different competitors.
MATERIALS AND METHODS
Chemicals
All reagents were of reagent grade or purer. N,N-dimethyl-
A typical test solution consisted of 1 to 2 ␮M of competitor
aniline,
trophenol, phenol,
isole, and methyl phenyl sulfoxide were obtained from Aldrich
Chemical (Oakville, ON, Canada). 4-Bromoaniline and -an-
p
-chloroaniline,
p
-nitroaniline,
N-methylaniline, p-ni-
and an appropriate concentration of probe in 20 ml 0.01 M
NaHCO3 solution. The solution was mixed thoroughly with a
magnetic stirrer, then injected into the HPLC for analysis of
the peak area. A second solution was made of 150 mg
KOONO/KNO3 weighed accurately and quickly dissolved in
0.6 ml 0.1 N NaOH solution; the high pH stabilized the per-
p
-toluidine,
p
-amino benzoic acid, thioan-
p
isidine were purchased from Acros Organics (Pittsburgh, PA,
USA). Aniline was obtained from BDH Chemicals (Toronto,
ON, Canada). The pesticides flumeturon, benthiocarb, atrazine,
hexazinone, fenthion, malathion, and methyl parathion were
purchased from Chem Service (West Chester, PA, USA); pho-
rate and thiazafluron were obtained from Riedel-deHaen (Su-
pelco, Oakville, ON, Canada) and Pestanal (Caledon Labs,
Georgetown, ON, Canada), respectively. All the above re-
agents were dissolved in high performance liquid chromatog-
raphy (HPLC) grade water; all solvents purchased from Cal-
edon were HPLC grade. Potassium peroxonitrite (KOONO)
was generated by photolysis of ground KNO3 under a 254-nm
ultraviolet lamp for 10 h; the final solid material was a mixture
of KOONO and KNO3.
oxonitrite ion. After adding 100 ␮l of a KOONO/KNO3/0.1 N
NaOH solution to all solutions, the reaction mixture was stirred
for 3 min and subsequently analyzed by HPLC. The 3-min
period was chosen to allow adequate time for the subsequent
HPLC injection and resulted in high reproducibility between
measurements; early method development indicated the re-
action was complete in much less time. The pH value of the
mixed solutions was approximately 9.0. At this pH, the yield
of hydroxyl radical was less than 5% [21], and the corre-
sponding quantity of carbonate radical produced from quench-
ing by bicarbonate ion would be quite small. Pseudo–first-
order kinetic analysis resulted in a rate constant for each
(probe)/(competitor) ratio. These data were transformed into
Analytical methods
plots of (probe)/(competitor) versus So
slope of kp c. Each measurement was replicated at least twice,
and the data was highly reproducible ( 98%).
Substituted anilines with PCA or PNA as competitor;
/S, which resulted in a
N,N-dimethylaniline,
p-chloroaniline, and p-nitroaniline
/
k
were selected as model competitors and were monitored by
HPLC using an Alltech Econosphere reverse-phase 25-cm C-
18 column (5 micron), aqueous acetonitrile as the mobile phase
under isocratic conditions, and a detection wavelength appro-
priate for each competitor; the HPLC system was a Waters
600 pump and 486 tunable absorbance detector. For phenol,
the pH of the aqueous phase was modified to pH 2.7 with
acetic acid. The HPLC analysis conditions are illustrated in
Table 1.
Ͼ
N
-
methyl aniline, phenol, and thioanisole with DMA, PCA, and
PNA as competitor, respectively; and nine pesticides with PNA
as competitor were selected to measure the rate constant in
this system.
Stopped-flow kinetic experiment
In order to confirm the formation of carbonate radicals in
our competition kinetic system, a Hi-Tech Scientific stopped-
flow spectrophotometer (Varian, Walton-on-Thames, UK) was
used. By making 0.2 M NaHCO3 solution and KOONO/NaOH
solution in two syringes, the mixing solution with an equal
volume was detected at a wavelength of 600 nm. After rapid
mixing of the two solutions, the absorption signal of the car-
bonate radical at 600 nm was observed.
Competition kinetics
We adapted the competition kinetics system [18] by em-
ploying three kinds of competitors, N,N-dimethylaniline
(DMA),
loss of which was monitored by HPLC. The rate constants of
carbonate radical to competitor are 1.8
109/M/s, 4.3 108/
M/s, and 7.3
107/M/s for DMA, PCA, and PNA, respectively
p-chloroaniline (PCA), and p-nitroaniline (PNA), the
ϫ
ϫ
ϫ
[19]. These three competitors can be used to measure the rate
constants of probes to carbonate radical between 105/M/s and
1010/M/s. A particular strength of this method is the require-
ment to only monitor the model competitor while measuring
the reactivity of each probe compound. This precludes the
necessity of developing a new analytical method for each com-
RESULTS AND DISCUSSION
Formation of the carbonate radical
Previous studies have shown that the carbonate radical is
formed through the reaction with the hydroxyl radical in high
carbonate water. Carbonate radicals decay with a second-order

1504
Environ. Toxicol. Chem. 19, 2000
J. Huang and S.A. Mabury
Table 2. Rate constantsa and Hammett constantsb for substituted
anilines
KOONO solution should be kept highly basic to maintain its
stability and reproducibility. The ideal pH value of the final
solution for making carbonate radicals was approximately 9.0.
Rate constant
Reference kc
Ϫ1 Ϫ1
Ϫ
1
Ϫ
1
ϩ
Substituent
(M
s
)
␴
(M
s )
para
Nitrogen dioxide radicals in our system
-OCH3
-CH3
-H
-Cl
-Br
1.9
9.1
4.6
4.3
4.1
1.9
7.3
ϫ
ϫ
ϫ
ϫ
ϫ
ϫ
ϫ
109
108
108
108
108
108
107
Ϫ
Ϫ
0.78
0.31
0.0
0.11
0.15
0.42
0.79
9.1
5.0
4.3
3.8
2.0
7.3
ϫ
ϫ
ϫ
ϫ
ϫ
ϫ
108c
108d
108d
108d
108c
107d
The carbonate radical as a secondary free radical is formed
through reaction with the hydroxyl radical. In our system, the
hydroxyl radical was formed through the homolytic decom-
position of HOONO through Equation 1. At the same time,
nitrogen dioxide radicals were also formed, which have been
found to be stable in the gas phase but unstable in aqueous
solution [22]. The one-electron reduction potential of nitrogen
dioxide is 1.03 volts [22], and this value is lower than the
value of the carbonate radical of 1.59 volts. So it is impossible
for nitrogen dioxide to react with carbonate ions to form the
carbonate radical. Nitrogen dioxide undergoes a rapid dispro-
portionation and proceeds in the following two steps with rate
-COOH
-NO2
a Data from our competition kinetic method at pH 9.0.
b Hammett constants from [13].
c [18].
d [13].
rate constant of 2.0
ϫ
107/M/s at pH 12.8 in the absence of a
scavenger [20]; their fate in natural waters depends on their
steady-state concentration relative to the concentration of or-
ganic, and possibly inorganic, scavengers.
constants of 9.0
solution [23]:
ϫ
108/M/s and 1.0
ϫ
103M/s in aqueous
•
2 NO2
→
→
N2O4
(3)
In this study, carbonate radicals were produced in 0.01 M
NaHCO3 by adding a solution of KOONO/KNO3/0.1 N NaOH.
At pH 9.0, peroxonitrite ions can still acidify to pernitrous
acid HOONO. As a result, the hydroxyl radical produced from
HOONO homolysis would quickly be scavenged by bicarbon-
ate solution to form bicarbonate and carbonate radicals. At the
fixed pH of our system, the quantity of hydroxyl radicals pro-
duced should be reproducible based on an equivalent amount
of KOONO/KNO3 added to the system. The excess of bicar-
bonate, relative to probe/competitor, in the system assured the
hydroxyl radicals produced were scavenged to form carbonate
radical rather than react directly with the probe or competitor.
Evidence for this condition being met is available from the
high linearity observed for each probe tested (Fig. 3) and ex-
cellent agreement between second-order rate constants mea-
sured for substituted aromatic amines and literature values
(Table 2). Further evidence was obtained by directly observing
the formation of the carbonate radical. The pH-dependent yield
of hydroxyl radical products from peroxonitrite was reported
[21]. It was proposed that the pKa of 6.8 corresponded to the
protonation of cis-peroxonitrite anion whereas the loss of hy-
droxyl radical-like reactivity with a pKa of 8.0 corresponded
to that of the trans-peroxonitrite anion. Therefore, the pKa of
8.0 of trans-peroxonitrous acid was favorable for the produc-
tion of the hydroxyl radical in comparison with pKa of 6.8 of
peroxonitrous acid [15] in our pH 9.0 system. The formation
of bicarbonate and carbonate radicals can be identified by the
stopped-flow spectrophotometer with the detecting wavelength
set at 600 nm. By putting the 0.02 M NaHCO3 and KOONO
solution in dilute NaOH in two syringes, the absorption signal
of the carbonate radical at 600 nm appeared immediately. The
concentration of NaHCO3 was about 0.02 M and the NaOH
was about 0.005 N, which makes the final pH of the mixing
solution 9.0. It was consistent with the pH value in our com-
petition kinetic system. If the NaOH were 0.1 N, no absorption
signal appeared after mixing with 0.02 M NaHCO3 solution.
Since the rate of formation of pernitrous acid is proportional
Ϫ
Ϫ
ϩ
N2O4
ϩ
H2O
NO3
ϩ
NO2
ϩ
2H
(4)
The rate is faster than carbonate radical decay at 2.0
M/s. Additionally, nitrogen dioxide can also react with the
carbonate radical with a rate constant of 1.0
109/M/s, which
can decrease the amount of nitrogen dioxide radicals as well
as carbonate radicals. The amount of nitrogen dioxide would
thereby be lower in comparison with that of the carbonate
radical. Aniline and phenoxide were selected as electron-rich
compounds to determine whether nitrogen dioxide was formed
ϫ
107/
ϫ
in our system. By using 2
petition system at pH 9.0,
␮
M aniline and phenol in the com-
-nitroaniline and -nitrophenol,
p
p
respectively, were detected through HPLC as products. About
80% of the aniline and the phenol reacted, while 12% of the
aniline formed
p-nitroaniline and 10% of the phenol formed
p
-nitrophenol, which was presumably through reaction with
the nitrogen dioxide radical. These results indicate the car-
bonate radical was the most important reactant with the se-
lected electron-rich compounds such as aniline and phenoxide.
For sulfur-containing compounds, especially the aliphatic com-
pounds such as phorate, the rate constant toward nitrogen di-
oxide should be slower than that for the carbonate radical due
to the lower one-electron reduction potential of nitrogen di-
oxide and the lack of the aromatic ring for electrophilic sub-
stitution. As a result, the relative error of this method for such
compounds would presumably be small.
Competition kinetic method
In order to validate this method, aromatic compounds with
known rate constants determined using other methods were
selected as probes to measure the rate constant in our com-
petition kinetic system. The measured rate constant of
yl aniline was about 2.5
109/M/s with DMA as competitor,
which is close to the reference value of 1.8
109/M/s [19].
The rate constant of phenol was found to be 1.2
108/M/s
N-meth-
ϫ
ϫ
ϫ
ϩ
with PNA as competitor at pH 9.0. Phenol is an electron-rich
aromatic compound, especially at high pH. The rate constants
of carbonate radicals with phenol vary with the pH, which is
due to the ionization of the phenolic OH group. Because the
to the concentration of H , there is very limited pernitrous
acid at pH 13. If KOONO were in neutral water, there still
was no absorption signal from mixing with 0.02 M NaHCO3
because the KOONO likely decomposed before reacting with
the NaHCO3. The formation of the carbonate radical in our
system depends strongly on the pH value of the solution. The
p
Ka of phenol is 10.2, there was a rapid increase in the rate
constant between pH 9 and 11. Our measured rate constant

New method for measuring carbonate radical reactivity
Environ. Toxicol. Chem. 19, 2000
1505
Fig. 3. Typical results indicating pesticide reactivity toward carbonate
radicals.
Fig. 2. Hammett plot of substituted aniline reactivity toward carbonate
radicals.
step for the reaction is the removal of an electron, most likely
from the basic nitrogen atom by the carbonate radical, to form
an anilino radical [6]. The formed anilino radical can be sta-
(1.2
M/s at pH 7.0 and 1.2
Thioanisole was selected as a model sulfur-containing ar-
omatic compound, with the measured rate constant of 2.3
107/M/s toward the carbonate radical in our competition kinetic
system. One of the oxidized products of thioanisole is methyl
phenyl sulfoxide, which was identified through HPLC and
comparison of a standard, indicating that the reaction site is
presumably on the sulfur atom. However, the rate constant of
methyl phenyl sulfoxide could not be determined with PNA
as competitor even though probe concentration was 400 times
higher than that of PNA. This indicated that the rate constant
of methyl phenyl sulfoxide was lower than 104/M/s.
ϫ
108/M/s) was between the reference values 2.2
109/M/s at pH 11.2 [14].
ϫ
107/
ϫ
bilized through conjugation. The
␳ value of Ϫ0.89 is similar
ϫ
to that of the reference value, which provides good evidence
that this method is an accurate way to measure the rate constant
of the carbonate radical without any light interference.
Pesticide reactivity toward carbonate radicals
Many of the pesticides were sulfur-containing compounds,
and PNA was the model competitor in our competition kinetics
system because the rate constant of most sulfur-containing
compounds was around 107/M/s [2]. A requirement of the com-
petition kinetic system was that competition for carbonate rad-
icals be simply between PNA and the probe pesticide. The
PNA was stable in aqueous solution and was detected at 380
nm without interference from the selected pesticide, which
typically had absorption below 300 nm.
Based on the reactivity of pesticides toward the carbonate
radical and on their solubility, they were divided into two
groups to be measured more accurately. The two groups were
based on concentration ratio, with the concentration ratio of
pesticide to PNA being 0 to 20 in the more reactive group and
0 to 200 in the less reactive group. Following Equation 2,
typical results for five probe pesticides are shown in Figure
The rate constants of the substituted anilines
line, -chlorooaniline, -methylaniline, -methoxyaniline, an-
iline, and -amino-benzoic acid were measured with PCA and
p-bromoani-
p
p
p
p
PNA as probes. The measured and reference values as well as
the substituent constants are shown in Table 2; an excellent
correlation (R
2 ϭ 0.992) between measured and reference val-
ues was obtained, with a slope of 0.9997.
Many types of organic reactions are sensitive to electron
density at the site on the nitrogen atom and can be explored
using the Hammett equation relating reaction rates and struc-
ture for para- or meta-substituted benzene derivatives. The
carbonate radical reactivity toward substituted aniline was de-
scribed previously by the Hammett equation, with good linear
relationship yielding ␳ ϭ Ϫ1.0 [13]. Based on our measured
rate constant, the Hammett plot of the logarithm of the ratio
3. The steeper the slope (kp/kc), the more reactive the pesticide
toward the carbonate radicals; second-order rate constants are
shown in Table 3.
Fenthion was the most reactive toward carbonate radicals,
followed by the sulfur-containing pesticide phorate. The mea-
of the rate constants for the substituted anilines versus their
ϩ
substituent constant,
␴
, is shown in Figure 2. The Hammett
sured rate constant of fenthion (2.0
the rate constant of thioanisole (2.3
ϫ
ϫ
107/M/s) was close to
107/M/s). The electron
para
plot shows that the reaction constant is ␳ ϭ Ϫ0.89, with a
good correlation coefficient of 0.98. The negative values of
indicate the carbonate radical is strongly electrophilic and the
reaction rate depends on electron density in the ring. For
substituted anilines, the best correlation was obtained with
␳
density on the sulfur atom of fenthion would be the most likely
site of the oxidation reaction with carbonate radicals. Phorate
has two electron-rich sulfur atoms on the aliphatic chain, which
makes it highly reactive with carbonate radicals. This result
is consistent with the reference rate constants of aliphatic sul-
fur-containing compounds ranging from 106 to 107/M/s.
Fluometuron and atrazine had intermediate rate constants
106/M/s and 4.0 106/M/s, and these values are
of 4.2 ϫ ϫ
relatively lower than aniline derivatives. The attack of the
carbonate radical is presumably on the site of the nitrogen
p
-
ϩ
␴
, while other values proved unsatisfactory. It was also
para
found that the ionization potentials of
p
-substituted anilines
ϩ
correlate well with
␴
[24]. Thus, it is likely the rate con-
para
stants are influenced by the ionization potential of the amine,
suggesting possible electron transfer from aniline to the car-
bonate radical [13]. It is probable that the rate-determining

1506
Environ. Toxicol. Chem. 19, 2000
J. Huang and S.A. Mabury
Table 3. Measured rate constants for pesticides toward the carbonate
radical
droxyl radical is more reactive and less selective than the
carbonate radical.
In the near-surface natural waters, the steady-state concen-
Measured rate constant
Theoretical
half-life,a t1/2
(days)
Ϫ
Ϫ
tration of the carbonate radical is about 10 13 to 10 14 M, which
Ϫ
1
Ϫ
Ϫ
1
Ϫ
Pesticide
(M
s
1), km
(M
s
1), km
is much higher than that for the hydroxyl radical, with 1.5
ϫ
Ϫ
Ϫ
10 18 to 5
ϫ
10 16 M [25], especially in carbonate-rich waters
Fenthion
Phorate
Fluometuron
Atrazine
Malathion
Thiazafluron
Benthiocarb
Hexazinone
Methyl parathion
2.0
1.2
4.2
4.0
8.9
3.6
2.8
2.4
2.0
ϫ
ϫ
ϫ
ϫ
ϫ
ϫ
ϫ
ϫ
ϫ
107
107
106
106
105
105
105
105
105
8.0
13
38
[6]. Even though the rate constant of the carbonate radical was
lower, it may still play an important role by limiting the per-
sistence of reactive pesticides given the higher steady-state
concentration. Suppose that the steady-state concentrations of
6.1
ϫ
106 b
40
180
446
573
685
802
• Ϫ
Ϫ14
the carbonate radical [CO3
] is equal to 5.0 ϫ 10 M in
ss
sunlit near-surface natural waters. The pseudo–first-order rate
• Ϫ
constant for the reaction of the pesticide is
kpesticide ϭ [CO3 ]
ss
ϫ
k
m, where km is the measured second-order rate constant of
Ϫ
a Based on a steady-state concentration of carbonate radical [CO3· ]ss
the pesticide. The resulting theoretical half-life of each pes-
ticide is shown in Table 3 without considering direct photol-
ysis, hydrolysis, or other oxidants such as the hydroxyl radical,
singlet oxygen, or other radicals. The theoretical surface half-
life of fenthion is about 8.0 d, while that of methyl parathion
is about 802 d in natural waters. In comparison, the steady-
Ϫ
14
ϭ
5 ϫ 10 M.
b From data of Larson and Zepp [6]; reaction rate presumably included
direct photolysis.
atom of those compounds. The electron density on the nitrogen
atom of fluometuron is decreased by the adjacent carbonyl
group, and the electronegative nitrogen atoms on the hetero-
atom aromatic ring decrease electron density on the nitrogen
•
state concentration of the hydroxyl radical [HO ]ss varies con-
siderably, but a typical value would be around 1.0
Ϫ
ϫ
10 17 M
in sunlit near-surface natural waters [25]. Taking the measured
second-order rate constant for atrazine and hexazinone (8.2
108/M/s and 6.2 108/M/s, respectively [17]), the near-surface
half-life for reaction with hydroxyl would be 978 and 1,294
d, respectively. These are longer than the values of 40 and
685 d for reaction with the carbonate radical. These calcula-
tions indicate the carbonate radical likely will contribute to
the degradation of highly reactive pesticides that are not ef-
ficiently removed via other pathways.
ϫ
atom of the
N-isopropyl group of atrazine. From the pseudo–
ϫ
first-order rate constant of atrazine toward carbonate radicals
using Larson and Zepp’s method [6], the second-order rate
constant of atrazine toward carbonate radicals was obtained
as 6.1
ϫ
106/M/s. This value is calculated from the pseudo–
first-order rate constant with carbonate radicals, and it may
include the contribution of direct photolysis. Our value is low-
er; however, it does not include any direct photodegradation.
Malathion has only one sulfur atom on the aliphatic carbon
chain, and the diester group would decrease the electron den-
sity on that sulfur atom; the rate constant was lower than that
for phorate with two sulfur atoms and an electron-donating
ethyl group. This indicates that the number of sulfur atoms on
the carbon chain can influence the rate constant toward car-
bonate radicals.
CONCLUSIONS
This study demonstrated that carbonate radicals can be gen-
erated through a solid hydroxyl radical source without ultra-
violet light interference. The formation of carbonate radicals
was detected by stopped-flow spectroscopy at 600 nm. The
new competition kinetic method was used to measure the rate
constant of carbonate radicals toward a suite of chemical
probes using a model competitor; competition kinetics require
only the monitoring of the model competitor for each unknown
probe being measured, thus greatly simplifying the analysis.
The measured rate constants of selected compounds were close
to reference values; for substituted anilines, the reaction con-
The other pesticides, i.e., thiazafluron, benthiocarb, hexa-
zinone, and methyl parathion, showed low reactivity toward
carbonate radicals. The lone pair electron of the sulfur atom
in thiazafluron maintains the aromatic characteristic of 6 e on
the heterocycle, so the reactivity was decreased compared with
the aliphatic sulfur-containing compounds. The lower reactiv-
ity of hexazinone could be due to the reduced presence of
olefinic character. The electron-withdrawing effect of chlorine
and carbonyl groups connecting with sulfur contributes to the
low reactivity observed for benthiocarb. Compared with fen-
thion, the strong electron-withdrawing effect of the nitro group
gives methyl parathion the lowest reactivity, with a second-
stant
␳ was Ϫ0.89. Of nine pesticides tested, fenthion was
most reactive, followed by phorate, with both rate constants
above 107/M/s, while fluometuron and atrazine were inter-
mediate and methyl parathion was the slowest (2.0
s). The relative error of this method is approximately 10%.
This method will assist in our efforts to fully characterize the
role of the carbonate radical in sunlight-driven natural cleans-
ing of field waters.
ϫ
105/M/
order rate constant of only 2.0
ϫ
105/M/s. It is likely the
reaction occurred at the sulfur atom and converted the pesticide
into the oxon derivative.
Acknowledgement—This project was funded by Natural Science and
Engineering Research Council of Canada and Dupont Canada. The
authors also wish to thank Debby Repka for her careful reading of
the manuscript.
Compared with the reactivity of the hydroxyl radical toward
pesticides [17], the second-order rate constant of the hydroxyl
radical was greater than that of the carbonate radical and varied
over one order of magnitude. The most reactive pesticide was
carbaryl, with a rate constant of 3.4
reactive pesticide was quinclorac, with a rate constant of 3.6
108/M/s. The rate constants of pesticides toward carbonate
radicals vary over two orders of magnitude, from the most
reactive fenthion (2.0
107/M/s) to the least reactive methyl
parathion (2.0
105/M/s). These results indicate that the hy-
ϫ
109/M/s, and the least
REFERENCES
1. Zepp RG. 1991. Photochemical fate of agrochemicals in natural
waters. Pesticide chemistry: Advances in international research,
development, and legislation. Proceedings, 7th International Con-
gress of Pesticide Chemistry, Hamburg, Germany, August 5–10,
1990, pp 329–345.
ϫ
ϫ
Ϫ
ϫ
2. Chen SN, Cope VW, Hoffman MZ. 1973. Behavior of CO3 rad-

New method for measuring carbonate radical reactivity
Environ. Toxicol. Chem. 19, 2000
1507
icals generated in the flash photolysis of carbonatoamine com-
plexes of cobalt (III) in aqueous solution. J Phys Chem 77:1111–
1116.
bonate radical anion with substituted phenols. Int J Radiat Biol
31:603–605.
15. King PA, Anderson VE, Edwards JO, Gustafson G, Plumb RC,
Suggs JW. 1922. A stable solid that generates hydroxyl radical
upon dissolution in aqueous solutions: Reaction with proteins and
nucleic acid. J Am Chem Soc 114:5430–5432.
16. Sedlak DL, Andren AW. 1991. Aqueous-phase oxidation of poly-
chlorinated biphenyls by hydroxyl radicals. Environ Sci Technol
25:1419–1427.
17. Mabury SA, Crosby DG. 1996. Pesticide reactivity toward hy-
droxyl and its relationship to field persistence. J Agric Food Chem
44:1920–1924.
18. Willson RL, Greenstock CL, Adams GE, Wageman R, Dorfman
LM. 1971. The standardization of hydroxyl radical rate data from
radiation chemistry. Int J Radiat Phys Chem 3:211–220.
19. Neta P, Huie RE. 1988. Rate constants for reactions of inorganic
radicals in aqueous solution. J Phys Chem Ref Data 17:1028–
1079.
20. Elango TP, Ramkrishnan V, Kuriacose JC. 1985. Generation and
reaction of inorganic radicals with special reference to carbonate
radicals. J Indian Chem Soc LXII1033–1037.
21. Crow JP, et al. 1994. On the pH-dependent yield of hydroxyl
radical products from peroxynitrite. Free Radic Biol Med 16:
331–338.
22. Huie RE, Neta P. 1986. Kinetics of one-electron reactions in-
volving ClO2 and NO2. J Phys Chem 90:1191–1198.
23. Forni LG, Mora-Arellano VO, Packer JE, Willson RL. 1986. Ni-
trogen dioxide and related free radical: Electron-transferreactions
with organic compounds in solutions containing nitrite or nitrate.
J Chem Soc 2:1–6.
3. Chen SN, Hoffman MZ, Parsons GH Jr. 1975. Reactivity of the
carbonate radical toward aromatic compounds in aqueous solu-
tion. J Phys Chem 79:1911–1912.
4. Eriksen TE, Lind J, Merenyi G. 1985. On the acid-base equilib-
rium of the carbonate radical. Radiat Phys Chem 26:197–199.
5. Chen SN, Hoffman MZ. 1975. Effect of pH on the reactivity of
the carbonate radical in aqueous solution. Radiat Res 62:18–27.
6. Larson RA, Zepp RG. 1988. Reactivity of the carbonate radical
with aniline derivatives. Environ Toxicol Chem 7:265–274.
7. Clifton CL, Huie RE. 1993. Rate constants for some hydrogen
abstraction reactions of carbonate radical. Int J Chem Kinet 25:
199–203.
8. Huie RE, Shoute LCT, Neta P. 1991. Temperature dependence of
the rate constants for reactions of the carbonate radical with or-
ganic and inorganic reductants. Int J Chem Kinet 23:541–552.
9. Adams GE, Bisby RH, Cundall RB, Redpath JL, Willson RL.
1972. Selective free radical reactions with proteins and enzymes:
The inactivation of ribonuclease. Radiat Res 49:290–299.
10. Chen SN, Hoffman MZ. 1973. Rate constants for the reaction of
the carbonate radical with compounds of biochemical interest in
neutral aqueous solution. Radiat Res 56:40–47.
11. Chen SN, Hoffman MZ. 1974. Reactivity of the carbonate radical
in aqueous solution. Tryptophan and its derivatives. J Phys Chem
78:2099–2102.
12. Anbar M, Meyerstein, Neta P. 1966. The reactivity of aromatic
compounds toward hydroxyl radicals. J Phys Chem 70:2660–
2662.
13. Elango TP, Ramkrishnan V, Vancheesan S, Kuriacose JC. 1984.
Reaction of the carbonate radical with substituted anilines. Proc
Indian Acad Sci 93:47–52.
24. Crable GF, Kearns GL. 1962. Effect of substituent groups on the
ionization potentials of benzenes. J Phys Chem 66:436–439.
25. Mabury SA, Crosby DG. 1994. The relationship of hydroxyl re-
activity to pesticide persistence. In Helz GR, Zepp RG, Crosby
DG, eds, Aquatic and Surface Photochemistry, CRC, Boca Raton,
FL, USA, pp 149–161.
14. Moore JS, Phillips GO, Sosnowski A. 1977. Reaction of the car-