Subscriber access provided by MIDWESTERN UNIVERSITY
Characterization of Natural and Affected Environments
Complex Mixtures of Brominated/Chlorinated Diphenyl Ethers and Dibenzofurans in Soils from the Agbogbloshie E-waste Site (Ghana): Occurrence, Formation and Exposure Implications Nguyen Minh Tue, Takafumi Matsushita, Akitoshi Goto, Takaaki Itai, Kwadwo Ansong Asante, Samuel Obiri, Saada Mohammed, Shinsuke Tanabe, and Tatsuya Kunisue Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.8b06929 • Publication Date (Web): 22 Feb 2019 Downloaded from http://pubs.acs.org on February 23, 2019
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 42
Environmental Science & Technology
1
Complex Mixtures of Brominated/Chlorinated
2
Diphenyl Ethers and Dibenzofurans in Soils from
3
the Agbogbloshie E-waste Site (Ghana): Occurrence,
4
Formation and Exposure Implications
5
Nguyen Minh Tue,† Takafumi Matsushita,† Akitoshi Goto,† Takaaki Itai,† Kwadwo
6
Ansong Asante,†,‡ Samuel Obiri,‡ Saada Mohammed,‡ Shinsuke Tanabe,† Tatsuya
7
Kunisue*,†
8
†Center for Marine Environmental Studies (CMES), Ehime University, 2-5 Bunkyo-cho,
9
Matsuyama 790-8577, Japan
10
‡CSIR Water Research Institute, P.O. Box AH 38, Achimota, Accra, Ghana
11
ACS Paragon Plus Environment
1
Environmental Science & Technology
Page 2 of 42
12
ABSTRACT: The distribution and toxic equivalents (TEQs) of brominated, chlorinated
13
dibenzo-p-dioxins/dibenzofurans (PBDD/Fs and PCDD/Fs) in soils from Agbogbloshie
14
e-waste site (Ghana) were investigated. The composition of brominated/chlorinated
15
dibenzofurans (PXDFs) and diphenyl ethers (PBDEs, PCDEs and PXDEs) were
16
examined using two-dimensional gas chromatography–time-of-flight mass spectrometry
17
to elucidate possible formation pathways of dioxins from e-waste recycling. The highest
18
concentrations of PCDD/Fs and PBDD/Fs were found respectively in the open burning
19
(1.3–380 ng/g dry weight) and dismantling areas (11–1000 ng/g dry weight), and were
20
comparable to the highest reported for informal e-waste sites. PXDFs and PXDEs were
21
detected at up to the range of hundreds of nanogram per gram. The homologue profiles
22
suggest that PXDFs were formed mainly from PBDFs through successive Br-to-Cl
23
exchange. However, monobromo-PCDFs were also derived from de novo-generated
24
PCDFs in open burning areas. PBDFs contributed similar or higher TEQs (7.9–
25
5400 pg/g dry weight) compared with PCDD/Fs (6.8–5200 pg/g dry weight), whereas
26
PXDFs were also substantial TEQ contributors in open burning areas. The high TEQs of
27
PBDFs in the dismantling area (120–5200 pg/g dry weight) indicate the need to consider
ACS Paragon Plus Environment
2
Page 3 of 42
Environmental Science & Technology
28
brominated dioxins besides chlorinated dioxins in future studies on health implications
29
for e-waste workers and local residents.
30
KEYWORDS: dioxins, e-waste, Ghana, PBDD/Fs, PXDD/Fs, PXDEs
31
ACS Paragon Plus Environment
3
Environmental Science & Technology
32
33
Page 4 of 42
INTRODUCTION Waste electrical and electronic equipment (e-waste) has become increasingly difficult
34
to manage because of the steady growth, with an estimated annual global volume of
35
44.7 million metric tons in 2016.1 E-waste contains not only valuable metals for
36
reclamation but also a large number of toxic substances such as heavy metals,
37
brominated flame retardants (BFRs) and various plastic additives.2 A large volume of
38
these hazardous waste materials have been treated informally in developing countries,
39
especially in Asia3,4 and more recently in Africa4,5 using crude thermal processes such
40
as circuit board heating and open burning of wires. These processes result in serious
41
environmental pollution caused by not only the emission of contaminants contained in e-
42
waste2 but also the unintentional formation of complex mixtures of thermally produced
43
toxic chemicals including chlorinated and/or brominated dioxin-like compounds (DLCs),
44
polycyclic hydrocarbons and related compounds.6–9 There is an increasing number of
45
reports on multiple adverse health effects in populations involved with informal e-waste
46
recycling.10
ACS Paragon Plus Environment
4
Page 5 of 42
Environmental Science & Technology
47
Although DLCs are known for their potency to cause many toxic effects,11,12 it is still
48
difficult to assess the environmental and health impacts of all DLCs from e-waste due to
49
their compositional complexity. Informal e-waste recycling sites (EWRSs) have been
50
widely known to be contamination hotspots of polychlorinated dibenzo-p-dioxins and
51
dibenzofurans (PCDD/Fs), with polyvinyl chloride (PVC)-coated wire burning as the
52
major emission source.6,13 The number of studies on brominated dibenzo-p-dioxins and
53
dibenzofurans (PBDD/Fs) in EWRSs is more limited, however, higher environmental
54
contamination levels of PBDFs compared with PCDD/Fs have been reported.6 The
55
abundance of PBDFs can be explained by their occurrence at up to hundreds of
56
microgram per gram in e-waste such as circuit boards and TV cases,14,15 and by their
57
possible formation through thermolytic/photolytic degradation of brominated flame
58
retardants in e-waste plastics, especially polybrominated diphenyl ethers (PBDEs)14,16–
59
18.
60
additives can also generate mixed brominated/chlorinated dibenzo-p-
61
dioxins/dibenzofurans (PXDD/Fs)16,19. However, because of the difficulties in analyzing
62
PXDD/Fs with 4700 possible congeners, only a few studies have attempted to analyze
Combustion of e-waste containing both chlorinated polymers and brominated
ACS Paragon Plus Environment
5
Environmental Science & Technology
63
these mixed halogenated DLCs in a limited number of environmental samples from
64
EWRSs.20,21 These studies found higher abundance of PXDFs as compared with
65
PXDDs but reported different PXDF homologue profiles, suggesting that the
66
composition of PXDFs varies with the waste materials and thermal conditions. In the
67
light of possible inclusion of PBDD/Fs and PXDD/Fs in the World Health Organization
68
Toxicity Equivalency Factor (WHO-TEF) concept,23 their formation, environmental fate
69
and potential human exposure risk in informal EWRSs need to be investigated.
70
Page 6 of 42
Another related group of mixed halogenated compounds that can be formed during
71
informal e-waste recycling is the mixed brominated/chlorinated analogues of PBDEs
72
(PXDEs). Despite possible formation of PXDEs during combustion processes,24,25
73
information on their occurrence in EWRSs is still severely lacking. In our knowledge,
74
there is only one study22 investigating a very small subset of PXDE homologues (nona-
75
brominated mono-chlorinated). Compounds tentatively identified as PXDEs were also
76
detected in a soil sample from an e-waste burning area in our previous study on PXDF
77
profiling using two-dimensional gas chromatography–time-of-flight mass spectrometry
78
(GC×GC–ToFMS),21 but could not be quantified reliably due to extensive debromination
ACS Paragon Plus Environment
6
Page 7 of 42
Environmental Science & Technology
79
on the second GC column. Considering the structural similarity with PBDEs which are
80
major toxic contaminants in e-waste,2 PXDEs need to be investigated in order to assess
81
their contribution to the environmental and human impacts of PBDEs. The possible role
82
of PXDEs as precursors in the formation of PXDFs also needs to be explored to better
83
estimate the emission of PXDFs in EWRSs.
84
The Agbogbloshie scrap yard is the major hub of informal waste recycling in
85
Ghana.26,27 At the time of our first survey in 2010,21 the site was primarily a burning
86
ground for the recovery of copper from e-waste. Surface soils were sampled at ten
87
spots, and PCDD/Fs and PBDD/Fs were detected at very high concentrations in the
88
open burning areas.21 PXDFs were analyzed for one open burning soil sample using
89
GC×GC–ToFMS and were found to contribute approximately 16% to the total dioxin-like
90
toxic equivalents (TEQs). During our second survey in 2013, we observed at
91
Agbogbloshie the stockpiling of not only e-waste but also end-of-life vehicle (ELV)-
92
related waste and common solid waste, the presence of rudimentary dismantling
93
workshops, and an increase in the number of open e-waste burning spots as well as the
94
number of workers handling waste materials (Figure S1). The present study
ACS Paragon Plus Environment
7
Environmental Science & Technology
95
investigated a larger set of surface soil samples collected during the second survey to
96
elucidate the contamination status of PCDD/Fs, PBDD/Fs and PBDEs as well as the
97
composition of PXDFs and PXDEs in different areas, in order to obtain a more
98
comprehensive view on the formation, toxic contribution and human exposure
99
implications of various DLCs generated during informal e-waste recycling/disposal.
100
101
Page 8 of 42
MATERIALS AND METHODS Sample collection. In August 2013, surface soil samples (0–2 cm depth) were
102
collected at the Agbogbloshie e-waste site using a stainless steel auger. Each sample
103
consisted of five sub-samples covering a square area of approximately 1 m2. The
104
samples were kept in a freezer in the Environmental Chemistry Laboratory of the Water
105
Research Institute of the Council for Scientific and Industrial Research in Accra, then
106
transported to Japan under frozen condition and stored at –25 °C until analysis. The soil
107
samples were imported to Japan under permission (no. 25-300) from Ministry of
108
Agriculture, Forestry and Fisheries, Japan based on Plant Protection Act. Forty one
109
samples were selected for analysis in this study: 14 samples from open e-waste burning
ACS Paragon Plus Environment
8
Page 9 of 42
Environmental Science & Technology
110
areas (areas #1 to #4), 14 from the e-waste/ELV dismantling area (consisting of sub-
111
areas #6 and #7), and 13 from other areas (Figure S2).
112
Sample pretreatment and extraction. Soil samples were oven-dried at 35 °C in
113
separate batches according to the sampling area, and then each sample was
114
homogenized and sieved through a 250-µm mesh. Extraction was carried out based on
115
a previous method.28 Briefly, approximately 5 g of each sample was extracted using a
116
high speed solvent extractor (SE-100, Mitsubishi Chemical Analytech) with a
117
hexane/acetone mixture (1:1 v/v, 35 °C, 4 ml/min, 60 min) and then with toluene (80 °C,
118
4 ml/min, 30 min). The two fractions were solvent-exchanged into hexane and
119
combined.
120
Quantitative analysis of PCDD/Fs, PBDD/Fs and PBDEs. A portion of the extract (up
121
to 1-g equivalent) was spiked with 13C12-PCDD/Fs (2,3,7,8-, 1,2,3,7,8-, 1,2,3,4,7,8-,
122
1,2,3,6,7,8-, 1,2,3,7,8,9-, 1,2,3,4,6,7,8-, octa-CDDs and 2,3,7,8-, 1,2,3,7,8-, 1,2,3,6,7,8-,
123
1,2,3,7,8,9-, 1,2,3,4,6,7,8-, octa-CDFs; Wellington Laboratories), 13C12-PBDD/Fs
124
(2,3,7,8-, 1,2,3,7,8-, 1,2,3,4,7,8-, 1,2,3,4,6,7,8-, octa-BDDs and 2,3,7,8-, 2,3,4,7,8-,
ACS Paragon Plus Environment
9
Environmental Science & Technology
Page 10 of 42
125
1,2,3,4,7,8-, 1,2,3,4,6,7,8-, octa-BDFs; EDF-5408, Cambridge Isotope Laboratories)
126
and 13C12-PBDEs (BDE-28, -47, -99, -100, -153, -154, -183, -197, -207, -209; Wellington
127
Laboratories), and then cleaned-up with concentrated sulfuric acid and multi-layer silica
128
gel. The cleaned-up extract was loaded onto an activated carbon-impregnated silica gel
129
reversible cartridge (Kanto Chemical) for separation into a non-dioxin fraction containing
130
PBDEs (eluted with 30 ml of a mixture of 25% dichloromethane (DCM) in hexane) and
131
then a dioxin fraction (reverse-eluted with 80 ml of toluene). The non-dioxin fraction was
132
further cleaned-up using gel-permeation chromatography (packed Bio-Bead S-X3, Bio-
133
Rad Laboratories, eluted with 150 ml of a mixture of 50% DCM in hexane), and then
134
spiked with 13C12-BDE126 and -BDE205 (Wellington Laboratories), whereas the dioxin
135
fraction was spiked with 13C12-1,2,3,4-/-1,2,3,7,8,9-CDD (Wellington Laboratories) and
136
13C
137
concentrated under gentle nitrogen stream down to 50 µl.
12-1,2,3,7,8,9-BDD
(Cambridge Isotope Laboratories). Both fractions were
138
PCDD/Fs and PBDD/Fs were determined using a HP-6890 gas chromatograph (GC,
139
Agilent Technologies) connected to a JMS-700D magnetic sector high resolution mass
140
spectrometer (HRMS, JEOL) operating in electron ionization selective ion monitoring
ACS Paragon Plus Environment
10
Page 11 of 42
Environmental Science & Technology
141
(EI-SIM) mode at a resolution of >10,000 (10% valley). For tetra- to octaCDD/Fs, the
142
GC column was a BPX-Dioxin-I (SGE Analytical Science, 0.15 mm internal diameter ×
143
unspecified film thickness × 30 m length) and the temperature program was 160 °C
144
(1 min), 20 °C/min to 200 °C, and then 3 °C/min to 300 °C. For tetra- to hexaBDD/Fs, a
145
DB-17ht column (Agilent Technologies, 0.25 mm × 0.15 µm× 30 m) was used with a
146
temperature program of 160 °C (1 min), 20 °C/min to 220 °C, 4 °C/min to 300 °C (hold
147
10 min). For hepta- to octaBDD/Fs, a DB-5ms column (Agilent Technologies, 0.25 mm ×
148
0.1 µm× 15 m) was used with a temperature program of 170 °C (1 min), 15 °C/min to
149
260 °C, 10 °C/min to 310 °C (hold 8 min). Tri- to decaBDEs (BDE-17, -28, -30, -47, -49,
150
-66, -71, -77, -79, -85, -99, -100, -118, -119, -126, -138 to -140, -153 to -156, -171, -180,
151
-183, -184, -191, -196, -197, -201, -203 to -209) were determined using a 7890A GC
152
(Agilent Technologies) connected to a 5975C quadrupole MS (qMS, Agilent
153
Technologies) operating in EI-SIM mode. The GC column was a DB-5ht (Agilent
154
Technologies, 0.25 mm × 0.1 µm× 15 m) and the temperature program was 110 °C
155
(1 min), 40 °C/min to 135 °C, 15 °C/min to 215 °C, 5 °C/min to 250 °C, and then
156
40 °C/min to 310 °C (hold 10 min).
ACS Paragon Plus Environment
11
Environmental Science & Technology
157
Page 12 of 42
Every set of seven samples was accompanied by a procedural blank. The
158
concentration of each homologue group was calculated from the areas of both identified
159
and unidentified peaks. The recoveries of the labelled surrogate standards were from
160
66% to 138%, and concentrations of the native compounds were adjusted accordingly.
161
All results were expressed on a dry weight basis.
162
Screening and semi-quantitative analysis of PXDFs and PXDEs. Six samples from
163
two open burning areas and the dismantling area were subjected to comprehensive
164
screening of PXDFs and PXDEs, two samples with the highest concentrations of
165
PCDD/Fs and PBDD/Fs were selected for each of these three areas to confirm the
166
representativeness of the chemical profiles. A 0.5-g equivalent portion of each selected
167
soil extract was subjected to a non-destructive clean-up procedure consisting of
168
activated silica gel chromatography (Wakogel® DX, Wako, 4 g, elution with 80 ml of a
169
mixture of 5% dichloromethane in hexane) and gel-permeation chromatography, and
170
then separated into a non-dioxin fraction and a dioxin fraction using an activated
171
carbon-impregnated silica gel reversible cartridge as described above. Both fractions
ACS Paragon Plus Environment
12
Page 13 of 42
Environmental Science & Technology
172
were analyzed using a GC×GC–ToFMS system consisting of a KT2006 modulator
173
(Zoex, 8 s modulation time) installed on a 7890A GC (Agilent Technologies) and a JMS-
174
T100GCV MS (JEOL) operating at a nominal sampling rate of 25 Hz and a resolution of
175
>5000 (full width at half maximum). The first column was a DB-1ms Ultra Inert (Agilent
176
Technologies, 0.18 mm × 0.18 µm × 20 m) and the second a BPX-50 (SGE Analytical
177
Science, 0.1 mm × 0.1 µm × 0.75 m). The temperature program was 160 °C (1 min),
178
20 °C/min to 250 °C and then 4 °C/min to 350 °C. Mass drift compensation was
179
performed by adjusting with reference PBDF and PBDE congeners (listed in Table S1).
180
Blobs corresponding to tetra–heptaXDFs and tetra–heptaXDEs were then identified on
181
the two-dimensional chromatogram using the following criteria: presence of ions
182
matching the theoretical m/z (±0.02 window) of the two most abundant ions of the
183
molecular ion cluster (M+), their intensity ratio within ±15% of the theoretical value,
184
presence of characteristic ions ([M−COBr]+ for PXDFs, [M−2Br]+ or [M−BrCl]+ for
185
PXDEs), and retention times either close to those of the authentic standards or
186
predictable from those of known homologues (for homologues without authentic
187
standards). Highly brominated PXDFs and PXDEs could not be detected because of
ACS Paragon Plus Environment
13
Environmental Science & Technology
Page 14 of 42
188
thermal degradation in the GC×GC modulator. However, four highly brominated PXDE
189
homologues (Br7–8Cl2DEs and Br8–9ClDEs) were identified with GC–qMS by monitoring
190
the theoretical m/z of their characteristic [M−2Br]+ ions using the same GC conditions as
191
PBDE analysis. PXDFs were semi-quantified based on the concentrations of PBDFs
192
obtained with HRMS, and PXDEs based on qMS-derived PBDE concentrations. The
193
response factor of a compound was assumed as identical to that of the standard isomer
194
if available, or interpolated from those of the two closest homologue groups in terms of
195
molecular weight (Table S1). Identification of PXDDs was not attempted considering
196
their considerably lower expected concentrations.
197
RESULTS AND DISCUSSION
198
Distribution of PCDD/Fs, PBDD/Fs and PBDEs. The concentrations of PCDD/Fs,
199
PBDD/Fs and PBDEs in surface soil samples from the Agbogbloshie e-waste site varied
200
widely, ranging over one to two orders of magnitude even within the same area
201
(Table 1). The contamination levels of PCDD/Fs and PBDD/Fs were not significantly
202
different between the open burning areas and the dismantling area (Wilcoxon’s rank
ACS Paragon Plus Environment
14
Page 15 of 42
Environmental Science & Technology
203
sum test, p > 0.05). However, the highest concentrations of total PCDD/Fs were found
204
in the open burning areas (1.3–380, median 33 ng/g as opposed to 5.6–230, median
205
16 ng/g), whereas the highest concentrations of total PBDD/Fs were found in the
206
dismantling area (23–1000, median 450 ng/g as opposed to 11–870, median 180 ng/g).
207
The dismantling area also had significantly higher concentrations of total PBDEs (140–
208
6900, median 2800 ng/g as opposed to 21–2100, median 480 ng/g; p < 0.001). In the
209
other areas, the concentrations of all target compounds were lower by at least an order
210
of magnitude.
211
Table 1. Concentrations (ng/g dry weight) of PCDD/Fs (Cl4–8DD/Fs), PBDD/Fs (Br4–
212
8DD/Fs)
213
waste site (Ghana) in 2013
and PBDEs (Br3–10DEs) in soil samples collected from the Agbogbloshie e-
Open burning areas (n = Dismantling area (n = 14)
Other areas (n = 13)
14) Range
Median
Range
Median
Range
Median
Cl4DDs
0.11–19
1.2
0.76–45
1.4
0.043–0.73
0.19
Cl5DDs
0.13–32
1.5
0.70–46
1.6
0.036–1.1
0.21
Cl6DDs
0.081–23
0.79
0.37–23
0.98
0.026–1.0
0.13
ACS Paragon Plus Environment
15
Environmental Science & Technology
Page 16 of 42
Cl7DDs
0.12–19
0.42
0.18–14
0.85
0.025–0.94
0.11
Cl8DD
0.11–12
0.32
0.16–8.5
0.85
0.054–0.51
0.10
Total PCDDs
0.75–110
4.4
2.2–140
8.0
0.18–4.2
0.69
Cl4DFs
0.22–100
8.1
1.5–43
4.2
0.082–2.0
0.70
Cl5DFs
0.20–100
6.2
0.98–31
4.6
0.066–3.0
0.77
Cl6DFs
0.064–42
2.1
0.36–14
1.5
0.027–2.6
0.28
Cl7DFs
0.039–24
1.4
0.15–6.7
0.69
0.013–2.7
0.13
Cl8DF
0.0098–5.9
0.25
0.046–1.5
0.16
0.0036–0.77 0.027
Total PCDFs
0.53–280
18
3.3–97
11
0.19–11
Br4DDs