6748 J. Phys. Chem. A, Vol. 106, No. 29, 2002
Zona et al.
•
•+
for the following compounds are all within 10%: phenol (y )
Cl-PhO ), measured by TMPD formation (reaction 10) via
pulse radiolysis, Table 1. The chloride yields after steady-state
2
2
2
3
2
0
82x, R ) 0.998), 2-chlorophenol (y ) 283x, R ) 0.999),
2
-
-chlorophenol (y ) 315x, R ) 0.997), 4-chlorophenol (y )
radiolysis, however, gave an initial concentration of G(Cl ) )
2.07, corresponding to 33% of OH which is much higher than
the phenoxyl radical concentration. This indicates that the
product chloride results from both the first-order decay of the
ipso-chloro- OH adducts, reactions 5 and 7, Scheme 1, and also
2
2
•
87x, R ) 0.996), and 2,4-dichlorophenol (y ) 258x, R )
.999). Therefore, a mean value of these data, y ) 285x, was
taken for the quantification of the hydroxylated derivatives of
,4-D. Similar experiments and analysis were also carried out
•
2
for PAA measurements.
from second-order reactions of other chloro-hydroxycyclohexa-
dienyl radicals. The meta OH-adducts (reaction 6 and 8) may
undergo disproportionation with a chloride release reaction 18
or combination at the chlorine substituted carbons. Such dimers
•
The G values, obtained from the linear part of the yield vs
dose plots for the radiolytic products of 2,4-D and PAA under
various oxidation conditions, are summarized in Tables 3 and
are expected to eliminate HCl very efficiently, e.g., reaction 19
4.
-
(
R ) -CH2COO ). An HX (X ) F, Cl) release from a product
For PAA in the presence of K3[Fe(CN)6], a quantitative
material balance could be achieved. The fractions of OH-adduct
distribution on the ring positions ipso:ortho:meta:para was 0.06:
•
of similar disproportionations and/or dimerizations has been
observed by pulse radiolysis with conductivity measurements
•
20
for the OH adducts of fluoro- and chlorobenzene.
0.45:0.11:0.36. Attack on the side chain obviously plays a minor
role. These results are comparable to that reported for anisole,
ipso:ortho:meta:para ) 0.07:0.48:0.07:0.38.22
The presence of two chlorine atoms on the phenoxyacetic
•
acid molecule changes the pattern of OH addition into a more
statistical distribution. For 2,4-D, the total yield of hydroxylation
on C6, C3, and C5 (2,4-diCl-OH-PAA), detected by oxidation
with K3[Fe(CN)6], is 47%. Because the product at 4.5 min
(
Figure 8, Table 3) is formed with the highest yield, 20% of
•
the initially generated OH radicals, it can be assumed to be
2
,4-dichloro-6-hydroxyphenoxyacetic acid, arising from the
•
OH adduct on the free ortho position of the molecule, reaction
9
, Scheme 1. The products resulting from the meta adducts, 3-
and 5-hydroxy-2,4-dichlorophenoxyacetic acid, contribute 27%
of the total reaction. Because of the ortho-para directing
•
chlorine substituents, the OH attachment on the meta positions
of 2,4-D is significantly higher than for PAA, where only 11%
of 3- and 5-hydroxyphenoxyacetic acid were produced. The
yield of 2,4-DCP (17%) corresponds to the fraction of the OH
Summary
•
attack on C1.
Electron pulse radiolysis and gamma radiolysis experiments,
together with product analysis by HPLC, have been performed
to elucidate the primary patterns of OH radical attack on
Because phenoxyl radicals, which are precursors of 2-chloro-
•
4
-hydroxy- and 4-chloro-2-hydroxyphenoxyacetic acid (OH-
phenoxyacetic acid (PAA) and 2,4-dichlorophenoxyacetic acid
(2,4-D). In the gamma radiolysis experiments, the yield of
Cl-PAA), are not oxidized by K3[Fe(CN)6], the yield of 6%
•
(Table 3) does not reflect the primary OH attack. This has to
•
primary OH adducts on positions not occupied by chlorine
be determined from the TMPD experiments, where about 20%
of the phenoxyl radicals could be detected (Table 1). Sum-
marizing the data from both series, as fractions of OH reaction
atoms were determined by oxidation with K3[Fe(CN)6] to the
•
•
corresponding hydroxylated compounds. The ipso-chloro- OH
adducts have been quantified via their phenoxyl radicals, formed
in a spontaneous HCl elimination step. These rapidly oxidize
with 2,4-D, gives C1, 0.17; C2/C4, 0.20; C3/C5, 0.27; and C6,
0
.20, demonstrating little preference for any of the positions
•
+
•
TMPD to TMPD , as measured by pulse radiolysis. For PAA,
on the ring. The material balance is 84% of the total OH
reaction. From this result, one could conclude that 16% of
hydroxyl radicals react with the side chain, however, this appears
relatively high compared to PAA, where this process contributed
•
the OH attachment preferentially takes place on ortho and para
positions to the acetic acid group, wheras for the dichlorinated
•
compound 2,4-D, a more statistically distribution of OH
5
-1 -1
addition to the molecule takes place. In contrast to PAA, the
only <3%. However, the low value of k17 < 5 × 10 M
s
•
OH adducts of 2,4-D showed no significant reaction with
also means that bimolecular decay of the OH adducts will also
occur in competition to the K3[Fe(CN)6] oxidation process which
would readily account for the lower material balance observed.
oxygen.
Acknowledgment. The use of the equipment of the RISOE
National Laboratory is gratefully acknowledged. S.P.M. ac-
knowledges the time and help of the personnel at the Radiation
Laboratory, University of Notre Dame, which is supported by
the Office of Basic Energy Sciences of the U.S. Department of
Energy. R.Z. thanks the University of Vienna for a stipend for
his stay in Denmark.
In N2O/O2 (4:1 v/v) saturated solutions, the total formation
of phenolic products from 2,4-D is very similar to that in the
presence of N2O, 28% and 32% (Table 3). This confirms the
pulse radiolysis results, showing that the hydroxycyclohexadi-
enyl radicals of 2,4-D have a very low reactivity toward oxygen.
•
For PAA, where a reaction of oxygen by the OH adducts was
8
-1 -1
observable (k15 ) (5.1 ( 0.2) × 10 M s ), a considerable
increase of 2-hydroxyphenoxyacetic acid from 5.5% in N2O to
References and Notes
2
2.7% in N2O/O2 (Table 4) occurred.
(1) Sun, Y.; Pignatello, J. J. EnViron. Sci. Technol. 1993, 27, 304.
(2) Sun, Y.; Pignatello, J. J. EnViron. Sci. Technol. 1995, 29, 2065.
(3) Modestov, A. D.; Lev, O. J. Photochem. Photobiol. A: Chem. 1998,
For 2,4-D, product chloride determination after gamma
radiolysis in N2O has also been performed to correlate its
production to that of chlorophenoxyl radical generation (R-
112, 261.