Natural formation of chlorinated phenols, dibenzo-p-dioxins, and dibenzofurans in soil of a Douglas fir forest

Article abstract of DOI:10.1021/es9900104

The natural formation of 4-MCP, 24/25- and 26-DCP, and 245-TrCP was detected in four selected areas of a rural Douglas fir forest where the humic layer was spiked in situ with a solution of Na³⁷Cl and covered by an enclosure, after 1 year of incubation. Chlorinated phenols (CP) can be formed naturally from organic matter and inorganic chloride by either de novo synthesis or chloroperoxidase (CPO)-catalyzed chlorination. The natural CP congeners were found to be present in high concentrations in soil compared to the other congeners, except for 245-TrCP which was present in a relatively low concentration. This study did not reveal which source, natural or anthropogenic, caused the observed concentrations. Some 20 chlorinated dibenzo-p-dioxins and dibenzofurans (CDD/F) were found to be formed naturally in soil of the Douglas fir forest; the formation of three 2,3,7,8-substituted congeners, 2378-TeCDD, 12378-PeCDD, and 123789-HxCDD, deserves special attention. A formation mechanism has been proposed which starts from naturally formed CP congeners and which probably involves peroxidase mediation. Chlorination of CDD/F congeners by the CPO-mediated reaction cannot be ruled out, but seems to be less likely due to the absence of several predicted congeners. The natural formation of 4-MCP, 24/25- and 26-DCP, and 245-TrCP was detected in four selected areas of a rural Douglas fir forest where the humic layer was spiked in situ with a solution of Na³⁷Cl and covered by an enclosure, after 1 year of incubation. Chlorinated phenols (CP) can be formed naturally from organic matter and inorganic chloride by either de novo synthesis or chloroperoxidase (CPO)-catalyzed chlorination. The natural CP congeners were found to be present in high concentrations in soil compared to the other congeners, except for 245-TrCP which was present in a relatively low concentration. This study did not reveal which source, natural or anthropogenic, caused the observed concentrations. Some 20 chlorinated dibenzo-p-dioxins and dibenzofurans (CDD/F) were found to be formed naturally in soil of the Douglas fir forest; the formation of three 2,3,7,8-substituted congeners, 2378-TeCDD, 12378-PeCDD, and 123789-HxCDD, deserves special attention. A formation mechanism has been proposed which starts from naturally formed CP congeners and which probably involves peroxidase mediation. Chlorination of CDD/F congeners by the CPO-mediated reaction cannot be ruled out, but seems to be less likely due to the absence of several predicted congeners.

Full text of DOI:10.1021/es9900104

Environ. Sci. Technol. 1999, 33, 2543-2549
of polychlorinated dibenzo-p-dioxins and dibenzofurans
Natural Formation of Chlorinated
Phenols, Dibenzo-p-dioxins, and
Dibenzofurans in Soil of a Douglas
Fir Forest
two decades ago (9). Therefore, natural (forest) fires have
been suspected to contribute significantly to the global
CDD/ F emission (10). Harrad and Jones (11), on the other
hand, argued that this is not true for the United Kingdom.
Actually, next to fires, there could be another natural
source for CDD/ F in the environment. O¨ berg et al. (12)
showed the formation of several CDD/ F congeners, including
some of the 2,3,7,8-substituted congeners, in the horseradish
peroxidase-mediated reaction of 2,4,5-trichlorophenol (245-
TrCP) and hydrogen peroxide. In addition, Wagner et al. (13)
observed PeCDD/ F and HxCDD/ F formation in a mixture of
E D D O J . H O E K S T R A , * ,
†
†
H E N K D E W E E R D ,
‡
E D W . B . D E L E E R , A N D
§
U D O A . T H . B R I N K M A N
2
45-TrCP, hydrogen peroxide, and a culture filtrate of the
TNO Environment, Energy and Process Development,
Apeldoorn, The Netherlands, NMi Van Swinden Laboratory,
Delft, The Netherlands, and Department of Analytical
Chemistry, Free University, Amsterdam, The Netherlands
soil fungus Phanerochaete chrysosporium which showed
peroxidase (PO) activity. PeCDD/ F and HxCDD/ F formation
was also demonstrated in a mixture of 245-TrCP, hydrogen
peroxide, and a fresh milk extract.
O¨ berg et al. (14) reported the formation of 13C-CDD/ F,
mainly 13C-HpCDD and C-OCDD, from 13C-PCP in activated
sewage sludge. O¨ berg et al. (15) observed an up to 3-fold
increase of CDD/ F congeners expressed as international toxic
equivalents (I-TEQ) during the composting of uncontami-
nated garden waste. De novo formation of CDD/ F was held
responsible because no contamination could have occurred
and loss of biomass could not explain the increase of tox-
icity. However, PCP did not appear to be a precursor for
CDD/ F in 13C-PCP-spiking experiments with garden waste.
The role of PO in the synthesis of CDD/ F during composting
is not clear yet either. Although several bacteria and white-
rot fungi that excrete PO are active in composting processes,
chemical oxidation could also occur at the reported tem-
peratures of 60-70 °C.
13
The natural formation of 4-MCP, 24/25- and 26-DCP, and
245-TrCP was detected in four selected areas of a rural
Douglas fir forest where the humic layer was spiked in
37
situ with a solution of Na Cl and covered by an enclosure,
after 1 year of incubation. Chlorinated phenols (CP) can
be formed naturally from organic matter and inorganic chloride
by either de novo synthesis or chloroperoxidase (CPO)-
catalyzed chlorination. The natural CP congeners were found
to be present in high concentrations in soil compared to
the other congeners, except for 245-TrCP which was present
in a relatively low concentration. This study did not
reveal which source, natural or anthropogenic, caused
the observed concentrations. Some 20 chlorinated dibenzo-
p-dioxins and dibenzofurans (CDD/F) were found to be
formed naturally in soil of the Douglas fir forest; the formation
of three 2,3,7,8-substituted congeners, 2378-TeCDD, 12378-
PeCDD, and 123789-HxCDD, deserves special attention.
A formation mechanism has been proposed which starts
from naturally formed CP congeners and which probably
involves peroxidase mediation. Chlorination of CDD/F
congeners by the CPO-mediated reaction cannot be ruled
out, but seems to be less likely due to the absence of
several predicted congeners.
The above references all suggest the formation of CDD/ F
from anthropogenic chlorinated phenol congeners by natural
processes. However, CP congeners can also be formed
biogenically, and so real natural formation of CDD/ F comes
into perspective. Nine tick species are known to produce
2
2
6-DCP (16). Eisner et al. (17) reported the identification of
5-DCP in the froth of a grasshopper. However, they ascribed
this to the uptake of 2,4-dichlorophenoxyacetic acid (2,4-D)
or its degradation products in the plants which are eaten by
the grasshopper. Ando et al. (18) reported the de novo
synthesis of 24-DCP by a soil fungus of the group of
Penicillium. CP congeners can be formed from phenol (19)
and from compounds present in natural waters (20) in the
chloroperoxidase-mediated reaction. The omnipresence of
PO activity suggests that the natural formation of CDD/ F
from CP congeners can also occur in soil.
Introduction
The presence of low concentrations of (poly)chlorinated
compounds in soil is, no doubt, primarily due to anthro-
pogenic activity, i.e., environmental pollution. However, it
is known today that quite a number of chlorinated com-
pounds can also be formed naturally by either biochemical
In this paper, we report the natural formation of CP and
CDD/ F congeners in the humic soil layer of a Douglas fir
3
7
forest which was spiked in situ with a solution of Na Cl. The
isotope ratios of the CP and CDD/ F congeners were compared
to those in nonspiked soil samples from the Douglas fir and
a beech forest. The possible formation mechanisms are
discussed.
(
1-3) or geochemical (4-6) processes. Some of these
compounds, such as chloromethane (7) and polychlorinated
acetones (1), can have both natural and anthropogenic
sources. The combustion of natural organic material, such
as wood and fuels, which contains appreciable amounts of
inorganic chloride (8), has already been suggested as a source
Experimental Section
37Cl Isotope Spiking. Four sites (B1, C1, F2, F3) in a Douglas
fir forest in the rural area Speulderbos in The Netherlands
were selected because they showed high concentrations of
chloroform in soil air and natural chloroform formation was
*
Corresponding author address: European Commission, Joint
Research Centre, Environment Institute, I-21020 Ispra (Va), Italy;
phone: +39-0332785319; fax: +39-0332786351; e-mail: eddo.
hoekstra@jrc.it.
2
proven (21). At each site a soil surface of about 120 cm was
3
7
†
spiked with a solution of sodium chloride-37 (95% Na Cl;
TNO Environment, Energy and Process Development.
NM: Van Swinden Laboratory.
Free University.
3
7
2
‡
§
CIL CHLM-1225) at a concentration of 10 g of Na Cl per m .
An enclosure was put over the spiked area and the natural
1
0.1021/es9900104 CCC: $18.00

1999 Am erican Chem ical Society
VOL. 33, NO. 15, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2 5 4 3
Published on Web 08/01/1999

formation of chloroform was studied during 1 year (21). After
year of incubation the soil under the enclosure was sampled
in glass jars for CP and CDD/ F analysis.
Sam pling. Two nonspiked soil cores were taken from the
Douglas fir and a beech forest in Speulderbos. The soil top
layer was sampled with a stainless steel tube (l ) 30 cm; φint
Identification and Quantification. The CP and CDD/ F
congeners were identified on the basis of their retention time
and the isotope ratio of their molecular ion. 24- and 25-DCP
coeluted on the GC column used. Each CP congener was
1
1
3
quantified by means of the corresponding
6
C -labeled
internal standard in the congener group, i.e., 4-MCP, 24/
25-DCP, 245- and 246-TrCP, 2345-TeCP, and PCP, and by
)
8 cm) after removing the litter layer. The soil core was
pressed from the tube by a plunger with a PTFE head and
put on a stainless steel plate. The soil core was divided into
subsamples based on their color, and the subsamples were
put in glass jars (800 mL) which were closed by Bakelite caps
with PTFE inner layers. Litter (dead leaves) of the beech forest
was sampled in the same type of jars. The samples were
stored at -20 °C.
the response factor obtained using an external standard
containing the 12
C
- and
13
6
6
C -labeled congeners and 1234-
TCN as mentioned above. The recoveries of the internal
standards were calculated on the basis of the results obtained
with the external standard and the injection standard in the
sample extract. The recovery data were not used for
quantification and only give information about the perfor-
mance of the extraction and cleanup procedure. The
recoveries were in an acceptable range (see Table 2).
The non-2,3,7,8-substituted CDD/ F congeners were iden-
tified by comparing our chromatograms of a fly ash sample
containing all congeners with those published by Professor
Ballschmiter’s group (22-24). After the assignment of all the
congeners in the chromatogram of our fly ash sample, the
congeners present in the soil samples could be identified
based on their retention time. Not all DCDF and TrCDF
congeners could be identified from data in the quoted
references. Therefore, we could not properly quantify 34-
DCDF and 126-, 128-, 129-, 134-, 139-, 146-, 147-, 148-, 168-,
178-, 237-, 267-, 346-, and 347-TrCDF. It should be noted
that Bacher et al. (22, 23) mentioned 148-TrCDD in their list
which is, however, identical with 147-TrCDD, while 139-
TrCDD was missing. From correspondence with G. Schreiner
of Professor Ballschmiter’s group it appeared that the
systematic numbering of CDD congeners has to be corrected
as follows: D22 ) 139-TrCDD, D23 ) 146-TrCDD, and D24
) 147-TrCDD.
Extraction and Cleanup. The samples were dried for 24
h at 105 °C and, after grinding, sieved through a 1 mm sieve.
1
3
An internal standard (100 µL) which contained
6
C -labeled
CP congeners (4-MCP, 24/ 25-DCP, 245- and 246-TrCP, 2345-
1
3
TeCP, and PCP) and
C12-CDD/ F congeners (2378-TeCDD/
F, 12378-PeCDD, 23478-PeCDF, 123478-HxCDD, 123678-
HxCDF, 123789-HxCDD/ F, 1234678-HpCDD/ F, 1234789-
HpCDF, and OCDD/ F) in a concentration range of 40-500
-
1
37
µg L nonane was added to about 100 g of non-Na Cl-
spiked soil. The internal standard was not added to the
3
7
Na Cl-spiked soil samples since it would give interference
3
5
37
with the native congeners when analyzing Cl and Cl ions.
The samples were Soxhlet extracted with toluene for about
2
0 h; 90% of the extract was used for the CDD/ F analysis and
the rest for the analysis of the CP congeners.
The extract for the CDD/ F analysis was passed successively
through an activated carbon, a modified silica, and an
aluminum oxide column. The cleaned extract was evaporated
-
1
to dryness, and 50 µL of an injection standard of a 250 µg L
1
3
solution of
C12-1234-TeCDD in nonane was added. The
3
7
Each CDD/ F congener was quantified using the corre-
injection standard was not added to the extracts of Na Cl-
spiked soil samples since it would give interference as
mentioned above.
sponding 13
C
12-labeled internal standard in the congener
group, i.e., 2378-TeCDD/ F, 12378-PeCDD, 23478-PeCDF,
23478-HxCDD, 123678-HxCDF, 123789-HxCDD/ F, 1234678-
1
The extract for the analysis of the CP congeners was
reduced by evaporation to 1 mL of nonane which had been
added beforehand. The extract was diluted with 10 mL of
hexane, and the CP congeners were extracted into a 0.1 M
HpCDD/ F, 1234789-HpCDF, and OCDD/ F, and by the
response factor obtained with an external standard containing
the 12
C
12- and
13
C12-labeled congeners as mentioned above.
The MCDD/ F, DCDD/ F, and TrCDD/ F congeners were
2 3
K CO solution. After phase separation, 1 mL of acetic acid
1
3
quantified via
C
C
12-27/ 28-DCDD, because this was the only
anhydride was added to the aqueous solution, and the CP
acetates formed were extracted three times with 25 mL of
hexane. The hexane extracts were dried on anhydrous
1
3
12-labeled standard available for the lower chlorinated
CDD/ F. The recoveries of the internal standards were in an
acceptable range (see Table 4).
Na
2 4
SO and concentrated by evaporation to 1 mL. After the
-
1
addition of 100 µL of an injection standard of a 1 mg L
Results and Discussion
solution of 1,2,3,4-tetrachloronaphthalene (TCN) in hexane,
the extract was diluted to 2 mL.
GC-MS Analysis. Separation was performed by gas
chromatography (Carlo Erba GC8000). One microliter was
injected at 260 °C in the split-splitless mode. The CP
congeners were separated on a DB5MS column (l ) 60 m;
To verify whether natural production of CP and CDD/ F
congeners can occur, four sites in a rural Douglas fir forest
3
7
were spiked with a solution of Na Cl. Natural chlorinated
organic compounds are formed from a natural carbon source
and inorganic chloride. The detection of a compound
containing a 37Cl isotope percentage higher than the natural
percentage of 24% implies that this compound is formed
φ
int ) 0.25 mm; film thickness ) 0.1 µm) using a temperature
-
1
program from 50 °C (1 min) to 75 °C (25 °C min ; 0 min)
to 200 °C (3 °C min ; 0 min) to 300 °C (15 °C min ; 7 min).
The CDD/ F were separated on a SP2331 column (l ) 60 m;
-
1
-1
37
naturally in the Na Cl-spiked soil. At the selected sites, the
natural production of chloroform was observed for a period
of 1 year (21); after that study the soil under the enclosure
was sampled and analyzed by GC-PEI-MS and GC-NCI-
MS for the presence of CP and CDD/ F congeners with a
φ
int ) 0.25 mm; film thickness ) 0.2 µm) using a temperature
-
1
program from 130 °C (1 min) to 180 °C (25 °C min ; 0 min)
-
1
-1
to 250 °C (2.5 °C min ; 15 min) to 260 °C (2 °C min ; 35
min). The GC effluent was transported to the mass spec-
trometer through a direct coupling interface kept at 275 °C,
and the analytes were detected using a Fisons TRIO1000.
Detection in the positive electron ionization (PEI) mode was
done by detecting the mass ions either in the full scan (m/ z
3
7
deviating Cl isotope percentage.
CP Congeners. The 37Cl isotope percentage in the various
CP congeners was calculated from two isotope masses of the
molecular ion in the PEI-MS analysis and from m/ z 35 and
37 in the NCI-MS analysis. The reason for analyzing the
sample extracts by two ionization methods was that the
signals of 245-TrCP and 2346- and 2356-TeCP were disturbed
in PEI-MS. All analytes except the MCP congeners show
excellent detectability in NCI-MS at the concentrations
present in the samples, and the natural 37Cl isotope percent-
)
34-475) or in the selected ion mode. Detection in the
negative chemical ionization (NCI) mode using methane as
ionization gas was done by detecting the mass ions either in
the full scan mode (m/ z ) 34-475) or by detecting the mass
ions 35 and 37.
2
5 4 4
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 33, NO. 15, 1999

37
a-c
TABLE 1. Cl Isotope Percentage in the Molecular Mass Ion of CP Congeners
37
spiked with Na Cl
B1
nonspiked
CE1638-1645
(x ( sd)
C1
F2
F3
CH057 humus
0-5 cm
CH058 black
5-10 cm
CH059 humus
0-5 cm
CH060 humus
0-5 cm
CH061 humus
0-5 cm
analytes
2
4
2
2
2
2
2
2
MCP
MCP
23 ( 2
25
61
27
28
29
26
25
26
23
24
25
32
25
28
nd
26
24
26
24
24
24
30
26
26
22
28
24
26
25
24
24
32
26
24
26
25
24
26
23
24
25
27
25
25
30
25
26
25
26
23
24 ( 2
4/25 DCP
6 DCP
45 TrCP
46 TrCP
345 TeCP
346 TeCP
23 ( 0.4
22 ( 1
23 ( 0.7
24 ( 0.5
24 ( 0.4
23 ( 0.8
23 ( 0.9
23 ( 0.8
PCP
234 TCN
1
a
Abbreviations: x, average; sd, standard deviation; nd, not detected. Bold values: isotope percentage of 37Cl higher than isotope percentage
b
of 37Cl + 3 sd in sam ple extract of nonspiked soil. Data based on NCI-MS except for MCP (PEI-MS).
c
TABLE 2. Concentrations (µg kg- dw) of CP Congeners in Soil of a Beech and a Douglas Fir Foresta
1
beech
Douglas fir
CE1638
leave
litter
CE1639
humus
0-5 cm
CE1640
1st black
5-10 cm
CE1641
2nd black
10-15 cm
CE1642
humus
0-5 cm
CE1643
black
5-10 cm
CE1644
gray
10-15 cm
CE1645
yellow
15-20 cm
recovery (%)
(x ( sd)
b
analyte
2
3
4
2
2
2
3
3
2
2
2
2
2
3
2
2
2
MCP
MCP
MCP
0.93
<0.3
45
<0.09
7.1
0.55
0.07
25
<0.2
10
6.3
<0.2
<1
<0.1
<0.1
<0.08
0.33
9.4
<0.1
0.57
0.47
<0.09
2.1
0.21
0.04
6.5
<0.04
4.7
0.08
<0.01
1.4
<0.01
1.1
0.25
<0.04
18
<0.1
22
0.09
<0.04
2.2
<0.05
3.8
0.06
<0.02
0.70
<0.06
0.90
0.05
<0.02
0.55
<0.03
0.45
110 ( 20
90 ( 10
3 DCP
4/25 DCP
6 DCP
4 DCP
5 DCP
34 TrCP
35 TrCP
36 TrCP
45 TrCP
46 TrCP
45 TrCP
345 TeCP
346 TeCP
356 TeCP
12
1.5
0.37
2.8
1.1
0.32
0.14
<0.09
<0.4
<0.2
<0.3
<0.2
0.22
9.6
<0.3
0.32
0.44
0.10
2.3
<0.04
<0.1
0.11
<0.04
<0.03
0.20
9.5
<0.05
0.49
0.32
0.12
1.0
<0.01
<0.01
<0.01
0.02
<0.01
0.03
5.7
<0.02
0.04
<0.1
<0.1
<0.09
<0.1
<0.07
0.32
19
<0.1
0.54
0.55
0.09
4.9
<0.05
<0.06
<0.05
<0.06
<0.04
0.14
17
<0.06
0.17
<0.06
<0.07
<0.02
<0.03
<0.02
0.04
5.4
<0.03
0.03
<0.03
<0.04
<0.03
<0.04
<0.03
0.04
c
110 ( 10
100 ( 10
1.4
<0.04
<0.03
0.12
<0.03
0.33
120 ( 20
c
c
0.10
0.06
0.36
0.20
0.09
0.60
0.14
0.04
0.35
PCP
100 ( 10
a
, below detection lim it indicated. Detection lim its are seen to differ for different sam ple types. b x, average; sd, standard deviation. c Interferences
<
observed in the quantifying m ass. Quantification was done by correcting the response area of the quantifying m ass by m eans of the natural chlorine
isotope ratio, Cl: Cl ) 76:24, and the response area of the next highest m /z value in the m olecular m ass ion on the assum ption that this m /z
was undisturbed.
3
5
37
age of 24% could generally be determined with a lower
standard deviation by NCI than by PEI, i.e., about 1%.
The natural CP congeners, 4-MCP, 24/ 25-DCP, 26-DCP,
and 245-TrCP, might be present naturally in soil due to de
novo synthesis by, e.g., fungi (18) or the chloroperoxidase
(CPO)-catalyzed chlorination of aromatic structures (RH)
such as phenol (19) or humic material (20):
3
7
Formation of CP congeners from Na Cl was considered
3
7
to have been demonstrated if the Cl isotope percentage in
the sample extracts of the spiked soil was higher than the
percentage plus 3 times the standard deviation in the sample
extracts of the nonspiked soil. Table 1 shows the average 37Cl
isotope percentages and the standard deviations for the
nonspiked soils from the beech and Douglas forests (CE1638-
CPO
-
+
RH + Cl + H + H O2
8 RCl + 2H O
(1)
2
2
3
7
37
1
645) and the Cl isotope percentages for the Na Cl-spiked
CPO activity has been observed in several soil extracts, and
the chlorination activity in the organic layer was about 1000-
fold higher than in deeper soil layers (25). The CPOs are
most probably excreted by fungi similar to Caldariomyces
fumago, a deuteromycete, which is known to produce CPO.
Penicillium species that produce 24-DCP (18) also belong to
the deuteromycetes, but CPO production has not been
studied. Basidiomycetes do not show CPO activity; however,
in pure cultures they produce de novo large amounts of
chlorinated 4-methoxybenzaldehydes which constitute up
to 55% of the total amount of organohalogens (TOX) (26).
Basidiomycetes may therefore be responsible for the very
high concentrations of chlorinated 4-methoxybenzaldehydes
soils in the Douglas forest (CH057-061). The formation of
3
7
37
4
-MCP (32-61% Cl), 24/ 25-DCP (26-27% Cl), and 26-
37 37
DCP (26-28% Cl) from Na Cl was detected in three out of
five samples. Two humic layers distinctly show the formation
3
7
37
of 245-TrCP (29-30% Cl) from Na Cl because the use of
NCI-MS rather than PEI-MS eliminated the interference
problem and considerably reduced the standard deviation
of the Cl isotope percentage of this congener. The data for
the humic layer of site C1 suggests the natural formation of
3
7
3
7
2
46-TrCP (28% Cl). However, one should be careful here,
3
7
because an increase of the Cl isotope percentage was not
observed in PEI-MS although the standard deviations of both
methods were essentially the same in this case.
-
1
(0.04-20 mg of organic Cl kg dw) found in rotting wood,
VOL. 33, NO. 15, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2 5 4 5

37
a,b
TABLE 3. Cl Isotope Percentage in the Molecular Mass Ion of CDD/F Congeners Obtained from PEI-MS
37
spiked with Na Cl
nonspiked
CE1638-1645
(x ( sd)
B1
C1
CH059
F2
F3
analyte
CH057
CH058
CH060
CH061
1
1
1
2
1
1
1
1
1
1
MCDD
247/1248/1369 TeCDD
378 TeCDD
378 TeCDD
2346 PeCDD
24c
29
33
30
33
37
29
27
25
26
24
24
42
31
22
50
30
25
28
25
24
26
25
36
24
26
26
24
23
26
29
27
25
23
34
45
38
nd
34
28
nd
17
30
24
26
27
25
23
110
31
29
31
22
25
30
35
36
24
24
24
nd
22
26
25
27
23
25
29
33
26
nd
35
24
18
25
25
24
24
32
22
24
63
29
33
25
34
24
26
25
21
24
26
24
20
20
25
28
27
24
24
31
46
32
23
30
27
nd
27
35
24
23
34
23
25
54
31
30
26
24
23
28
25
22
23
26
28
nd
20
26
29
26
24
22
27
38
29
21
35
26
19
20
24
24
25
nd
22
27
46
34
24
26
25
25
29
27
24
24
26
25
nd
21
26
23
28
22
27
24 ( 1
24 ( 2
25 ( 2
24 ( 2
24 ( 0.8
25 ( 2
25 ( 0.8
25 ( 2
24 ( 1
24 ( 2
2378 PeCDD
23478 HxCDD
23678 HxCDD
23789 HxCDD
234678 HpCDD
OCDD
MCDF
c
1
2
3
4
1
1
1
1
2
1
1
1
2
1
1
1
2
1
1
1
1
26 ( 2
c
MCDF
MCDF
MCDF
7 DCDF
24/167 TrCDF
234/1678 TeCDF
347 TeCDF
24 ( 1
c
24 ( 1
c
23 ( 3
25 ( 0.6
25 ( 0.8
25 ( 1
26 ( 2
378 TeCDF
25 ( 2
2346 PeCDF
2348/12378 PeCDF
2489 PeCDF
3478 PeCDF
23478/123479 HxCDF
23678 HxCDF
23789 HxCDF
34678 HxCDF
234678 HpCDF
234679 HpCDF
234689 HpCDF
234789 HpCDF
24 ( 1
24 ( 1
25 ( 1
24 ( 0.2
24 ( 2
24 ( 2
25 ( 2
24 ( 0.8
24 ( 1
25 ( 0.7
24 ( 0.4
24 ( 0.4
24 ( 0.8
OCDF
a
2
,3,7,8-Substituted congeners in italics; x, average; sd, standard deviation; nd, one of the two isotope m asses in the m olecular ion not detected.
37 c
b
Bold values: isotope percentage of Cl exceeds that in the nonspiked sam ples + 3 sd. Values obtained using another MS instrum ent.
moldy leaf litter, and soil from a Lepista nuda fairy ring (27,
dw, respectively, and were high compared to those for which
natural formation was not observed. The concentrations of
245-TrCP observed in soil were invariably low.
2
8) compared to the TOX concentration (4-920 mg of Cl
-
1
kg dw (29)). How basidiomycetes synthesize chlorinated
aromatics is not clear yet.
From our results it cannot be concluded that the presence
of 4-MCP, 24/ 25-DCP, 26-DCP, and 245-TrCP are only due
to natural formation in soil. Emissions from production and
use of CP congeners or herbicides synthesized from CPs will
also contribute to the presence of CP congeners in soil of
rural areas by evaporation, atmospheric transportation, and
deposition. Industrial chlorination of phenol to obtain
commercial wood preserving agents (31) mainly proceeds
via 4-MCP, 24-DCP, 246-TrCP, and 2346-TeCP to PCP.
Therefore, the presence of high concentrations of CP
congeners in rural soils may also be due to anthropogenic
sources, and the relative contributions of the different inputs
remain to be elucidated.
Since the two nonspiked soil cores contained an internal
standard, the concentrations of the CP congeners in these
samples could be calculated to get an impression about the
order of magnitude (Table 2). Since the internal standard
was added after the drying procedure (24 h at 105 °C), the
recoveries shown in Table 2 do not take into account the
whole treatment procedure, and therefore the results may
have been influenced by the experimental setup. We know
of only one other study which describes background levels
of CP congeners in rural soil. Edelman (30) screened 40
different sites spread over The Netherlands by taking 10 cm
cores and drying the soil samples for 48 h at 35 °C. He found
2
<
4-DCP, 246-TrCP, and PCP concentrations of <0.5-18,
CDD and CDF Congeners. Since analysis of the samples
from the Na Cl-spiked Douglas soils revealed that several
-1
0.2-2.7, and <0.1-4.4 µg kg dw, respectively. Since these
37
results are of the same order of magnitude or lower than
ours (i.e., the average of our 0-5 cm and 5-10 cm samples
in Table 2), the influence of the sample drying can be
considered to have been small.
MCP to TrCP congeners are formed naturally, the nonspiked
CE1638-1645 and the spiked CH057-061 sample series were
also analyzed by GC-PEI-MS for the presence of CDD/ F
37
congeners. The Cl isotope percentage was calculated from
Five congeners were not found in any sample, and three
more showed up in one or two samples only. For all 11
remaining analytes, the concentrations are seen to decrease
with increasing soil depth, and there are no real differences
in the concentration profiles in the soils of the beech and the
Douglas forest. The concentrations of the natural CP
congeners, 4-MCP, 24/ 25-DCP, and 26-DCP, in the humic
the two isotope masses of the molecular ion with the highest
intensity, i.e., under the natural nonspiked conditions of
samples CE1638-1645. Natural formation was demonstrated
in the same way as for CP congeners. Table 3 shows the Cl
isotope percentage of CDD/ F congeners with a 37Cl isotope
percentage above the natural level and the data for all 2,3,7,8-
substituted congeners. From among the latter group of
analytes, 12378-PeCDD showed a significantly increased
3
7
-
1
layers were found to be 18-25, 10-22, and 2.8-6.3 µg kg
2
5 4 6
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 33, NO. 15, 1999

TABLE 4. Concentrations of Selected CDD/F Congeners and TEQ Values in Soil of a Beech and a Douglas Fir Forest (ng kg- Dw)a
1
beech
CE1639
douglas fir
recovery (%)
(x ( sd)
b
analyte
CE1638
CE1640
CE1641
CE1642
CE1643
CE1644
CE1645
2
1
1
1
2
1
1
1
1
1
1
1
1
1
2
2
2
1
2
1
1
2
1
1
1
2
1
1
1
2
1
1
1
1
1
3/27/28 DCDD
39/237 TrCDD
247/1248/1369 TeCDD
378 TeCDD
2.2
0.58
3.4
2.5
0.43
<2.2
1.9
<6
3.0
1.3
9.5
6.6
1.1
2.6
3.8
3.9
8.3
8.2
1.8
0.71
4.4
2.7
0.37
1.8
2.1
2.1
3.8
4.2
37
<0.1
<0.08
0.37
0.18
0.04
0.34
0.22
0.31
0.31
0.32
4.2
5.8
0.65
3.3
2.1
0.46
1.2
2.0
2.1
4.1
5.3
44
3.5
0.22
1.5
0.61
0.23
0.51
0.81
1.0
1.5
1.8
18
18
70
0.31
0.53
1.8
<0.05
0.75
2.2
0.70
2.6
2.7
1.1
3.6
0.44
3.1
1.3
<0.06
0.36
0.13
0.05
0.20
0.15
0.19
0.17
0.38
2.4
0.35
<0.08
0.10
0.05
0.05
<0.2
0.07
<0.2
<0.2
<0.3
0.80
0.76
5.4
0.06
0.08
0.57
0.39
<0.08
0.19
<0.1
0.59
0.20
<0.1
0.15
<0.1
0.16
0.21
0.16
0.30
0.15
1.4
120 ( 20
378 TeCDD
90 ( 10
2346 PeCDD
2378 PeCDD
23478 HxCDD
23678 HxCDD
23789 HxCDD
234678 HpCDD
234679 HpCDD
2346789 OCDD
7 DCDF
3 DCDF
7 DCDF
8 DCDF
24/167 TrCDF
34/238 TrCDF
234/1678 TeCDF
347 TeCDF
378 TeCDF
2346 PeCDF
90 ( 20
80 ( 20
<6
4.1
38
39
70 ( 20
60 ( 20
95
86
360
0.94
36
4.4
33
47
2.2
12
140
0.50
0.51
2.5
3.7
1.2
4.2
6.7
2.6
4.1
2.2
5.2
<2
4.3
4.1
3.7
<7
<8
21
170
1.6
1.9
4.8
9.2
1.8
7.6
7.8
6.8
5.3
4.8
14
1.2
8.2
13
8.8
6.5
6.9
88
190
0.53
0.57
3.1
0.20
<0.08
5.8
2.3
4.8
7.4
2.1
6.3
0.77
6.2
6.9
5.8
1.9
5.0
33
60 ( 30
0.08
0.21
0.53
<0.03
<0.03
0.95
3.4
0.83
0.67
0.34
1.0
0.14
0.91
1.4
0.98
0.70
0.81
8.5
0.10
0.25
0.69
0.15
0.18
0.55
<0.1
0.54
0.51
0.27
0.74
0.11
0.61
0.93
0.68
0.53
0.55
6.1
0.94
4.0
6.7
1.9
6.9
5.3
6.4
11
5.0
13
1.5
10
13
11
4.2
9.8
58
80 ( 10
80 ( 20
2348/12378 PeCDF
2489 PeCDF
3478 PeCDF
23478/123479 HxCDF
23678 HxCDF
23789 HxCDF
34678 HxCDF
234678 HpCDF
234679 HpCDF
234689 HpCDF
234789 HpCDF
2346789 OCDF
4.6
3.3
1.8
2.7
29
2.5
5.3
2.6
80 ( 20
90 ( 30
70 ( 20
3.9
3.2
<7
<60
6.1
7.3
7.4
9.6
<7
42
18
21
6.4
20
5.1
120
13
0.87
1.2
0.91
16
1.4
1.6
3.6
7.2
3.0
35
9.8
11
0.54
0.74
0.74
12
0.98
1.1
0.09
0.15
0.38
2.6
0.31
0.34
80 ( 30
80 ( 40
44
4.9
5.4
I-TEQ
BGA-TEQ
16
a
Naturally form ed congeners and 2,3,7,8-substituted CDD/F congeners (italics) including DCDD/F and TrCDD/F that are substituted at one or
m ore of the 2,3,7,8 positions were selected. <, below detection lim it indicated. Detection lim its are seen to differ for different sam ple types.b x,
average; sd, standard deviation.
37Cl isotope percentage in two out of five samples and 2378-
TeCDD and 123789-HxCDD in one sample. Of the non-
dw in the humic layer of the forests. The concentrations of
all CDD/ F congeners found in both forest soils decrease with
increasing depth, and the concentrations in the litter sample
are lower than in the humic layer. Other than the natural CP
congeners, the natural CDD/ F congeners are not present in
the highest concentrations but are always below 10 ng kg
dw in the humic layer.
Another mechanism that could explain the formation of
CDD/ F congeners enriched with 37Cl is the peroxidase (PO)-
mediated coupling of CP congeners as studied by O¨ berg et
al. (12) who found the following main reactions:
2
1
,3,7,8-substituted congeners, 4-MCDF, 17-DCDF, and 1247/
248/ 1369-TeCDD were found to be formed naturally in all
3
7
the Na Cl-spiked soils and 1-MCDD, 12346-PeCDD,k and
234689-HpCDF in four out of five. Nine non-2,3,7,8-
-
1
1
substituted congeners were formed in less than four spiked
soils.
The natural formation of CDD/ F congeners in the soil of
the Douglas forest may be caused by the chloroperoxidase-
mediated chlorination of CDD/ F congeners. Since the
chloroperoxidase-mediated reaction chlorinates organic
material randomly, the CDD/ F congeners are expected to be
chlorinated according to their relative concentration ratio,
and thus the natural CDD/ F congeners are expected to be
mainly formed from the congeners present at the highest
concentrations in soil. This is the constraint of the CPO
mechanism because most congeners with a relative high
concentration in soil appear not to lead to one of the natural
CDD/ F congeners found. Table 4 shows the concentrations
of the naturally formed congeners and the 2,3,7,8-substituted
CDD/ F congeners including DCDD/ F and TrCDD/ F that are
substituted at one or more of the 2,3,7,8 positions, in the soil
core samples from the Douglas fir and beech forests. All
MCDD/ F congeners were found to be below their detection
limits. Next to the congeners reported in Table 4, 35 non-
PO
2
CP + 2H O2 8 CDD + 4H O
(2)
(3)
2
2
PO
2
CP + H O2
8 CDF + 3H O
2
2
The formed CDD/ F congeners were interpreted as being
direct coupling products or indirect products from chlorine
rearrangements, abstractions, and additions. The authors
argued that CDD/ F congeners are formed via chlorophenoxy
radicals. However, they did not mention or discuss the
reaction pathways which they assumed. Therefore, we studied
two reaction mechanisms which are proposed to occur in
gas-phase pyrolysis (32) and fly-ash-catalyzed oxidation (33,
34).
2
,3,7,8-substituted TeCDD/ F to HxCDD/ F congeners were
In the gas-phase pyrolysis of CP congeners, CDF congeners
are assumed to be formed via o-C-o-C coupling of two
-
1
found to be present in concentrations higher than 10 ng kg
VOL. 33, NO. 15, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2 5 4 7

chlorophenoxy radicals to a 2,2′-dihydroxybiphenyl, whereas
CDD congeners are supposed to be formed via the attack of
a chlorophenoxy radical on a CP congener at the ortho
position, forming a 2-phenoxyphenol radical (32). The
presence of a chlorine atom at the ortho position of the two
reacting CP congeners appears to be essential for the
formation of CDD congeners. Since, in our study, all the
naturally formed CP congeners and all CP congeners with
high concentrations in soil have a chlorine atom in the ortho
position (except 4-MCP), we expected that combinations of
these congeners will most probably cause the formation of
CDD/ F congeners. However, the mechanism proposed by
Born et al. (32) for the formation of CDD/ F congeners in
pyrolysis processes does not predict our results in soil except
for the formation of 1378-TeCDD, 12346-PeCDD, and
1
234689-HpCDF. We, therefore, tentatively conclude that a
gas-phase pyrolysis-like mechanism does not take place in
soil.
The formation of CDD congeners from CP congeners on
fly ash is thought to follow the same mechanism as in gas-
phase pyrolysis and, in parallel, a Smiles rearrangement of
the 2-phenoxyphenol anion is suggested (33). The presence
of a chlorine atom in the ortho position of the reacting CP
congeners is not necessary for the formation of CDD
congeners since meta- and para-substituted CP congeners
are easily isomerized to ortho-substituted analogues on fly
ash (34). No formation of CDF congeners by the fly-ash-
mediated reaction is reported, and CP congeners are not
expected to be CDF precursors in combustion (34). Therefore,
the fly-ash-mediated mechanism cannot explain our results
any better than the gas-phase pyrolysis of CP congeners
referred to above.
The PO-mediated reaction and the fly-ash-mediated
mechanism have in common that ortho substitution is not
important while they differ with regard to the products
formed. The CDD and CDF congeners formed in the PO-
mediated reaction appear to have the same number of
chlorine atoms as their reactants (12). Taking these observa-
tions into account, we adapted the fly-ash-mediated mech-
anism. Figure 1 shows our proposed mechanism that
proceeds via anions; however, a radical mechanism would
lead to the same products (34). For the formation of CDD
congeners it was assumed that hydrogen can also be a leaving
group (35), and for the formation of CDF congeners we
suggest a completely new mechanism that closely resembles
the CDD formation mechanism. The hypochlorous acid that
can be formed during CDF formation may be responsible for
the chlorination reaction that O¨ berg et al. (12) observed in
the PO-mediated reaction of CP congeners.
FIGURE 1. Possible formation mechanism of CDD/F congeners
mediated by peroxidases.
TABLE 5. Natural Formation of CDD/F Congeners Explained by
PO-Mediated Formation from Combinations of CP Congenersa
naturally
formed CP
congeners 4-MCP 24-DCP 25-DCP 26-DCP 246-TrCP
CP congeners with highest concentrations in soil
PCP
4
2
-MCP
4-DCP
1369-D
17-F
25-DCP
26-DCP
1369-D
17-F 1369-D
12346-D
12346-F
2
45-TrCP 1247-D 1247-D 1247-D
1378-D 1234679-F
1234689-F
1
248-D 1248-D 1248-D
378-D 1347-F
Table 5 gives an overview of how the various CDD/ F
1
3
7
congeners that were formed in our Na Cl-spiked soil
experiments can be explained by specific combinations of
naturally formed CP congeners and congeners with a high
concentration in soil. 17-DCDF and 1247-/ 1248-/ 1369-
TeCDD seem to be typical products in soil since they can be
formed from naturally formed CP congeners by a natural
reaction. The formation of 12346-PeCDD, 12346-PeCDF, and
a
D, CDD; F, CDF.
on the natural formation of CDD/ F congeners, even though
it must be admitted that 245-TrCP plays a larger role than
would be expected on the basis of its rather low concentration
in soil. In addition, our mechanism predicts the formation
of more congeners than we actually found (see, e.g., Figure
1 where the formation of 137- and 237-TrCDD and 1248-
TeCDF is predicted (note that 237-TrCDF could not be
identified; cf. Experimental Section)). Although one can, of
course, argue that some congeners may well have been
formed in too low amounts to permit their detection by
the current analytical procedure, this is too unsubstantiated
a way out. It is obvious that further work will have to be
carried out to validate and/ or modify the proposed mech-
anism. Nevertheless, we feel that the present results, and
their tentative interpretation, do contribute to the unravelling
of the problem of the natural formation of CP and CDD/ F
congeners.
1
234679- and 1234689-HpCDF in soil cannot be called purely
natural because one of the reactants, PCP, originates from
anthropogenic sources. 1369- and 1378-TeCDD can be
formed from either purely natural CP congeners or a natural
congener and anthropogenic 246-TrCP. The formation of
1
-MCDD and 4-MCDF cannot be explained on the basis of
the above-mentioned CP congeners only: if our mechanism
holds true, the involvement of phenol is suggested. Regarding
the congeners that were less frequently found in the
3
7
Na Cl-spiked soil studies, their formation should probably
be explained also by chlorine atom rearrangements.
In summary, the mechanism proposed above provides a
rather satisfactory explanation of our experimental results
2
5 4 8
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 33, NO. 15, 1999

Finally, now that it has been demonstrated that natural
formation of CP and CDD/ F congeners in soil is possible,
subsequent research should be directed at evaluating the
significance of this natural formation compared to other
sources such as atmospheric deposition.
(9) Olie, K.; Vermeulen, P. L.; Hutzinger, O. Chemosphere 1977, 6,
55.
10) Sheffield, A. Chemosphere 1985, 14, 811.
11) Harrad, S.; Jones, K. C. Chem. Br. 1993, 29, 198.
12) O¨ berg, L. G.; Glas, B.; Swanson, S. E.; Rappe, C.; Paul, K. G. Arch.
Environ. Contam. Toxicol. 1990, 19, 930.
4
(
(
(
(
(
(
13) Wagner, H. C.; Schramm, K. W.; Hutzinger, O. Z. Umweltchem.
O¨ kotox. 1990, 2, 63.
Abbreviations
14) O¨ berg, L. G.; Andersson, R.; Rappe, C. Organohalogen Compd
CDD
CDF
CP
D
chlorinated dibenzo-p-dioxins
1
992, 9, 351.
chlorinated dibenzofurans
chlorinated phenols
di
15) O¨ berg, L. G.; Wågman, N.; Andersson, R.; Rappe, C. Organo-
halogen Compd 1993, 11, 297.
(16) Berger, R. S. J. Med. Ent. 1983, 20, 103.
(
(
(
17) Eisner, T.; Hendry, L. B.; Peakall, D. B.; Meinwald, J. Science
971, 172, 277.
18) Ando, K.; Kato, A.; Susuki, S. Biochem. Biophys. Res. Commun.
970, 39, 1104.
Hp
Hx
M
O
Pe
hepta
hexa
mono
octa
penta
tetra
tri
1
1
19) Wannstedt, C.; Rotella, D.; Siuda, J. F. Bull. Environ. Contam.
Toxicol. 1990, 44, 282.
Te
Tr
(20) Hodin, F.; Bor e´ n, H.; Grimvall, A.; Karlsson, S. Water Sci. Technol.
991, 24, 403.
1
(
(
(
(
21) Hoekstra, E. J.; De Leer, E. W. B.; Brinkman, U. A. Th. Environ.
Sci. Technol. 1998, 32, 3724.
22) Bacher, R.; Riehle, U.; Swerev, M.; Ballschmiter, K. Chemosphere
1991, 23, 1151.
Acknowledgments
Part of this work has been carried out within the frame-
work of the EC project “Natural halogenation of organic
macromolecules in the terrestrial environment” (STEP-CT91-
23) Bacher, R.; Swerev, M.; Ballschmiter, K. Environ. Sci. Technol.
1
992, 26, 1649.
24) Schimmel, H.; Riehle, U.; Reuter, U.; Ballschmiter, K. Chemo-
sphere 1992, 24, 413.
2
6). We thank Staatsbosbeheer (Driebergen, The Nether-
(25) Laturnus, F.; Mehrtens, G.; Grøn, C. Chemosphere 1995, 31,
lands) for the use of their nature reserves for sampling. We
greatly appreciated the help of G. Schreiner of Professor
Ballschmiter’s group to correct the systematic numbering
and to identify some DCDF and TrCDF congeners whose
relative retention times were not published yet.
3709.
(
(
26) Verhagen, F. J. M.; Swarts, H. J.; Wijnberg, J. B. P. M.; Field, J.
A. Chemosphere 1998, 37, 2091.
27) De Jong, E.; Field, J. A.; Spinnler, H. E.; Wijnberg, J. B. P. A.; De
Bont, J. A. M. Appl. Environ. Microbiol. 1994, 60, 264.
28) Hjelm, O.; Bor e´ n, H.; O¨ berg, G. Chemosphere 1996, 32, 1719.
29) Asplund, G. In Naturally-Produced Organohalogens; Grimvall,
A., De Leer, E. W. B., Eds.; Kluwer Academic Publishers:
Dordrecht, 1995; pp 35-48.
(
(
Literature Cited
(
1) Neidleman, S. L.; Geigert, J. Biohalogenation: principles, basic
roles and applications; Ellis Horwood: Chichester, 1986.
2) Gribble, G. W. Environ. Sci. Technol. 1994, 28, 310A.
3) Field, J. A.; Verhagen, F. J. M.; De Jong, E. Trends Biotechnol.
(
(
30) Edelman, T. Background concentrations of compounds in Dutch
soils (in Dutch), Ministry of VROM, Staatsuitgeverij: The Hague,
(
(
1
984.
31) Vartiainen, T.; Lampi, P.; Tolonen, K.; Tuomisto, J. Chemosphere
995, 30, 1439.
1
995, 13, 451.
4) Isidorov, V. A.; Zenkevich, I. G.; Ioffe, B. V. J. Atmos. Chem.
990, 10, 329.
5) Isidorov, V. A.; Povarov, V. G.; Prilepsky, E. B. J. Ecol. Chem.
993, N1, 19.
6) Isidorov, V. A.; Prilepsky, E. B.; Povarov, V. G. J. Ecol. Chem.
993, N2-3, 201.
1
(
(
(
(
(
(
32) Born, J. G. P.; Louw, R.; Mulder, P. Chemosphere 1989, 19, 401.
33) Born, J. G. P. Ph.D. Dissertation, Leiden University at Leiden,
1
1
991; pp 176-178.
1
(
(
34) Tuppurainen, K.; Halonen, I.; Ruokoj a¨ rvi, P.; Tarhanen, J.;
Ruuskanen, J. Chemosphere 1998, 36, 1493.
35) Butte, W.; Juhl, U.; Schwarting, W.; Witte, I. Chemosphere 1988,
1
7) Harper, D. B. In Naturally-Produced Organohalogens; Grimvall,
A., De Leer, E. W. B., Eds.; Kluwer Academic Publishers:
Dordrecht, 1995; pp 21-33.
8) Bumb, R. R.; Crummett, W. B.; Cutie, S. S.; Gledhill, J. R.; Hummel,
R. H.; Kagel, R. O.; Lamparski, L. L.; Luoma, E. V.; Miller, L.;
Nestrick, T. J.; Shadoff, L. A.; Stehl, R. H.; Woods, J. S. Science
1
7, 1189.
(
Received for review January 4, 1999. Revised manuscript
received April 15, 1999. Accepted April 19, 1999.
1
985, 210, 385.
ES9900104
VOL. 33, NO. 15, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2 5 4 9