Environ. Sci. Technol. 2004, 38, 2082-2088
Solvent-Specific Photolytic Behavior
of Octachlorodibenzo-p-dioxin
JINA CHOI AND WONYONG CHOI*
School of Environmental Science and Engineering,
Pohang University of Science and Technology,
Pohang 790-784, Korea
BYUNG JIN MHIN
Department of Chemistry, PaiChai University, 493-6
Doma-dong, Seoku, Taejun 302-735, Korea
The photolysis of octachlorodibenzo-p-dioxin (OCDD) was
investigated in various organic solvents under λ > 300
nm irradiation. The rates of OCDD photolysis were highly
solvent-specific. OCDD was photodegraded rapidly in toluene,
n-hexane, CCl4, and 1-octanol, whereas it underwent
negligible photodegradation in acetonitrile, acetone, and
methanol. Both OCDD photolysis and fluorescence emission
did not take place in very polar solvents because polar
solvent molecules efficiently quenched the excited OCDD
(OCDD*). The addition of acetonitrile to an OCDD solution
in toluene rapidly quenched both the fluorescence emission
and the photolysis rate, which can be described by SternVolmer analysis. The efficient photolysis in nonpolar (or
less polar) solvents seems to be mediated through a chargetransfer path where the solvent and OCDD* act as an
electron donor and acceptor, respectively. However, OCDD
photolysis in CCl4 seems to represent the opposite case
in which the solvent is an electron acceptor and OCDD* is
an electron donor. Hammett σ constants that approximately
represent the electron-donating power in structurally
related aromatic solvents show a good correlation with
the photolysis rates. We propose that the solvent specificity
in OCDD photolysis is mainly ascribed to the difference
in the electron donating (or accepting) tendency among
various solvents. When triethylamine that easily donates an
electron to form a charge-transfer exciplex with OCDD*
was added, a marked enhancement in the photolysis rate
was observed.
Introduction
Polychlorinated dibenzo-p-dioxins (PCDDs) are highly toxic
and widespread pollutants, which are generated from various
incomplete combustion processes (e.g., municipal solid waste
incinerators) and chemical processes dealing with chlorinated
aromatic compounds (1-3). There are 75 PCDD congeners
with one to eight chlorines, and their physicochemical
properties and biological toxicities are highly congenerspecific in general (4-8). In particular, seven PCDD congeners
with chlorine atoms substituted at four lateral positions (2,
3, 7, and 8) are considered highly toxic (9). Octachlorodibenzo-p-dioxin (OCDD) has the least toxicity among seven
toxic congeners but has the highest abundance in the
environment due to its persistency (10). OCDD could be
potentially more toxic when it transforms into less chlorinated
congeners through dechlorination.
* Corresponding author phone: +82-54-279-2283; fax: +82-54279-8299; e-mail: [email protected].
2082
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 7, 2004
Since PCDDs are chemically stable and accumulate in
various media of natural environments and living organisms,
the development of effective PCDDs destruction methods
and understanding their degradation mechanisms in the
environment are of significant importance. Many degradation
methods of PCDDs, which include incineration (11), catalytic
degradation (12, 13), radiolysis (14), ozonation (15), ozonation-photolysis (16), photo-Fenton reaction (17), direct
photolysis (18-20), and photocatalysis (21-23), have been
investigated. Among them, photolysis has been intensively
investigated since it is the most important degradation
pathway in environmental media where solar light can reach.
PCDDs can absorb near-UV light that is contained in sunlight
and subsequently undergo a series of photochemical processes to induce their transformation or degradation. The
photochemistry and photolytic properties of PCDDs have
been studied in various media such as organic solvents (22,
24, 25), acetonitrile/water mixture (26-30), and on solid
surfaces (25, 31-33). Most photoreactions of PCDDs were
carried out in organic solvents because their water solubility
is extremely low.
Many studies have shown that PCDDs can be successfully
photodegraded in various organic solvents and that the
presence of some additives (e.g., triethylamine, borohydride)
(34, 35) could enhance the photolysis rates significantly.
However, understanding detailed photolytic mechanisms and
identifying the photodegraded products and intermediates
remain largely elusive. It has been frequently reported that
photochemical conversion to dechlorinated (or less chlorinated) PCDDs is a minor path (19) and that the cleavage of
C-O bond is critical in initiating their photolytic reactions
(36). In particular, the photolysis of PCDDs is strongly
dependent on the nature of solvents or surrounding media
(24). However, there have been few attempts to elucidate the
relationship between solvent properties and photolysis rates
of PCDDs. Although it has been often proposed that the
solvent effects in dioxin photolysis are related to the hydrogen
donating potentials or solvent polarities (10, 24, 37), the level
of understanding is largely unclear and speculative.
In this work, we carried out the photolytic degradation
of OCDD, the most recalcitrant dioxin congener, in various
organic solvents, and tried to understand the origin of solvent
effects in its photolysis. The rates of OCDD photolysis were
highly solvent-specific. Based upon the kinetic and spectroscopic data, we propose that not only the solvent polarity
but also the electron-donating (or accepting) tendency of
the solvent is important in determining the photolysis rate
of OCDD. In general, the photolysis of PCDDs in various
environmental media should be sensitively affected by the
electron donating (or accepting) potential of the media.
Experimental Section
Chemicals. OCDD was purchased from Ultra Scientific.
A PCDD mixture containing 2,3,7,8-tetrachlorodibenzop-dioxin (TCDD), 1,2,3,7,8-petachlorodibenzo-p-dioxin
(PeCDD), 1,2,3,4,7,8-hexachlorodibenzo-p-dioxin (HxCDD),
1,2,3,4,6,7,8-heptachlorodibenzo-p-dioxin (HpCDD), and
OCDD was purchased from AccuStandard to be used as a
standard for intermediates identification. Dibenzo-p-dioxin
(DD) was also obtained from AccuStandard. Organic solvents
used in this study include the following: toluene, n-hexane,
methanol, ethanol, and acetone (analytic grade, J. T. Baker)
and 2-propanol, acetonitrile, carbon tetrachloride, methylene
chloride, 1-octanol, benzene, chlorobenzene, and benzonitrile (HPLC grade, Aldrich). When needed, triethylamine (TEA,
10.1021/es034916s CCC: $27.50
 2004 American Chemical Society
Published on Web 02/19/2004
Aldrich) was used as an additive to the organic solution of
OCDD.
Photolysis. A stock solution of OCDD of 218 µM (100
mg/L) was made in toluene. OCDD in other solvents was
prepared by evaporating toluene in an aliquot of the stock
solution and subsequently redissolving in a calculated
amount of a specific solvent. An OCDD solution in a specific
solvent (typically 2.2 µM, or 1 mg/L) was held in a 30 mL
photoreactor and stirred magnetically for 30 min prior to
illumination and constantly during illumination. Most photolysis experiments were performed using a 300-W Xe-arc
lamp (Oriel). A 450-W Xe-arc lamp was used for some cases
when the 300-W lamp was not available. Light passed through
a 10-cm IR water filter and a UV cutoff filter (λ > 300 nm)
to simulate the solar irradiation. Sample aliquots were
intermittently taken from the illuminated reactor and injected
into a 2-mL glass vial for analysis. Multiple photolyses (2-3
times) were carried out under an identical condition. A
photolysis of DD in toluene was also carried out in the same
way for comparison with the OCDD photolysis
Analysis. OCDD and intermediates were analyzed by using
a Hewlett-Packard gas chromatograph (HP 6890A) equipped
with a 63Ni electron capture detector (ECD) and a HP-5
column (30 m × 0.32 mm × 0.25 µm). Nitrogen was used as
a carrier gas. For the analysis of OCDD in CCl4 and CH2Cl2,
the solvent was replaced by toluene through evaporation
and redissolution prior to injection to the GC/ECD. Duplicate
injections were made for each sample analysis with a splitless
mode. The injector and detector temperature was 300 °C.
The oven temperature was held at 100 °C for 1 min, then
increased at a rate of 5 °C/min up to 280 °C, and held at 280
°C for 15 min. The formation of 1,2,3,4,6,7,9-HpCDD as a
main byproduct from OCDD degradation was suggested from
the analysis of GC/ECD chromatogram and was further
confirmed by HRGC/HRMS (JMS700T, JEOL, Japan) analysis
following SIM (Selected Ion Monitoring) method using
13C-internal standards (Wellington Laboratory, Canada). DD
analysis was performed on a reverse-phase HPLC column
with acetonitrile-water solvent and UV detection (Agilent
1100 Series). UV absorption spectra of OCDD in organic
solvents were obtained with a UV/vis spectrophotometer
(Shimadzu UV-2401 PC) and a 1-cm path length quartz cell.
Fluorescence spectra of OCDD in various solvents were
obtained using a spectrofluorometer (Shimadzu RF-5301).
Excitation wavelength used was 304 nm.
Results and Discussion
Solvent-Specific Photolysis of OCDD. The photolytic degradation of OCDD was carried out in various organic solvents
as shown in Figure 1. The photolysis rates of OCDD under
an identical irradiation condition were markedly affected by
the kind of solvents. OCDD was photodegraded rapidly in
toluene, n-hexane, and 1-octanol, whereas it underwent
negligible photodegradation in acetonitrile, acetone, and
methanol. The photolytic reactivity of OCDD seems to be related with the solvent polarity. Table 1 shows a correlation
between OCDD photolysis and solvent polarity: OCDD is
photodegraded faster in less polar solvents than in polar solvents. However, the correlation is not quantitative. The effect
of solvent polarity on the photolysis will be discussed later.
It should be considered the possibility that the different
photoreactivity might be due to different molar absorptivity
(m) of OCDD depending on the solvent polarity. The light
absorbed by OCDD (Ia) in a specific solvent can be expressed
by the Beer-Lambert law (eq 1)
Ia(λ) ) I0(λ)[1 - exp(-2.303mCL)] )
Ii(λ)T(λ)[1 - exp(-2.303mCL)] (1)
where Ia(λ) ) absorbed light intensity; I0(λ) ) light intensity
FIGURE 1. Photolytic degradation of OCDD (2.2 µM) in various
organic solvents under λ > 300 nm irradiation (300-W Xe lamp). The
solid lines represent the first-order fit: (a) acetonitrile, (b) methanol,
(c) acetone, (d) ethanol, (e) 2-propanol, (f) n-hexane, (g) toluene, (h)
1-octanol. The inset shows the photolysis of OCDD (2.2 µM) in CCl4
and CH2Cl2 with a 450-W Xe lamp.
TABLE 1. Comparison between Dielectric Constants of the
Solvent and the Photolysis of OCDD
solvent
dielectric
constant (39)
degradation (%)
after 6 h photolysis
toluene
n-hexane
CCl4
CH2Cl2
1-octanol
2-propanol
ethanol
acetone
methanol
acetonitrile
2.38
1.89
2.24
8.93
10.30
20.18
25.60
21.01
33.00
36.64
72
66
65a
22a
72
43
18
0
0
0
a Photolyzed with a 450-W Xe lamp while a 300-W Xe lamp was used
in other cases.
incident onto the reactor; Ii(λ) ) light intensity emitted from
the light source; T(λ) ) transmittance through the UV cutoff
filter; C ) OCDD concentration; and L ) reactor path length.
Under the condition of weak absorption by OCDD in the
present photolysis experiments (see Figure 2), eq 1 can be
approximated to eq 2 (38):
Ia(λ) ) 2.303Ii(λ)T(λ)m(λ)CL
(2)
The integrated total light absorption by OCDD is represented
by
Ia )
∫ I (λ)dλ ) 2.303CL ∫ I (λ)T(λ)
λ a
λ i
m(λ)dλ
(3)
Since the light intensity from the Xe-arc lamp is relatively
constant over the narrow wavelength range (280-330 nm)
of interests in this work, the total light absorption by OCDD
in a specific solvent, Ia, is proportional to the following integral
term.
Ia ∝
∫ T(λ)
λ
m(λ)dλ
(4)
The absorption spectra of OCDD in different solvents are
compared in Figure 2. The m at 313 nm and ∫λT(λ)m(λ)dλ
(over the wavelength range 250-350 nm) values are compared in Table 2, which shows that there is no correlation
VOL. 38, NO. 7, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2083
FIGURE 2. UV-vis absorption spectra of OCDD in several organic
solvents ([OCDD] ) 5.4 µM, cell path length ) 1 cm). The light
transmittance (T(λ)) profile through the UV cutoff filter used in this
study is compared along with the spectra. The inset shows the
enlarged absorption spectra of OCDD in acetonitrile and the
integrated area (shaded) of eq 4 (∫λT(λ)Em(λ)dλ) that is directly
proportional to the total light absorption by OCDD in a specific
solvent.
TABLE 2. Molar Absorptivity (Em) at 313 nm, the Integral Term
of Eq 4, and the Photolysis of OCDD in Several Organic
Solvents
solvent
Em (M-1 cm-1)
at 313 nm
∫λT(λ)Em(λ)dλa
degradation (%)
after 6 h photolysis
toluene
n-hexane
1-octanol
2-propanol
ethanol
methanol
acetonitrile
2370
1852
333
2981
1870
2389
2796
549b
502
108
617
450
507
693
72
66
72
43
18
0
0
a
Integration range λ: 250-350 nm.
b
FIGURE 3. (a) Fluorescence emission spectra of OCDD (5.4 µM) in
various organic solvents (λex ) 304 nm). (b) Effect of cosolvent on
the fluorescence intensity of OCDD in toluene (λex ) 304 nm,
cosolvent: toluene ) 50:50 (v/v)). The numbers in parentheses are
the dielectric constant of the cosolvent.
Arbitrary unit.
between the light absorption (Ia) and the photolysis rates of
OCDD in various organic solvents. Therefore, the solventspecific photolysis of OCDD should not be ascribed to the
difference in Ia among different organic solvents. Hung et al.
(37) also reported a similar result. They compared the molar
absorptivities of OCDD at 314 nm with its photolysis rates
in hexane and 60% water/acetonitrile and showed that the
photolysis of OCDD was about 30 times faster in hexane
(having lower molar absorptivity) than in 60% water/acetonitrile (having higher molar absorptivity). This implies that the
direct homolytic pathway (reactions 6 and 7) in OCDD photodegradation should not be important since the rate of direct
photolysis (kp) should be proportional to Ia as follows (38).
kp(λ) ) Φ(λ)Ia(λ)
[Φ, quantum yield of photolysis] (5)
On the other hand, the photolytic pathway could be
enhanced in the presence of efficient H-atom donors (DH)
(reaction 8) or electron donors (D) (reaction 9) (34). The role
of solvent molecules as an H-atom donor in PCDD photolysis
has been often discussed in the literature.
Ar-Cl + hv f [Ar-Cl]*
•
[Ar-Cl]* f Ar + Cl
(6)
•
(7)
[Ar-Cl]* + DH f ArH + D• + Cl•
•
[Ar-Cl]* + D f Ar + Cl + D
s
+•
(8)
(9)
Crosby et al. (10) reported that the direct photolysis of PCDDs
2084
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 7, 2004
requires the presence of organic hydrogen donors. Hung et
al. (37) carried out the photolysis of OCDD in hexane or
acetonitrile/water solution with cosolvents and concluded
that the H-atom donating potential of each solvent or
cosolvents that are better H-atom donors than the solvent
itself is important in determining the photolysis rates of
OCDD. However, the energy required to abstract an H atom
from the solvent molecule does not correlate with the
photolysis rate. For example, the C-H bond energy of
acetonitrile (393 kJ/mol) (39) in which OCDD is not photolyzed at all is lower than that of benzene (465 kJ/mol) (39)
in which the photolysis of OCDD is observed (compare Figures
1 and 6a). Furthermore, the inset in Figure 1 shows that the
photolysis of OCDD in CCl4 (no H-atom donor) is faster than
in CH2Cl2. Qin (40) also reported the similar result in the
photolysis of 2,3,7,8-TCDD. Therefore, it seems that the
H-atom donating potential of the solvent is not the dominating parameter to determine the photolysis rate of PCDDs.
The role of the solvent as an electron donor (reaction 8)
should be considered, which will be addressed later.
Fluorescence and Photolysis of OCDD. We next consider
the possibility that the solvent-dependence of OCDD photolysis could be related to varying lifetimes of excited OCDD
(OCDD*) depending on the solvent polarity. Figure 3a
compares the fluorescence spectra of OCDD in several
solvents. A broad and unstructured emission band (350500 nm) is clearly present in nonpolar solvents such as
n-hexane, CCl4, and toluene, whereas the fluorescence
intensity is significantly reduced in CH2Cl2 (polar solvent) or
completely absent in acetonitrile. Nevertheless, the OCDD
absorbance at the excitation wavelength (304 nm) is higher
in acetonitrile than in toluene (see the inset in Figure 3a).
The quenching of OCDD* by a highly polar solvent must be
very efficient. Figure 3b shows that the order of fluorescent
FIGURE 4. (a) Fluorescence emission spectra of OCDD (0.11 mM)
in toluene (λex ) 304 nm) in the presence of varying concentrations
of acetonitrile (MeCN). (b) Effect of the added acetonitrile
concentration on the fluorescence intensity. The inset shows a
Stern-Volmer plot (with R2 ) 0.989) of OCDD fluorescence quenching
by acetonitrile.
intensities of OCDD* in cosolvent-toluene mixtures is in
good agreement with the order of dielectric constants of the
cosolvent. Since an electronically excited OCDD* may be a
very polarizable species such as other excited molecules (41),
it can interact strongly with polar solvent molecules. As a
result, the polar solvent can quench OCDD* efficiently with
reducing both its fluorescence intensity and its photolysis.
This is consistent with the fact that OCDD is not photolyzed
at all in polar solvents (Figure 1).
The addition of acetonitrile to OCDD solution in toluene
rapidly quenches the fluorescence as shown in Figure 4. The
Stern-Volmer analysis (41) can be applied to the photolysis
of OCDD by assuming a simple competition among fluorescence (reaction 11), quenching (reaction 12), and photolysis (reaction 13).
OCDD + hv f OCDD*
r ) Ia
(10)
r ) kf[OCDD*]
(11)
r ) kq[OCDD*][Q]
(12)
r ) kph[OCDD*]
(13)
OCDD* f OCDD + hvf
OCDD* + Q f OCDD + Q
OCDD* f photolysis
where kf, kq, and kph represent the rate constant of unimolecular fluorescence, bimolecular quenching, and photolysis,
respectively.
The steady-state approximation for [OCDD*] yields
[OCDD*] ) Ia/(kf + kph + kq[Q]). The fluorescence intensity
(If) and the photolysis rate (rph) are proportional to kf[OCDD*]
and kph[OCDD*], respectively, then each of which can be
expressed as kfIa/(kf + kph + kq[Q]) and kphIa/(kf + kph + kq[Q]).
FIGURE 5. (a) Photolytic degradation profile of OCDD (2.2 µM) in
toluene solutions mixed with varying concentrations of acetonitrile
(MeCN). (b) Effect of the acetonitrile concentration on the OCDD
photolysis rate constants. The inset shows a Stern-Volmer plot
(with R2 ) 0.988).
If we denote the fluorescence intensity and the photolysis
rate in the absence of quencher (Q) as If° and rph°, respectively,
it follows that
If°/If ) rph°/rph ) 1 + kq[Q]/(kf + kph)
(14)
In acetonitrile (or other highly polar solvents), the quenching
path (reaction 12) must be dominant, and consequently the
emission (reaction 11) and the photolysis (reaction 13) should
be minimal. The inset in Figure 4b shows that the fluorescence
quenching by acetonitrile can be successfully described by
eq 14 with yielding the slope value of 0.42. On the other
hand, the effect of acetonitrile quenching on the photolysis
rate of OCDD is shown in Figure 5. The plot of kph°/kph vs
[MeCN] in Figure 5b (inset) also exhibits a good fit to eq 14.
(The ratio, rph°/rph, in eq 14 can approximate to kph°/kph when
initial reaction rates are compared.)
The Role of Solvent in OCDD Photolysis. All of the above
results strongly support that the photolysis of OCDD in very
polar solvents does not take place because OCDD* is
efficiently quenched by polar solvent molecules. However,
the solvent polarity alone does not account for the solventspecific photolysis. For example, the photolysis in 1-octanol
is slightly faster than in n-hexane although 1-octanol is much
more polar that n-hexane (Figure 1). This implies that the
role of the solvent in the photolysis mechanism is not simply
to quench OCDD*. Since neither the direct homolysis
(reaction 7) nor the H-atom transfer (reaction 8) seems to
be a major path in the OCDD photolysis, the charge-transfer
path (reaction 9) where the solvent molecule serves as an
electron donor (or acceptor) should be taken into account
to explain the solvent-specific photolysis.
VOL. 38, NO. 7, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2085
electronic effect of aromatic substituents can be expressed
by Hammett σ constants, which have positive values for
electron-withdrawing groups and negative values for electrondonating groups. The photolysis rates monotonically decrease
with increasing σ values (Figure 6b). The σ constants show
a better correlation with the photolysis rate than the IPs of
the solvent. As for the Hammett σ constants, an average value
of σm(meta-substituents) and σp(para-substituents) was used
for each solvent. In qualitative terms, aromatic solvents with
electron-donating substituents (e.g., toluene) should facilitate
the photoactivated charge transfer between the solvent and
OCDD and hence the photolysis of OCDD. By applying the
Hammett equation, the photolysis rate constants of OCDD
can be expressed as
log(ks/kbz) ) Fσ
FIGURE 6. (a) Photolytic decay profiles of OCDD (2.2 µM) in
structurally related aromatic solvents (benzonitrile, chlorobenzene,
benzene, toluene) under λ > 300 nm irradiation. (b) Correlation
between the OCDD photolysis and average Hammett constants (σav)
of the substituted aromatic solvents. The numbers in parentheses
represent the ionization potential (eV) of the solvent. The inset
shows the Hammett plot.
The excited OCDD molecules may form transient chargetransfer complexes with solvent molecules, which can be
followed by a series of degradation steps. Depending on the
electron-donating or electron-accepting potentials of the
solvent with respect to those of OCDD, it may serve as an
electron donor (D) or an electron acceptor (A).
OCDD* + D T (OCDDδ-...Dδ+)q ff degradation (15)
OCDD* + A T (OCDDδ+...Aδ-)q ff degradation (16)
In this case, the photolysis rates of OCDD can be directly
related with the electron-donating potential or electronaccepting potential of the solvent, which can be approximately represented by its ionization potential (IP) or
electron affinity (EA).
In terms of the transition state theory (42), the formation
of the photoactivated charge-transfer complexes between
the excited OCDD and the solvent molecules should be
equilibrated with the ground-state reactants prior to the
photolysis, and the magnitude of the activation energy (Ea)
for the charge transfer should dictate the photolytic rates. If
we assume that increasing the electron-donating (or electronaccepting) potential of the solvent decreases Ea and that the
tendency to donate (or accept) an electron is largely
determined by the electronic effects in solvent molecules,
we may apply the linear free energy relationship (LFER) (43)
to relate the photolysis rates of OCDD in structurally related
solvents. In Figure 6, we compare the photolytic decay profiles
of OCDD in structurally related aromatic solvents: toluene,
benzene, chlorobenzene, and benzonitrile. The inductive
2086
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 7, 2004
(17)
where ks and kbz represent the photolysis rate constant of
OCDD in a specific solvent of interest (substituted benzene)
and the reference solvent (benzene), respectively, and F is
an empirical parameter that is constant for a given set of
photolytic reactions. The inset in Figure 6b shows the
Hammett plot for the OCDD photolysis. Since the dielectric
constants of toluene (2.38), benzene (2.28), and chlorobenzene (5.69) are very similar, the solvent polarity should be
a minor factor in explaining the different photolysis rates
shown in Figure 6. Therefore, the electron-donating tendency
of the solvent seems to be critical in determining the
photodegradation efficiency of OCDD.
When the solvent has lower dielectric constant (nonpolar
or less polar) and lower IP (electron donating), the photolysis
of OCDD is mediated through the reaction 15 path where
the solvent and OCDD* act as an electron donor and acceptor,
respectively. This is the case for most nonpolar solvent
systems. On the other hand, the opposite case in which the
solvent is an electron acceptor and OCDD* is an electron
donor (reaction 16) enables the photolysis of OCDD as well.
The photolysis of OCDD in CCl4 should be the case. CCl4 has
a very high IP (11.47 eV) and cannot serve as an electrondonating medium. However, it may play the role of an
electron acceptor since CCl4 has much higher EA (2.0 eV)
(44) than OCDD (0.664 eV) (8). In addition, IP of OCDD should
be smaller than that of CCl4 (11.47 eV) judging from the fact
that the reported IP values for DD, 2-MCDD, and 2,8-DCDD
are 7.60, 7.71, and 7.80 eV, respectively (45). Therefore, the
photoinduced electron transfer from OCDD to CCl4 is
energetically favored, and OCDD might act as an electron
donor. As a result, the rate of OCDD photolysis in CCl4 is
comparable to that in toluene or n-hexane (see the inset in
Figure 1).
Effect of Triethylamine as an Electron Donor. It is known
that the photoreductive degradation of chlorinated aromatic
compounds is enhanced in the presence of triethylamine
(TEA) through the formation of intermediate exciplex (46,
47). Enhanced photolysis of OCDD in the presence of TEA
has been also reported and ascribed to the exciplex formation
between OCDD and TEA (34). Figure 7 confirms the previous
finding. The photolysis of OCDD in toluene is drastically
enhanced with TEA, and the photolysis in acetonitrile is
enabled only in the presence of TEA. Figure 8 shows that the
presence of TEA quenches the fluorescence as well. However,
this TEA-induced quenching is different in its origin from
the polarity-induced quenching shown in Figures 3b and 4a
because TEA is as nonpolar as toluene (dielectric constant
of TEA is 2.4). The exciplex formation (reaction 18) that is
subsequently followed by photolysis should compete with
the emission (reaction 11) with reducing the fluorescence.
OCDD* + Et3N f [Et3N δ+...OCDDδ-] ff degradation
(18)
FIGURE 9. Dependence of the photolysis rates on [OCDD] in toluene.
FIGURE 7. Photolytic degradation of OCDD with or without 1 M
triethylamine (TEA) (a) in toluene and (b) in acetonitrile solution.
the difference in electron donating (or accepting) tendency
among various solvents. When a reagent like TEA that easily
donates an electron to form a charge-transfer exciplex is
added, a marked enhancement in the photolysis rate should
be expected. Within the exciplex, an OCDD molecule carries
radical anionic character with a reduced photodissociation
energy. In addition, the photolysis rate of OCDD in N2-purged
toluene solution is comparable to that in air-equilibrated
solution, which precludes the role of singlet oxygen in this
photolytic mechanism. On the other hand, Figure 9 shows
that the photolysis rates of OCDD are independent of [OCDD],
which rules out the possibility that OCDD excimers might
be involved in photolysis.
The cleavage of C-Cl bond and the subsequent H-atom
abstraction from the solvent molecules should produce
dechlorinated congeners.
[OCDD δ-...D-Hδ+] f HpCDD + D•+ + Cl-
FIGURE 8. Fluorescent emission spectra of OCDD (0.11 mM) in
toluene (λex ) 304 nm) in the presence of TEA.
The exciplex is known to be formed between the lowest
excited singlet states of chlorinated aromatic compound and
the ground-state amine with the electron transferred from
the amine to aromatic compound (41). TEA is obviously a
better electron donor (with IP ) 7.5 eV) than the solvents
used in this study (with IP > 8.8 eV) and preferably forms an
exciplex with OCDD. The positive TEA effects were also
observed in other organic solvents used in this study. An
experimental evidence for the exciplex formation between
naphthalene and TEA (structurally similar to OCDD-TEA
exciplex) was recently reported by observing a fluorescence
band red-shifted by 30 nm-1 from the original fluorescence
band of naphthalene (48).
OCDD Photolytic Mechanism. We propose that the
formation of the partially ionic OCDD (as in reactions 15 and
16) is a prerequisite for its efficient photolysis and that the
solvent specificity in OCDD photolysis is mainly ascribed to
(19)
In the photolysis of OCDD in toluene, we could detect
1,2,3,4,6,7,9-HpCDD as a main byproduct and a small amount
of 1,2,3,4,6,7,8-HpCDD. Their intensity ratio was measured
to be 20:1 after 6 h of irradiation. 2,3,7,8-TCDD, the most
toxic congener, was not produced at all. In the presence of
TEA, 1,2,3,7,8-PeCDD, 1,2,3,4,7,8-HxCDD, and other unidentified dechlorinated dioxin congeners were detected after
6 h of photolysis of OCDD in toluene. On the other hand, no
GC/ECD peaks for dechlorinated dioxin congeners were
detected at all during the photolysis of OCDD in CCl4 since
CCl4 is not an H-atom donor. Detailed and quantitative
intermediates and products analysis was not carried out in
this study. However, the photoconversion to dechlorinated
congeners is only minor (estimated to be less than 20%) and
cannot account for the major photolytic pathway of OCDD
in this study. Similar observations have been frequently
reported, and the exact photodegradative mechanisms of
PCDDs are largely unknown (28, 30). A recent photolytic
study of 2,3,7,8-TCDD proposed that the cleavage of the C-O
bond leads to the production of chlorinated dihydroxybiphenyls with a satisfactory photochemical mass balance (36).
To assess the possibility of C-O bond cleavage in the PCDD
photolysis under the present experimental condition, the
photolysis of non-chlorinated dibenzo-p-dioxin (DD) in
toluene was tested. Since DD does not have any C-Cl bond,
its photolysis should be initiated by the cleavage of the C-O
bond. DD in toluene (1.1 mM) could be rapidly photodegraded under λ > 300 nm with the half-life of about 0.5 h.
Therefore, the cleavage of the C-O bond in the dioxin
molecular framework should take place. We believe that such
a C-O bond cleavage is a major initiation step in dioxin
photolysis. The bond breakage process may not be strictly
homolytic but should require attaining some ionic (heterolytic) character at the transition state. When the solvent serves
VOL. 38, NO. 7, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2087
as neither an electron donor nor electron acceptor, the
photolysis of PCDDs should be negligible. The degradation
products could be different depending on whether OCDD
acts as an electron acceptor or donor. The reaction pathways
following the C-O cleavage and the product formation should
depend on the availability of H-atom donors. Further studies
addressing this issue are required.
Acknowledgments
This work was supported by KOSEF through the Center for
Integrated Molecular Systems and by the Brain Korea 21
project. Helpful discussions with Profs. Min Joong Yoon and
Ja Kang Ku are appreciated.
Literature Cited
(1) Addink, R.; Olie, K. Environ. Sci. Technol. 1995, 29, 1425.
(2) Dickson. L. C.; Lenoir, D.; Hutzinger, O. Environ. Sci. Technol.
1992, 26, 1822.
(3) Paasivirta, J. Water Sci. Technol. 1988, 20, 119.
(4) Choudhary, G. Chlorinated Dioxins and Dibenzofurans in the
Total Environment; Choudhary, G., Keith, L. H., Rappe, C., Eds.;
Butterworth: Boston, 1983; Chapter 19.
(5) Mhin, B. J.; Choi, J.; Choi, W. J. Am. Chem. Soc. 2001, 123, 3584.
(6) Mhin, B. J.; Lee, J. E.; Choi, W. J. Am. Chem. Soc. 2002, 124, 144.
(7) Lee, J. E.; Choi, W.; Mhin, B. J. J. Phys. Chem. A 2003, 107, 2693.
(8) Lee, J. E.; Choi, W.; Mhin, B. J. Bull. Korean Chem. Soc. 2003,
24, 792.
(9) Safe, S. Crit. Rev. Toxicol. 1990, 21, 51.
(10) Crosby, D. G.; Wong, A. S.; Plimmer, J. R.; Woolson, E. A. Science
1971, 173, 748.
(11) Kasai, E.; Harjanto, S.; Terui, T.; Nakamura, T.; Waseda, Y.
Chemosphere 2000, 41, 857.
(12) Ide, Y.; Kashiwabara, K.; Okada, S.; Mori, T.; Hara, M. Chemosphere 1996, 32, 189.
(13) Hiraoka, M.; Takeda, N.; Kasakura, T.; Imoto, Y.; Tsuboi, H.;
Iwasaki, T. Chemosphere 1991, 23, 1445.
(14) Hilarides, R. J.; Gray, K. A.; Guzzetta, J.; Cortellucci, N.; Sommer,
C. Environ. Sci. Technol. 1994, 28, 2249.
(15) Palauschek N.; Scholz, B. Chemosphere 1987, 16, 1857.
(16) Muto, H.; Shinada, M.; Takizawa, Y. Environ. Sci. Technol. 1991,
25, 316.
(17) Pignatello, J. J.; Huang, L. Q. Water Res. 1993, 27, 1731.
(18) Botre, C.; Memoli, A.; Alhaiqui, F. Environ. Sci. Technol. 1978,
12, 335.
(19) Dobbs, A. J.; Grant, C. Nature 1979, 278, 163.
(20) Colombini, M. P.; Francesco, F. D.; Fuoco, F. Microchem. J. 1996,
54, 331.
(21) Pelizzetti, E.; Borgarello, M. Chemoshere 1988, 17, 499.
(22) Pelizzetti, E.; Minero, C.; Carlin, V.; Borgarello, E. Chemosphere
1992, 25, 343.
2088
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 7, 2004
(23) Choi, W.; Hong, S. J.; Chang, Y. S.; Cho, Y. Environ. Sci. Technol.
2000, 34, 4810.
(24) Dung, M. H.; O’Keefe, P. W. Environ. Sci. Technol. 1994, 28, 549.
(25) Kieatiwong, S.; Nguyen, L. V.; Hebert, V. R.; Hackett, M.; Miller,
G. C. Environ. Sci. Technol. 1990, 24, 1575.
(26) Desideri, A. Boll. Chim. Farm. 1979, 118, 274.
(27) Choudhary, G. G. J. Agric. Food Chem. 1989, 37, 254.
(28) Kim, M.; O’Keefe, P. W. Chemosphere 2000, 41, 793.
(29) Friesen, K. J.; Muir, D. C. G.; Webster, G. R. B. Environ. Sci.
Technol. 1990, 24, 1739.
(30) Dulin, D.; Drossman, H.; Mill, T. Environ. Sci. Technol. 1986,
20, 72.
(31) Crosby, D. G.; Wong, A. S. Science 1977, 195, 1337.
(32) Miller, G. C.; Hebert. V. R.; Miille, M. J.; Mitzel, R.; Zepp, R. G.
Chemosphere 1989, 18, 1265.
(33) Koester, C. J.; Hites, R. A. Environ. Sci. Technol. 1992, 26, 502.
(34) Konstantinov, A.; Bunce, N. J. J. Photochem. Photobiol. A: Chem.
1996, 94, 27.
(35) Epling, G. A.; Qin, Q.; Kumar, A. Chemosphere 1989, 19, 329.
(36) Rayne, S.; Wan, P.; Ikonomou, M. G.; Konstantinov, A. D. Environ.
Sci. Technol. 2002, 36, 1995.
(37) Hung, L. S.; Ingram, L. L., Jr. Bull. Environ. Contam. Toxicol.
1990, 44, 380.
(38) Finlayson-Pitts, B. J.; Pitts, J. N., Jr. Atmospheric Chemistry:
Fundamentals and Experimental Techniques; John Wiley &
Sons: New York, 1986; Chapter 3.
(39) CRC Handbook of Chemistry and Physics; Lide, D. R., Ed.; CRC:
Boca Raton, FL, 1996.
(40) Qin, Z. Chemosphere 1996, 33, 91.
(41) Turro, N. J. Modern Molecular Photochemistry; University
Science Books: Sausalito, CA, 1991; Chapters 5 and 8.
(42) Atkins, P. W. Physical Chemistry; Oxford University Press:
Oxford, UK, 1993; Chapter 27.
(43) Schwarzenbach, R. P.; Gschwend, P. M.; Imboden, D. M.
Environmental Organic Chemistry; John Wiley & Sons: New
York, 1993; Chapter 8.
(44) Lacmann, K.; Maneira, M. J. P.; Moutinho, A. M. C.; Wigman,
U. J. Chem. Phys. 1983, 78, 1767.
(45) Zimmermann, R.; Boesl, U.; Lenoir, D.; Kettrup, A.; Grebner, T.
L.; Neusser, H. J. Int. J. Mass Spectrom. Ion Processes 1995, 145,
97.
(46) Bunce, N. J.; Safe, S.; Ruzo, O. J. Chem. Soc. Perkin I 1975, 1607.
(47) Ohachi, M. O.; Tsujimoto, K.; Seki, K. J. Chem. Soc. Chem.
Commun. 1973, 389, 384.
(48) Andrews, D. P.; Beddard, G. S.; Whitaker, B. J. J. Phys. Chem. A
2000, 104, 7785.
Received for review August 19, 2003. Revised manuscript
received January 14, 2004. Accepted January 20, 2004.
ES034916S