Photolytic method for destruction of dioxins in liquid laboratory waste and identification of the photoproducts from 2,3,7,8-TCDD

Article abstract of DOI:10.1021/es990078j

Analytical and other research laboratories that generate small volumes of dioxin-containing wastes have no convenient method for their disposal. We have used ultraviolet photolysis with a low-pressure mercury lamp to destroy dioxin-like compounds, both as

Full text of DOI:10.1021/es990078j

Environ. Sci. Technol. 2000, 34, 143-148
reductive dechlorination, the yields of dechlorination prod-
Photolytic Method for Destruction of
Dioxins in Liquid Laboratory Waste
and Identification of the
ucts are low (5, 12), and the “missing material” is hitherto
unidentified. Photolysis in sunlight was used to attempt to
destroy 2,3,7,8-TCDD following its accidental release in
Seveso, Italy, in 1976 (13); photolysis in isopropyl alcohol
was used to decontaminate 2,3,7,8-TCDD residues at a former
trichlorophenol manufacturing plant in Verona, MO (14).
Ritterbusch et al. (15, 16) proposed UV treatment of laboratory
waste containing PCDD/ PCDF in hexane, isooctane, and
toluene, using a 150-W mercury lamp but did not characterize
the products completely. No remediation technology can be
put into practice, however, unless all the products are
identified and shown to be less problematic than the original
contaminants. In the present study, we developed a photolytic
method for the destruction of dioxin-like compounds in liquid
laboratory waste using inexpensive and easily available
equipment, identified the photolysis products from 2,3,7,8-
TCDD, and carried out preliminary toxicity assays.
Photoproducts from 2,3,7,8-TCDD
A L E X A N D R E D . K O N S T A N T I N O V ,
A N D R E A M . J O H N S T O N , B R I A N J . C O X ,
J O H N R . P E T R U L I S ,
M O N I K A T . O R Z E C H O W S K I , A N D
N I G E L J . B U N C E *
Department of Chemistry and Biochemistry, University of
Guelph, Guelph, Ontario, Canada N1G 2W1
C O L L E E N H . M . T A S H I R O A N D
B R O C K G . C H I T T I M
Wellington Laboratories, 398 Laird Road,
Guelph, Ontario, Canada N1G 3X7
Experimental Section
Chem icals. Solvents were from Fisher Scientific. OCDD (98%
purity; 17), 2,3,7,8-TCDD, 2,3,7-TrCDD, 2,3-DiCDD, 2-MCDD,
and dibenzo-p-dioxin (all >99% purity) were previously
synthesized in one of our laboratories; 1,2,3,4-TCDD and
decachlorobiphenyl (99% purity) were purchased from Accu-
Standard. 1,6-[3H]-2,3,7,8-TCDD (specific activity 37 Ci/
mmol, >98% purity) was purchased from ChemSyn Labo-
ratories, 2,4,6,7-[3H]estradiol was from Amersham (specific
activity 87 Ci/ mmol), and Cytoscint ES scintillation fluid was
from ICN Pharmaceuticals.
Analytical Methods. Calibration curves for dibenzo-p-
dioxin, 2-MCDD, 2,3-DiCDD, 2,3,7-TrCDD, 1,2,3,4-TCDD,
1,3,6,8-TCDD, 2,3,7,8-TCDD, OCDD, and DCB were con-
structed on a Varian Saturn 3 ion trap MS using hexachlo-
robenzene as the internal standard. The MS was coupled to
a Varian Star 3400CX gas chromatograph, the latter being
equipped with SPI injector and DB-5MS capillary column
(30 m × 0.25 mm × 0.25 µm). The samples were injected in
hexane, with the volumes of injection not exceeding 1.0 µL.
The mass spectra were analyzed using Saturn 4D software
Version 5.2, instrument detection limit for TCDD 10 pg on-
column.
Photolyzed dioxin waste samples were analyzed using a
Micromass VG70SE high-resolution mass spectrometer
(HRMS), coupled with a Hewlett-Packard 5890 series II gas
chromatograph, equipped with a DB-5 capillary column
(J&W, 60m ×0.25 mm ×0.25 µm). The instrumental detection
limits were 0.25 pg for tetra congeners; 1.0 pg for penta,
hexa, and hepta congeners; and 2.0 pg for OCDD and OCDF.
HRMS analyses were performed according to U.S. EPA
Method 1613 protocols (18).
Qualitative mass spectral analyses of the 2,3,7,8-TCDD
photoproducts were performed using a VG Quattro II (Fisons
UK Ltd.) triple quadrupole mass spectrometer equipped with
an atmospheric pressure ion source and MassLynx software
package. The sample solutions were introduced into the mass
spectrometer via a 10-µL Rheodyne 7010 injection valve. The
mobile phase [50/ 50 v/ v Nanopure water (Barnstead) and
acetonitrile (Caledon)] was delivered using a Hewlett-Packard
1090 series II/ L binary LC pump at 15 µL/ min. The mass
spectra were acquired and averaged over at least 8 scans in
multichannel analysis (MCA) data acquisition mode by
scanning the first quadrupole in 0.1 amu increments from
m/ z 100 to 800 in 1.2 s.
Analytical and other research laboratories that generate
small volumes of dioxin-containing wastes have no convenient
method for their disposal. We have used ultraviolet
photolysis with a low-pressure mercury lamp to destroy dioxin-
like compounds, both as individual congeners and in
actual waste analytical samples, down to nondetect levels.
Photolysis promises to be an efficient, safe, and inexpensive
method for on-site treatment of liquid laboratory wastes
that are contaminated by dioxin-like compounds, allowing
the treated materials to be discarded as regular organic
solvent waste. Experiments with 1,6-[3H]-2,3,7,8-TCDD revealed
that the principal photolytic pathway involves cleavage
of C-O bonds rather than C-Cl bonds, giving chlorinated
hydroxydiphenyl ethers as the initial products and
accounting for the low material balances of reductive
dechlorination products previously found upon photolysis
of PCDDs. The photolysis products from 2,3,7,8-TCDD do not
bind to either the Ah receptor or the estrogen receptor
in vitro, making it unlikely that the products from UV treatment
of PCDD/PCDF in laboratory waste will show either Ah
or estrogen receptor-mediated toxicological effects.
Introduction
Halogenated aromatic compounds (HACs), including poly-
chlorinated dibenzo-p-dioxins (PCDDs), polychlorinated
dibenzofurans (PCDFs), and polychlorinated biphenyls
(PCBs), are well-known toxic pollutants that are the subject
of intense analytical and toxicological research interest (1,
2). Numerous laboratories around the world generate small
amounts of waste containing these compounds, creating a
need for an inexpensive, safe, simple, and efficient method
of on-site detoxification.
PCDDs and PCDFs decompose under UV irradiation (3,
4), a reaction that is more efficient in hydrogen donor solvents
(5, 6) than in acetonitrile/ water mixtures (7-10). Although
this suggests that photodegradation may occur by sequential
Apparatus. The reaction vessel was a standard 100-mL
three-neck round-bottom flask equipped with a magnetic
* Corresponding author phone: (519)824-4120, ext. 3962; fax:
(519)766-1499; e-mail: bunce@chembio.uoguelph.ca.
9
10.1021/es990078j CCC: $19.00
Published on Web 12/02/1999
2000 Am erican Chem ical Society
VOL. 34, NO. 1, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 1 4 3

TABLE 1. Photolysis of High-Strength Dioxin Waste
concentration, mol L-1 (conversion, %)
TCDD, 322a
time, h
Cl5 PCB, 326a
TriCDF, 272a
PeCDF, 340a
PeCDD, 356a
0
1.5
4
24
27
1.53 × 10-4 (0.0)
6.61 × 10-6 (95.7)
2.36 × 10-7 (99.8)
ND (100)
3.68 × 10-4 (0.0)
3.62 × 10-4 (1.6)
9.37 × 10-5 (74.6)
1.13 × 10-6 (99.7)
ND (100)
1.88 × 10-4 (0.0)
1.75 × 10-4 (6.4)
8.07 × 10-5 (57.0)
1.35 × 10-6 (99.3)
ND (100)
1.18 × 10-4 (0.0)
1.02 × 10-5 (91.3)
NDb (100)
ND (100)
ND (100)
1.40 × 10-5 (0.0)
8.35 × 10-6 (40.5)
2.33 × 10-6 (83.4)
ND (100)
ND (100)
ND (100)
a
b
Com pound and ion m onitored, m /z. ND, not detected.
TABLE 2. Yields of Dechlorination Products in Photolysis of 2,3,7,8-TCDD
-1
concentration, mol L
dioxin congener
0a,b
4a
10a
20a
2,3,7,8-TCDD
2,3,7-TriCDD
2,3-DiCDDc
1.18 × 10-5
4.22 × 10-8
1.44 × 10-8
1.64 × 10-8
1.56 × 10-8
1.19 × 10-5
0
7.44 × 10-6
1.97 × 10-7
ND
1.84 × 10-6
1.13 × 10-7
ND
2.16 × 10-7
5.63 × 10-8
ND
2-MCDD
1.64 × 10-8
1.56 × 10-8
7.67 × 10-6
36.8
ND
ND
1.64 × 10-8
dibenzo-p-dioxin
total concn, M
conversion, %
total dechlorination
products, M (%)
ND
1.95 × 10-6
2.88 × 10-7
98.2
84.4
8.86 × 10-8
2.29 × 10-7
1.13 × 10-7
(0.96)
7.27 × 10-8
(0)
(1.95)
(0.62)
a
b
c
Irradiation tim e, m in. 0.75% of less chlorinated dioxins were present originally. Includes 2,6-DiCDD and 2,7-DiCDD, which were assum ed
to have the sam e detector response as 2,3-DiCDD.
stirrer. The three necks were provided with a reflux condenser
sealed with a rubber septum and a balloon and needle to
relieve gas pressure, a rubber septum for taking samples
during the irradiations, and a quartz sleeve attached to a
commercially available T24 quartz male ground joint. A
PenRay 4.6-W low-pressure mercury lamp was placed in the
quartz sleeve.
Photolyses. For single congeners, a hexane solution (25.0
mL) of concentration ∼100 µM was irradiated at 254 nm;
0.5-mL aliquots were withdrawn at appropriate intervals,
mixed with 0.5 mL of a hexane solution of the internal
standard, and analyzed by GC-MS.
Two laboratory waste samples were processed; one
containing high levels and the other containing low levels of
PCDD and PCDF. The ‘high’ concentration waste sample
(5-100 µg/ mL) contained a mixture of waste standards in
isooctane, principally 3,3′,4,4′,5-pentachlorobiphenyl, 1,4,7,8-
TCDD, 1,2,3,7,8-PeCDD, an unknown trichlorodibenzofuran,
and an unknown pentachlorodibenzofuran (Table 1). This
solution was irradiated at 254 nm, with 0.5 mL withdrawn
after 1.5, 4, 24, and 27 h.
The ‘low’ concentration waste sample (in isooctane) was
prepared by combining approximately 600 cleaned-up
extracts of environmental samples that had been archived
awaiting disposal. It contained native PCDD/ PCDF (mostly
from incineration/ combustion sources; variable concentra-
tions with maximum 1 ng/ mL) and also 13C12-labeled PCDD/
PCDF (≈10 ng/ mL). A total of 30 mL of the waste sample was
photolyzed, with 1.0-mL aliquots removed after 1.5, 3.5, 6,
and 27 h for analysis by HRMS. The concentrations of PCDD/
PCDF (congener group totals) were determined by peak area
comparison to U.S. EPA Method 1613 calibration standards
using the external standard method.
yield by the Sandmeyer reaction to 2-hydroxy-4,4′,5′-trichlo-
rodiphenyl ether, which gave 1 in 90% yield (20 mg) as a
colorless solid upon overnight stirring over SO2Cl2: mp 71
1
°C, purity 98% by GC-MS. H NMR (δ, CDCl3), 5.58, s, 1H;
4
4
6.90, dd, J ) 8.8 Hz, J ) 2 Hz, 1H; 6.95, s, 1H; 7.14, d, J )
2 Hz, 1H; 7.17, s, 1H; 7.44, d, J ) 8.8 Hz, 1H. 13C NMR (δ,
CDCl3), 117.6, 118.2, 120.0, 120.2, 123.6, 128.2, 128.6, 131.5,
133.8, 142.0, 146.5, 154.8. MS (EI): 326 (33), 324 (100), 322
(69), 254 (16), 252 (26) 223 (6), 177 (8), 146 (25), 109 (14).
(b) 2,2′-Dihydroxy-4,4′,5,5′-tetrachlorobiphenyl (2). 2
was obtained in four steps from 3,4-dichlorophenol, which
was treated with iodine in boiling water in the presence of
NaOH to give 2-iodo-4,5-dichlorophenol (33% yield). This
was converted to 2-iodo-4,5-dichloroanisole by reflux with
methyl iodide in acetone in the presence of K2CO3 (90% yield),
followed by Ullmann coupling in a sealed ampule at 220 °C
in the presence of 40 molar excess of copper-bronze powder
to give 2,2′-dimethoxy-4,4′,5,5′-tetrachlorobiphenyl in 14%
yield (isolated). Treatment of the latter compound with BBr3
in dry CH2Cl2 at -20 °C gave 2,2′-dihydroxy-4,4′,5,5′-
tetrachlorobiphenyl (2) in 70% yield (8 mg) as a colorless
solid, mp 178-179 °C (decomp), no impurities detected by
1H NMR: (δ, CD3OD), 7.05, s, 2H; 7.33, s, 2H. 13C NMR (δ,
CD3OD), 118.4, 123.0, 125.7, 132.7, 133.6, 155.5.
Photodechlorination of 2,3,7,8-TCDD. Two milliliters of
1.2 × 10-5 mol L-1 solution of 2,3,7,8-TCDD in hexane in an
8 mm Pyrex tube, sealed with a rubber septum, were
photolyzed at 300 nm in a Rayonet photoreactor. Aliquots
(200 µL) were removed after 4, 10, and 20 min; mixed with
200 µL of HCB (5 × 10-6 mol L-1 in hexane); and analyzed
by GC-MS to give the results shown in Table 2.
In the photolysis of 2,3,7,8-TCDD spiked with 1,6-[3H]-
2,3,7,8-TCDD, the stock solution of 1,6-[3H]-2,3,7,8-TCDD
in hexane (10 µL) had activity 2.21 ( 0.11 × 104 DPM (nominal
concentration 3.7 × 10-8 mol L-1). A solution of 2,3,7,8-TCDD
(9.32 × 10-5 mol L-1, 0.9 mL) was placed into each of two 8
mm Pyrex tubes. Hexane (0.1 mL) was added to one tube,
and 0.1 mL of the stock solution of 1,6-[3H]-2,3,7,8-TCDD in
hexane was added to the other. The tubes were capped with
rubber septa, placed in a Rayonet merry-go-round photo-
Syntheses. (a) 2-Hydroxy-4,4′,5,5′-tetrachlorodiphenyl
ether (1). 1 was synthesized according to the method of
Humpi (19) by condensation of 2,4-dichloronitrobenzene
with 3,4-dichlorophenol (78% yield) followed by reduction
of the resulting 2-nitro-4,4′,5′-trichlorodiphenyl ether with
Sn/ AcOH/ HCl to 2-amino-4,4′,5′-trichlorodiphenyl ether
(95% yield). The amino compound was converted in 25%
9
1 4 4 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 1, 2000

reactor, and irradiated at 300 nm. Three 10-µL aliquots were
taken with a gas-tight 25-µL syringe through the rubber
septum from the tube containing 1,6-[3H]-2,3,7,8-TCDD
before and after 3 h irradiation, by which time the conversion
of unlabeled 2,3,7,8-TCDD had reached 99.9% (GC-MS of
the unlabeled sample). Each aliquot of the radiolabeled
sample was analyzed by scintillation counting after adding
2.0 mL of scintillation cocktail. The radiolabeled samples
were also separated by HPLC (Perkin-Elmer model 250
isocratic LC pump, Rheodyne model 7010 injector equipped
with 20-µL sample loop, Waters µ-Bondapack C18 3.9 × 300
mm column, and Gilson model 202/ 204 fraction collector,
pure methanol as the mobile phase). The fractions were
collected directly into scintillation vials and counted (Beck-
man LS 7000 scintillation counter). Chromatograms were
obtained by plotting the radioactivity of each fraction vs
elution time. Wipe tests were performed after all experiments
with radiochemicals.
Hydroxylapatite Assay with the Products of Photolysis
of 2,3,7,8-TCDD. A solution of 2,3,7,8-TCDD in hexane (2.40
mL, 7.8 × 10-5 mol L-1) was placed in an 8 mm Pyrex tube
and irradiated at 300 nm in the merry-go-round photoreactor.
Aliquots (200 µL) taken after 0, 0.5, 1, 2, 4, 8, 12, 20, 35, and
60 min were placed in 1-mL sample vials, then 25 µL was
drawn from each aliquot and mixed with 25 µL of 5 × 10-6
HCB in hexane prior to analysis by GC-MS (triplicate
injections). The remaining 175 µL of hexane was carefully
removed under a slow stream of nitrogen, and each residue
was redissolved in 50 µL of DMSO (Fisher, Spectranalyzed
Grade) and then diluted 1000-fold before use.
Long-Evans rats as described previously (22). The HPLC
analysis was as described previously, except that a Pheno-
menex Biosap S400 size exclusion column (300 mm × 7.8
mm) was used with flow rate of mobile phase (25 mM
phosphate, 1.5 mM EDTA, 100 mM KCl, 10% glycerol, pH
7.1) of 0.5 mL/ min. For each chromatogram, 0.5-mL fractions
were collected between 6.5 and 12 min after the injection.
Results and Discussion
Photolysis of Waste. Our objective was to develop meth-
odology for the removal of dioxins to below the detection
limits of high resolution GC-MS, using inexpensive, com-
mercially available equipment with minimal custom modi-
fication. We wanted the equipment to be compact, easily
assembled, and able to operate unattended in a laboratory
fume hood.
We chose a 254-nm PenRay lamp as the UV source. The
advantages of low-pressure mercury lamps over the high- or
medium-pressure mercury lamps proposed previously (15)
include low power (4.6 W), high photon efficiency, and low
operating temperature (<60 °C) to avoid the need for external
water cooling. Insertion of the light source in a quartz sleeve
avoided deposition of materials on the surface of the lamp
and completely circumvented heating the reaction mixture
(the temperature inside the reactor never exceeded ambient
by 3 °C). Evaporative solvent loss during overnight photolyses
in hexane was avoided by sealing the reflux condenser with
a septum (equipped with a balloon for pressure relief). The
fire hazard was minimal, and the equipment could run
unattended in a fume hood.
The assay procedure followed that of Gasiewicz and Neal
(20) with only minor modifications. Hepatic cytosol from
immature Sprague-Dawley rats was prepared as described
previously (21). Aliquots of rat liver cytosol were thawed and
diluted to 2.0 mg/ mL with HEGD buffer (1 mM N-2-
hydroxyethylpiperazine-N′-2-ethanesulfonic acid, 1 mM
EDTA sodium salt, 1 mM dithioerythritol, 10% v/ v glycerol,
pH 7.6). One milliliter of diluted cytosol was added in
quadruplicate to 10 µL of 1 × 10-7 mol L-1 1,6-[3H]-2,3,7,8-
TCDD and 10 µL of each (diluted) photolyzed solution and
incubated for 45 min at 23 °C. Then 0.20 mL was withdrawn
and incubated for 10 min on ice with 0.25 mL of freshly
washed hydroxylapatite slurry in ice-cold HEGD buffer. One
milliliter of ice-cold HEGD buffer containing 1% Triton X-100
surfactant was added; after mixing, the mixture was cen-
trifuged at 2000g for 2 min, and the supernatant was
discarded. After two more similar washings, the HAP pellet
was transferred quantitatively to a plastic 20-mL scintillation
vial with 3 × 0.75 mL of ethanol. Ten milliliters of scintillation
cocktail was added, and the radioactivity was counted.
Aliquots of cytosol incubated with [3H]TCDD alone gave a
value for “total binding”, while samples incubated with [3H]-
TCDD and a 200-fold excess of TCDF were used to determine
“nonspecific” binding. Quadruplicate values of the activities
of the samples (sample binding), measured by scintillation
counting, were averaged, and the percent specific binding
of [3H]TCDD to the Ah receptor was calculated using
Preliminary photolyses involved individual dioxin con-
geners in hexane. Decachlorobiphenyl (DCB) was also
included in the trials to demonstrate the applicability of the
method for detoxification of PCB-containing waste. As shown
in Figure 1, irradiation of 1,2,3,4-TCDD (1.0 × 10-5 mol L-1
in hexane), 1,3,6,8-TCDD (1.3 × 10-4 mol L-1), OCDD (5.0 ×
10-5 mol L-1), and DCB (3.0 × 10-5 mol L-1) gave >99%
conversion in <2 h. In every case, the concentrations of
starting material and secondary products dropped below the
detection limit of Saturn 3 GC-MS within 5 h. Pseudo-first-
order kinetics were followed for the disappearance of each
compound, consistent with low light absorption [Iabs ) I0-
(1-10-Abs), which is directly proportional to substrate
concentration at low absorbance]. No attempt was made to
determine quantum yields because they were not the
objective of this project.
Some analytical wastes of dioxin-like compounds are
generated in toluene, whose high absorbance at 254 nm out-
competes HACs for incident radiation. We considered
whether photolysis might still proceed through energy
transfer from toluene to the chlorinated substrates [triplet
energy of toluene ) 347 kJ mol-1, those of chlorinated
benzenes (model for PCDDs) are 335-345 kJ mol-1, and that
of biphenyl (model for DCB) is 274 kJ mol-1 (23)]. In practice,
toluene was unsatisfactory as a solvent: 1,3,6,8-TCDD (2.3
× 10-5 mol L-1) was unchanged after 3 h of irradiation in
toluene, and DCB was almost unreactive. In toluene-hexane
mixtures, the photolysis of 1,3,6,8-TCDD slowed progressively
with the proportion of toluene: in 1% toluene, 99.8%
conversion was achieved after 18 h, while 13% of 1,3,6,8-
TCDD remained after 31 h in 10% toluene. An attempt to
assist the photolysis by adding triethylamine as an electron
donor (24) to toluene was unsuccessful. We conclude that
PCDD wastes generated in toluene should be solvent-
exchanged into an alkane solvent ahead of photolysis; a few
percent of residual toluene can be tolerated at the expense
of longer irradiation times.
% specific binding )
[sample (DPM) - nonspecific (DPM)]
[total (DPM) - nonspecific (DPM)]
× 100
Estrogen Receptor Assay with the Products of Photolysis
of 2,3,7,8-TCDD. The gel filtration chromatographic method
for determining relative estrogenic binding affinities devel-
oped by Cox and Bunce (22) was used without alterations,
using 2,4,6,7-[3H]estradiol (1.5 ×10-8 mol L-1) as the reference
radioligand. Five samples of the photolyzed 2,3,7,8-TCDD
were used in the assay, using 10-µL aliquots of the (undiluted)
DMSO solutions. Rat liver cytosol was obtained from female
Two types of samples of liquid waste were tested. A “high-
concentration” sample contained 5-100 µg/ mL each of
1,4,7,8-TCDD and 1,2,3,7,8-PeCDD and unspecified pen-
9
VOL. 34, NO. 1, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 1 4 5

FIGURE 1. Disappearance of 1,2,3,4-TCDD, 1,3,6,8-TCDD, OCDD, and DCB during photolysis at 254 nm.
tachlorobiphenyl, trichlorodibenzofuran, and pentachlo-
rodibenzofuran. The mixture was irradiated at 254 nm; after
27 h, the concentrations of all starting materials and
chlorinated intermediates dropped below the level of detec-
tion of the Saturn 3 GC-MS (Table 1). The “low-concentra-
tion” waste was a mixture of ca. 600 environmental samples,
mostly extracts of PCDD/ PCDF from fly ash, mixed with 13C-
labeled standards. It contained 5-100 ng/ mL of PCDD/ PCDF,
which was below the detection limits of the Saturn 3 GC-
MS, so the samples were analyzed by HRMS at 10000
resolution. After 27 h, all analytes were below the instrumental
detection limit (Figure 2).
Products and Mechanism of PCDD Photolysis. Practical
implementation of photolytic waste treatment requires
detailed product studies to ensure that the procedure will
not form harmful contaminants. Table 2 shows that PCDD/
PCDF give only low yields of photodechlorination products
in hydrogen donor solvents, consistent with many previous
studies (4, 25-27). No other products were observed by GC-
MS, casting doubt on the suggestion that the major photolysis
pathway of PCDDs involves successive C-Cl homolysis (28).
Previous explanations for the low material balance include
PCDD heterolysis to yield a carbene that reacts with a second
molecule of the parent molecule (29) and the formation of
polychlorobenzenes during photolysis of 1,2,3,7,8-PeCDD
in CCl4 (30); neither has any precedent in the photochemistry
of haloaromatic compounds (25).
After photolysis to 99.9% conversion of 2,3,7,8-TCDD
spiked with 1,6-[3H]-2,3,7,8-TCDD, a 10-µL aliquot exhibited
almost equal radioactivity as an unirradiated control (3165
( 110 vs 3380 ( 60 dpm). This showed that the photolysis
products remained in solution rather than adsorbing to the
reaction vessel, precipitating, or volatilizing. No significant
peaks were seen upon HPLC analysis of an irradiated solution
of unlabeled 2,3,7,8-TCDD, but scintillation counting of
fractions from 20 µL of irradiated 1,6-[3H]-2,3,7,8-TCDD
separated by C18 reverse-phase HPLC revealed products with
short retention times (Figure 3). These more polar products
were observed by conventional HPLC analysis of unlabeled
TCDD (UV absorbance detector, λ ) 227 nm) when the
solution was concentrated 500 times.
by scintillation counting and accounted for ∼90% of the total
radioactivity of the starting material. Analysis by negative
ion ES-MS, which detects quasi-molecular ions at m/ z values
corresponding to (M - H)- showed three major peaks in
fraction A (m/ z 271, 303, and 321; ratio 1:3:4; with isotope
patterns indicating 3, 3, and 4 chlorines, respectively). The
parents corresponding to these quasi-molecular ions were
C12H7Cl3O (m/ z 271), C12H7Cl3O3 (m/ z 303), and C12H6Cl4O2
(m/ z 321). All these compounds must be phenolic, with the
protons lost from hydroxyl groups. Only m/ z 321 appeared
in the major fraction B, which could be either 2-hydroxy-
3′,4,4′,5-tetrachlorodiphenyl ether (1) or 2,2′-dihydroxy-
4,4′,5,5′-tetrachlorobiphenyl (2).
Preparation of authentic samples indicated 1 rather than
2 as the likely major product based on HPLC retention time.
C-O rather than C-Cl homolysis is therefore the major
pathway in the photolysis of dioxins, explaining the low
dechlorination yields. Literature precedents to support C-O
bond cleavage include 10% of 2 (as the dimethyl ether, after
diazomethane treatment) upon photolysis of 2,3,7,8-TCDD
in isooctane; however >80% of the starting material was
undetected (5). Photolysis of dibenzo-p-dioxin (31, 32) gives
2,2′-dihydroxybiphenyl and 4-hydroxydibenzofuran (33, 34);
diphenyl ether photorearranges to o- and p-phenylphenol
(35).
Receptor Binding Assays of the Products of Photolysis
of 2,3,7,8-TCDD. We carried out two bioassays to investigate
whether the products of photolysis of PCDD/ PCDF might
pose a threat of toxicity. Binding to the Ah receptor protein
was studied, since it is well-known that the major toxic
responses to dioxins are Ah receptor-mediated (1, 2). Product
mixtures from various stages of photolysis of 2,3,7,8-TCDD
in hexane were incubated with 1,6-[3H]-2,3,7,8-TCDD and
rat hepatic cytosol containing the Ah receptor. The radio-
labeled TCDD competed with residual unphotolyzed (un-
Fractions of the radiolabeled material eluting after 3-4
min (fraction A) and 4-6 min (fraction B), collected from
five injections on the analytical column, had mole ratios 1:3.3
9
1 4 6 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 1, 2000

FIGURE 2. Change in concentrations (congener totals) of PCDD (a) and PCDF (b) during photolysis of the low-strength waste at 254 nm.
TCDD was in large excess over the radioligand, and specific
binding was low. Specific binding increased with the percent
photoconversion of TCDD, eventually reaching 100% at 100%
conversion, when no unlabeled TCDD remained to compete
with the radioligand (Figure 4). Because the reaction products
at 100% conversion have no Ah receptor affinity, they will
not exhibit dioxin-like toxicity.
The increase of specific binding lagged behind the percent
conversion at intermediate stages of conversion (Figure 4).
FIGURE 3. HPLC radiochromatogram from photolysis of 1,6-[3H]-
2,3,7,8-TCDD in hexane at 300 nm: mobile phase, 100% MeOH flow
rate, 1 mL/min.
The binding affinities of 1 and 2 relative to unlabeled 2,3,7,8-
TCDD were 3.4 × 10-4 and 5.7 × 10-5, respectively; hence,
1 could contribute to the Ah receptor binding at intermediate
stages of photolysis. The relative binding affinities of a group
of chlorinated diphenyl ethers were in the same range as 1:
2,3′,4′,6-tetrachloro-, 1.2 × 10-4; 3,3′,4,4′-tetrachloro-, 1.0 ×
10-2; 3,3′,4,4′,5-pentachloro-, 1.4 × 10-2; 2,3′,4,4′,5-pen-
tachloro, 7.7 × 10-3. The 3,3′,4,4′-chlorines, corresponding
labeled) TCDD and its photolysis products for a fixed aliquot
of Ah receptor. In the absence of unlabeled TCDD, the
radioligand occupied all the receptor binding sites, and
maximum protein-bound radioactivity was observed (100%
specific binding). In the unphotolyzed sample, the unlabeled
9
VOL. 34, NO. 1, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 1 4 7

(3) Nestrick, T. J.; Lamparski, L. L.; Townsend, D. I. Anal. Chem.
1980, 52, 1865.
(4) Dobbs, A. J.; Grant, C. Nature 1979, 278, 163.
(5) Dulin, D.; Drossman, H.; Mill, T. Environ. Sci. Technol. 1986,
20, 72.
(6) Hung, L. S.; Ingram, L. L., Jr. Bull. Environ. Contam. Toxicol.
1989, 44, 380.
(7) Choundry, G. C.; Webster, G. R. B. Chemosphere 1985, 14, 893.
(8) Choudhry, G. G.; Webster, G. R. B. J. Agric. Food Chem. 1989,
37, 254.
(9) Choundry, G. C.; Webster, G. R. B. Chemosphere 1986, 15, 1935.
(10) Hung, L. S.; Ingram, L. L., Jr. Bull. Environ. Contam. Toxicol.
1989, 44, 380.
(11) Deleted in proof.
(12) Kleatwong, S.; Nguyen, L. V.; Herbert, V. R.; Hackell, M.; Miller,
G. Environ. Sci. Technol. 1990, 24, 1575.
(13) Mill, T. In Dioxins in the Environment; Kamrin, M. A., Rodgers,
P. W., Eds; Hemisphere: Bristol, PA, 1985; p 173.
(14) Esposito, M. T.; Tiernan, T. O.; Dryden, F. E. Dioxins; U.S. EPA
Report EPA-600/ 2-80-197; U.S. Government Printing Office:
Washington, DC, 1980; p 257.
FIGURE 4. Relationship between percent specific binding of 1,6-
[3H]-2,3,7,8-TCDD to rat hepatic Ah receptor and the photolytic
conversion of 2,3,7,8-TCDD.
(15) Ritterbusch, J.; Lorenz, W.; Bahadir, M. Chemosphere 1994, 29,
457.
(16) Ritterbusch, J.; Lorenz, W.; Bahadir, M. Chemosphere 1994, 29,
1829.
(17) Kulka, M. Can. J. Chem. 1961, 39, 1973.
(18) U.S. Environmental Protection Agency. Method 1613; Tetra-
through Octachlorinated Dioxins and Furans by Isotope Dilution
HRGC/HRMS; EPA 821-B-94-005, Revision B; U.S. Government
Printing Office: Washington, DC, October 1994.
(19) Humpi, T. Synthesis 1985, 919.
(20) Gasiewicz, T. A.; Neal, R. A. Anal. Biochem. 1982, 124, 1.
(21) Schneider, U. A.; Brown, M. M.; Logan, R. A.; Millar, L. C.; Bunce,
N. J. Environ. Sci. Technol. 1995, 29, 2595.
(22) Cox, B. J.; Bunce, N. J. Anal. Biochem. 1999, 267, 357.
(23) Murov, S. L.; Carmichael, I.; Hug, G. L. Handbook of Photo-
chemistry, 2nd ed.; Marcel Dekker: New York, 1993; Table 1, pp
4-53.
(24) Konstantinov, A.; Bunce, N. J. J. Photochem. Photobiol., A: Chem.
1996, 94, 27.
FIGURE 5. Specific and nonspecific binding of [3H]estradiol for the
estrogen receptor when competed against photolysis products
2,3,7,8-TCDD.
(25) Bunce, N. J. In Handbook of Organic Photochemistry; Horspool,
W. M., Ed.; CRC Press: Boca Raton, FL, 1995; Chapter 86.
(26) Crosby, D. G.; Wong, A. S. Science 1977, 195, 1337.
(27) Nash, R. G.; Beall, M. L., Jr. J. Agric. Food Chem. 1980, 28, 614.
(28) Stormer, K. A. In Detoxication of Hazardous Waste; Exner, J. H.,
Ed.; Ann Arbor Science: Ann Arbor, MI, 1982; p 280.
(29) Mamantov, A. Chemosphere 1985, 14, 897.
(30) Minghui, Z.; Zhicheng, B.; Xiaobai, X.; Keou, W. Chemosphere
1996, 32, 603.
(31) Guan, B.; Wan, P. J. Chem. Soc., Chem. Commun. 1993, 409.
(32) Guan, B.; Wan, P. J. Photochem. Photobiol., A: Chem. 1994, 80,
199.
(33) Plimmer, J. R.; Klingebiel, U. I.; Crosby, D. G.; Wong, A. S. In
Chlorodioxins, Origin and Fate; Blair, E. H., Ed.; American
Chemical Society: Washington, DC, 1973; p 49.
(34) Masse´, R.; Pelletier, B. Chemosphere 1987, 16, 7.
(35) Ogat, Y.; Tagai, K.; Ishino, I. Tetrahedron 1970, 26, 2703.
(36) Jobling, S.; Reynolds, T.; White, R.; Parker, M.; Sumpter, J.
Environ. Health Perspect. 1995, 103, 582.
to the 2, 3, 7, and 8 chlorines of PCDDs, impart moderately
strong Ah receptor binding activity, which is somewhat
diminished by the presence of the polar 2-OH group of 1.
Knowing that the major primary products of PCDDs are
hydroxylated, the second bioassay was a test of estrogenic
activity, which has been documented for numerous phenolic
compounds (36, 37), including hydroxylated PCBs (38-40).
The photolysis products from 2,3,7,8-TCDD were competed
against [3H]estradiol for occupation of the estrogen receptor
from female Long-Evans rat liver. Gel filtration chromatog-
raphy (22) allowed the direct determination of percent specific
binding vs extent of photoreaction. Samples taken at 0, 13,
43, 88, and 99% conversion of 2,3,7,8-TCDD showed no
significant decrease in the area of specific binding with
increasing conversion (Figure 5). Thus, the photolysis
products possess neither dioxin-like nor estrogenic activity,
indicating that dioxin wastes that have been photolyzed to
nondetect levels can be treated as regular organic waste.
(37) Routledge, E. J.; Sumpter, J. P. J. Biol. Chem. 1997, 272, 3280.
(38) Moore, M.; Mustain, M.; Daniel, K.; Chen, I.; Safe, S.; Zacharewski,
T.; Gillesby, B.; Joyeux, A.; Balaguer, P. Toxicol. Appl. Pharmacol.
1997, 142, 160.
(39) Safe, S.; Washburn, K.; Zacharewski, T.; Phillips, T. Chemosphere
1995, 31, 3017.
(40) Connor, K.; Ramamoorthy, K.; Moore, M.; Mustain, M.; Chen,
I.; Safe, S.; Zacharewski, T.; Gillesby, B.; Joyeux, A.; Balaguer, P.
Toxicol. Appl. Pharmacol. 1997, 145, 111.
Acknowledgments
We thank Dr. C. D. Metcalfe (Trent University) for the gift
of chlorinated diphenyl ethers and the Natural Sciences and
Engineering Research Council of Canada for partial support.
Literature Cited
Received for review January 22, 1999. Revised manuscript
received August 30, 1999. Accepted October 29, 1999.
(1) Safe, S. Crit. Rev. Toxicol. 1994, 24, 87.
(2) Rowlands, J. C.; Gustafsson, J.-A. Crit. Rev. Toxicol. 1997, 27,
109.
ES990078J
9
1 4 8 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 1, 2000