Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
This is a previous version of the article published in Chemosphere. 2017, 169: 361-368. doi:10.1016/j.chemosphere.2016.11.074
Study of the presence of PCDDs/PCDFs on zero-valent iron nanoparticles
Blanca Calderon*a,b, Lisa Lundina, Ignacio Aracilb, Andres Fullanab
a Department of Chemistry, Umeå University, SE-901 87 Umeå Sweden.
b Permanent address: Chemical Engineering Department, University of Alicante, San Vicente del Raspeig Road, s/n
03690 San Vicente del Raspeig, Alicante, Spain.
* Corresponding author: Tel. +34 965 903 400 X.1116; Fax: + 34 965 903 826; email address: [email protected]
ABSTRACT
Studies show that nanoscale zero-valent iron (nZVI) particles enhance the formation of
chlorinated compounds such as polychlorinated dioxins and furans (PCDD/Fs) during thermal
processes. However, it is unclear whether nZVI acts as a catalyst for the formation of these
compounds or contains impurities, such as PCDD/Fs, within its structure. We analyzed the
presence of PCDD/Fs in nZVI particles synthesized through various production methods to
elucidate this uncertainty. None of the 2,3,7,8-substituted congeners were found in the
commercially-produced nZVI, but they were present in the laboratory-synthesized nZVI
produced through the borohydride method, particularly in particles synthesized from iron (III)
chloride rather than from iron sulfate. Total PCDD/F WHO-TEQ concentrations of up to 35
pg/g were observed in nZVI particles, with hepta- and octa-chlorinated congeners being the
most abundant. The reagents used in the borohydride method were also analyzed, and our
findings suggest that FeCl3 effectively contains PCDD/Fs at concentrations that could explain
the concentrations observed in the nZVI product. Both FeCl3 and nZVI showed a similar
PCDD/F patterns with slight differences. These results suggest that PCDD/Fs might transfer
from FeCl3 to nZVI during the production method, and thus, care should be taken when
employing certain nZVI for environmental remediation.
KEYWORDS: toxicity; PCDD/F; nanoparticles; nZVI; iron chloride
1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
1. INTRODUCTION
Nanoscale zero-valent iron (nZVI) has proven effective for removing a wide range of
environmental contaminants including chlorinated compounds, heavy metals and nitrates
(Cundy, et al., 2008; Yan, et al., 2013). Due to its nanometric size, it presents a large overall
surface area and numerous active sites, thereby remarkably enhancing the efficiency of
contaminant degradation (Boparai, et al., 2011). Consequently, nZVI is considered a promising
solution for contaminant remediation of soils, groundwater and wastewater, and is being applied
in pilot plant and field tests in both underground water and soil throughout Europe, the United
States of America (the U.S.), and Asia (Karn, et al., 2009).
The synthesis of nZVI is commonly performed using the wet chemical reduction of iron (II) or
(III) with sodium borohydride (see equations 1 and 2) because this reduction can occur at
ambient temperature and pressure (Elliott and Zhang, 2001; Zhang, 2003). Nevertheless, this
method is very costly for large production of nanoparticles due to the high cost of sodium
borohydride (Li, et al., 2009) and the concurrent production of large quantities of hydrogen
(Crane and Scott, 2012). As a result, nowadays, large-scale production of commercial
nanoparticles occurs mainly by using high temperature reduction of oxides in a hydrogen
atmosphere (He and Zhao, 2007). Other alternative methods such as grinding, ultrasound or
electrochemical methods are currently being developed (Han, et al., 2015; Stefaniuk, et al.,
2016; Wiesner, et al., 2006).
4 Fe3+¿+3 BH 4−¿+9 H 2O →4 Fe0 ↓+3 H 2B O3
−¿+12 H+¿+6 H2↑¿
¿¿ ¿ (1)
Fe(H2 O)62+¿+2BH 4
−¿→Fe 0↓+2B( OH )3+7 H 2↑¿ ¿ (2)
Chlorinated organic compounds have received much attention from laboratory scale and field
test researchers studying contaminant degradation by nZVI, due to their widespread presence in
contaminated sites [13, 14]. Studies show that nZVI is effective at degrading a large number of
these compounds, such as trichloroethylene, polychlorinated biphenyls (PCBs), and
2
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
23
organochlorinated pesticides (Li, et al., 2006; Varanasi, et al., 2007). The mechanism for
degradation is the adsorption of the contaminants into the nZVI surface, followed by their
reductive dehalogenation by electron donation (Yan, et al., 2013) (see equations 3 and 4):
Fe0 → Fe2+¿+2 e−¿¿¿ E0 = -0.44 V (3)
RCl+2e−¿+H +¿→ RH+Cl−¿¿¿¿ E0 = 0.5-1.5 V at pH = 7 (4)
Polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs), which are also chlorinated
organic compounds, have caused special concern. PCDD/Fs are persistent, highly toxic, organic
pollutants, especially those with chlorine atoms present at positions 2, 3, 7, and 8 of the
compound. Long term exposure to PCDD/Fs causes immunological, neurological and
reproductive effects, among others, and 2,3,7,8-TCDD is classified as carcinogenic by the U.S.
Environmental Protection Agency (Vallejo, et al., 2015). Given the nature of PCDD/Fs, the
Stockholm Convention established in 2001 that they should irreversibly be transformed or
destroyed to mitigate their ubiquitous presence in the environment. A few studies have
demonstrated effective methods for degrading PCDD/Fs in gas streams (Shu, et al., 2008), fly
ash handling systems (Weber, et al., 2002) and aqueous solutions (Kim, et al., 2008) with
metallic iron nanoparticles or related compounds.
Nevertheless, the high reactivity of nZVI, which makes it effective for environmental
remediation, could have undesirable effects on living organisms, as evaluated in some recent
studies (Crane and Scott, 2012). In addition to their direct effects on cells and organisms,
nanoparticles possess high surface area and strong adsorbent capacity, and could serve as
transport vectors of other contaminants, such as heavy metals and organic substances (Brar, et
al., 2010; Karn, et al., 2009). Furthermore, some studies have shown that iron, including nZVI,
can increase the production of PCDD/Fs under certain conditions. In a recent study by our team,
the addition of nZVI during thermal degradation of PCDD/Fs resulted in decreased amounts of
PCDD/Fs released into the gas phase, and a remarkable increase of PCDD/F content in the soil
(Lundin, et al., 2013). Also, a study by Font et al. showed that combustion of polyvinyl chloride
3
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
45
(PVC) in the presence of nZVI led to increased PCDD/F emissions up to a factor of 104 (Font, et
al., 2010). These studies demonstrate that contaminant degradation treatment using iron
nanoparticles can cause increased production of PCDD/F content in some way. It is still unclear
which mechanisms lead to the formation of PCDD/Fs with use of iron nanoparticles because no
investigation of this phenomenon has been found in the literature. Moreover, it is still unclear
whether the PCDD/Fs come from the nanoparticles, or whether nZVI itself participates in the
formation of PCDD/Fs under certain specific conditions.
This work is aimed to investigate the possible presence of PCDD/Fs in nZVI synthesized in the
lab with varying starting materials and commercially-available nZVI purchased from different
companies. Commercially-obtained nZVI, presumably produced through the dry chemical
reduction method, and lab-synthesized nZVI, using the wet chemical reduction method were
evaluated. In the case of synthesized nZVI, different raw materials were studied to determine
whether differences in PCDD/F content in nZVI depend on the raw materials used, and which
mechanisms (if any) lead to their formation. This investigation is highly novel and could
provide a better understanding of the possible toxic effects of nZVI, which is being widely
applied for the remediation of contaminants.
2. MATERIALS AND METHODS
2.1. Chemical reagents
Chemical reagents used to synthesize nZVI were selected from various sources and differed in
quality. Specifically, five different kinds of iron (III) chloride (FeCl3·6H2O), three kinds of iron
sulfate (FeSO4·7H2O), and two kinds of sodium borohydrate (NaBH4) were used.
Commercially-produced nZVI was purchased from two different companies, Nano Iron s.r.o.,
and Sigma Aldrich. Table 1 shows the specific properties and the origins of the reagents used
for nZVI synthesis, and for the commercially-available nanoparticles purchased for the work.
The rest of the chemicals used for PCDD/Fs analysis (sulfuric acid (H2SO4), n-hexane,
4
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
67
tetradecane, methanol, dichloromethane, etc.) were of analytical grade, and all aqueous
solutions were prepared using high quality purified water (Millipore).
Table 1. Properties of the reagents used for the nZVI synthesis, and of the commercial nZVI
Reagent Company Reference Purity Form Price (€/100g) Abbreviation
FeCl3·6H2O
VWR 24.208.260 ≥ 99% Lumps 10.1 FeCl3-VWRAlfa Aesar A16231.36 > 96% Lumps 3.8 FeCl3-AA-IAlfa Aesar 12497 97-102% Lump 9.6 FeCl3-AA-II
Sigma-Aldrich F2877 ≥ 98% Chunks 18.1 FeCl3-SA-I
Sigma-Aldrich 236489 97% Chunk 13.1 FeCl3-SA-II
FeSO4·7H2O
VWR 24.244.298 99.5–105.0% Powder 4.5 FeSO4-VWR
Alfa Aesar 14498.3 ≥ 99% Crystalline 9.4 FeSO4-AASigma-Aldrich F7002 ≥ 99% Powder 13.8 FeSO4-SA
NaBH4
Alfa Aesar 38788 ≥ 97.1% Caplets 69.9 NaBH4-AASigma-Aldrich 452882 ≥ 98% Powder 59.6 NaBH4-SA
nZVI
Sigma Aldrich 746843 99.5% Powder (35-
45 nm) 1076.0 nZVI-SA
Nanoiron, s.r.o.
NANOFER 25
14-18% of Fe in water
Aqueous suspension
(nZVI size < 50 nm)
6.1 nZVI-NI
2.2. Synthesis of nZVI
The following wet chemical reduction method, known as the borohydride reduction method but
with slight modifications, was used to synthesize nZVI. Briefly, 1.5 molar excess of an aqueous
solution of selected NaBH4 was added slowly to a solution of FeCl3 or FeSO4 in water
(depending on the case) to produce the nanoparticles. After that, the nanoparticles were
separated from the water by decantation, with the aid of a magnet, and then washed four times
with distilled water to remove the remaining borohydride. The synthesis of nZVI was performed
using both ferrous and ferric iron, using different qualities of FeCl3 and FeSO4, respectively.
PCDD/Fs extraction and analysis were performed immediately after nZVI synthesis.
2.3. PCDD/Fs analysis
5
104
105
106
107
108
109
110
111
112
113
114
115
116
117
89
The FeCl3, FeSO4 and NaBH4 were dissolved in 250 mL of 0.5 M H2SO4. Then, the solutions
were extracted with n-hexane by three sequential extractions with 100 mL each, and the extracts
were combined and concentrated to generate 40 l of tetradecane which was then cleaned
according to a method by Liljelind et al. (Liljelind, et al., 2003).
In short, 13C-labelled internal standards were added to the extract. Apart from the additional use
of 1,2,3,4,6-pentachlorodibenzofuran (1,2,3,4,6-PeCDF) and 1,2,3,7,8,9-
hexachlorodibenzofuran (1,2,3,7,8,9-HxCDF) with the labeled standards and the use of super-
activated carbon AX21-carbon, subsequent treatment and analysis of PCDD/F in these samples
followed the EU standard methods 1948:1-3. Samples were purified prior to PCDD/F analysis
using a multilayer silica column, a super alumina column, and a column comprising a pre-
washed mixture of AX21-carbon and Celite (7.9:92.1 w/w). The concentrations of PCDD/F
congeners were determined using gas chromatography-high resolution mass spectrometry (GC-
HRMS), equipped with a DB5 column and a Waters AutoSpec ULTIMA NT 2000D high
resolution mass spectrometer. Recovery of the internal standard was well within the limits set
by the EU standard methods 1948:1-3.
The accuracy and precision of the analysis were checked by measuring blanks and certified
standards. The analytes were quantified by the isotope dilution technique using 13C-labeled
internal and recovery standards. The World Health Organization 2005 Toxic Equivalency
Factors (WHO2005-TEF) were applied to obtain the toxicity of the samples in terms of PCDD/F
content. All analyses were performed in triplicate, and the standard deviation of all
measurements was determined and is represented as error bars in all figures.
3. RESULTS AND DISCUSSION
3.1. PCDD/Fs in nZVI particles
6
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
1011
Four different kinds of nanoparticles, two commercially-available and two lab-synthesized
through the borohydride method, were analyzed for PCDD/F content. Table 2 shows the results
of the concentrations of the 2,3,7,8-substituted congeners and Table S1 in Supporting
Information the concentrations of the total homologues of PCDDs and PCDFs. No PCDD/Fs
were detected in either of the commercially-produced nZVI particles, except for MoCDF and
DiCDD in the nZVI from Sigma Aldrich. Of the two lab-synthesized nZVI particles, the ones
produced using FeCl3 (nZVI-B-ferric) contained about 100 times more PCDD/Fs than those
produced using FeSO4 (nZVI-B-ferrous). More explicitly, the nZVI-B-ferric contained high
amounts of the highly-substituted PCDD/Fs, i.e. the hepta- and octa-chlorinated congeners
which are more abundant, with concentrations of 850, 830 and 3500 pg g-1 for HpCDD, HpCDF
and OCDF respectively. Two noteworthy findings were that the PCDD/Fs present in the nZVI-
B-ferrous particles were also the highly-chlorinated hepta- and octa-chlorinated congeners, and
that no other degrees of chlorination were detected except for the mono-chlorinated congeners.
The nZVI-B-ferric particles possessed approximately 5-fold more PCDFs than PCDDs.
No 2,3,7,8-substituted congeners were found in the commercially-produced nZVI particles. The
lab-synthesized nZVI-B-ferrous particles only contained 1,2,3,4,6,7,8-HpCDF and OCDF,
while nZVI-B-ferric particles contained all the 2,3,7,8-substituted PCDF congeners. The nZVI-
B-ferric particles possessed lower concentrations of 2,3,7,8-substituted PCDD than PCDF, but
neither 2,3,7,8-TCDD nor 1,2,3,4,7,8-HxCDD were detected.
However, since nZVI-B-ferric particles also contained significant concentrations of the most
toxic homologues, such as PeCDD, TCDD and PeCDF, their WHO toxic equivalent (WHO–
TEQ) concentration reached a value of 35 pg g-1. In the case of nZVI-B-ferrous particles, again
only 1,2,3,4,6,7,8-HpCDF and OCDF were detected, thus giving the material a lower WHO-
TEQ concentration than the measurement detection limit (0.1 pg g-1).
Consequently, we concluded that the commercially-produced nZVI particles possess a
negligible PCDD/F content, while lab-synthesized nZVI particles produced using the
7
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
1213
borohydride method, particularly the ones synthesized with iron chloride, possess significant
PCDD/F content. The possible origin of these persistent compounds will be further assessed in
subsequent sections.
Table 2. Concentrations of 2,3,7,8-substituted congeners in commercially-produced and lab-
synthesized nZVI. Concentrations are in pg g-1 sample. The detection limit of the analysis was
0.5 pg g-1
Commercially-produced Lab-synthesizednZVI-SA nZVI-NI nZVI-B-ferric nZVI-B-ferrous
2-MoCDD n.d. - n.d. n.d.23-DiCDD n.d. - n.d. n.d.237-TriCDD n.d. n.d. 1±0.2 n.d.2378-TCDD n.d. n.d. n.d. n.d.12378-PeCDD n.d. n.d. 3±0.7 n.d.123478-HxCDD n.d. n.d. n.d. n.d.123678-HxCDD n.d. n.d. 4±0.1 n.d.123789-HxCDD n.d. n.d. 5±1 n.d.1234678-HpCDD n.d. 1±0.2 19±3.1 n.d.OCDD n.d. n.d. 78±14 n.d.2-MoCDF 15±1.5 n.d. 4±0.8 n.d.28-DiCDF n.d. n.d. n.d. n.d.248-TriCDF n.d. n.d. 3±0.6 n.d.2378-TCDF n.d. n.d. 8±1 n.d.12378-PeCDF n.d. n.d. 23±3.7 n.d.23478-PeCDF n.d. n.d. 13±2.2 n.d.123478-HxCDF n.d. n.d. 97±12 n.d.123678-HxCDF n.d. n.d. 51±2.3 n.d.123789-HxCDF n.d. n.d. 17±2.6 n.d.234678-HxCDF n.d. n.d. 25±6.1 n.d.1234678-HpCDF n.d. n.d. 460±96 7±11234789-HpCDF n.d. n.d. 75±7.0 n.d.OCDF n.d. n.d. 3500±370 43±3.5
WHO2005-TEQ 0.0 0.01±0.002 35±5 0.1±0.01
- : could not be quantified due to disturbances during GC/HRMS analysis.
n.d.: under detection limit (<0.5 pg g-1)nZVI-SA: commercially-produced nanoparticles purchased from
Sigma-Aldrich (their characteristics can be viewed in Table 1).
nZVI-NI: commercially-produced nanoparticles purchased from Nano Iron, s.r.o.
nZVI-B-ferric: nZVI synthesized through the borohydride method using the reactants FeCl3-SA-I and
NaBH4-AA as raw materials.
8
168
169
170
171
172
173
174
175
176
177
178
179
180
1415
nZVI-B-ferrous: nZVI synthesized through the borohydride method using the reactants FeSO4-AA and
NaBH4-AA as raw materials.
3.2. Determination of the origin of PCDD/F in the nZVI
Secondly, we analyzed the raw materials commonly used to synthesize nZVI for laboratory and
field applications for PCDD/Fs to determine the origin of the PCDD/Fs detected in the lab-
synthesized nZVI particles. Synthesis of nZVI particles with the borohydride method can be
performed using either ferrous or ferric iron (see equations 3 and 4). Studies of the
environmental remediation of contaminants with nZVI use iron sulfate as a source of ferrous
iron and iron chloride for ferric iron for nZVI particle synthesis. NaBH4 is the reducing agent
for nZVI particle synthesis in a majority of studies (Fu, et al., 2014). Thus, we evaluated
FeCl3·6H2O, FeSO4·7H2O, and NaBH4 from various sources for PCDD/Fs content. Table 3
shows the results of the PCDD/F 2,3,7,8-substituted congeners analyses for iron chlorides and
iron sulfates, Table S2 in the Supporting Information their concentration of the total
homologues of PCDD/Fs and Table S3, shows the PCDD/F analysis for NaBH4.
Table 3. Concentrations of 2,3,7,8-substituted congeners in reagents used for nZVI particle
synthesis using the wet chemical reduction method. Concentrations are in pg g-1 sample. The
detection limit of the analysis was 0.5 pg g-1
FeCl3-VWR
FeCl3-AA-I
FeCl3-AA-II
FeCl3-SA-I
FeCl3-SA-II
FeSO4-VWR
FeSO4-AA
FeSO4-SA
2-MoCDD - - - - - - - -23-DiCDD - - n.d. n.d. n.d. - - -237-TriCDD n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d.2378-TCDD n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d.12378-PeCDD n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d.123478-HxCDD n.d. n.d. n.d. 0.5±0.06 n.d. n.d. n.d. n.d.123678-HxCDD n.d. n.d. n.d. 1±0.1 n.d. n.d. n.d. n.d.123789-HxCDD n.d. n.d. n.d. 0.5±0.04 n.d. n.d. n.d. n.d.1234678-HpCDD 2±0.2 3±0.5 n.d. 5±0.5 1±0.1 1±0.1 n.d. n.d.OCDD 8±2 14±2 1±0.1 26±3.5 2±0.6 3±0.7 1±0.1 n.d.
9
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
1617
FeCl3-VWR
FeCl3-AA-I
FeCl3-AA-II
FeCl3-SA-I
FeCl3-SA-II
FeSO4-VWR
FeSO4-AA
FeSO4-SA
2-MoCDF 84±29 37±8 0.5±0.02 17±3.2 2±0.4 n.d. n.d. n.d.28-DiCDF n.d. n.d. n.d. 1±0.1 n.d. n.d. n.d. n.d.248-TriCDF n.d. 1±0.1 n.d. 1±0.2 n.d. n.d. n.d. n.d.2378-TCDF n.d. n.d n.d. 1±0.2 n.d. n.d. n.d. n.d.12378-PeCDF n.d. n.d. n.d. 1±0.01 3±0.9 n.d. n.d. n.d.23478-PeCDF n.d. n.d. n.d. 1±0.1 n.d. n.d. n.d. n.d.123478-HxCDF n.d. n.d. n.d. 7±1.2 n.d. n.d. n.d. n.d.123678-HxCDF n.d. n.d. n.d. 1±0.1 n.d. n.d. n.d. n.d.123789-HxCDF 1±0.1 1±0.1 1±0.1 1±0.1 1. ±0.1 0.6±0.2 1±0.2 n.d.234678-HxCDF n.d. n.d. n.d. 1±0.1 n.d. n.d. n.d. n.d.1234678-HpCDF n.d. 3±0.3 1±0.2 23±2.9 n.d. n.d. n.d. n.d.1234789-HpCDF n.d. 1±0.2 n.d. 6±0.4 n.d. n.d. n.d. n.d.OCDF 2±0.1 216±21 1±0.5 740±74 4±1 2±0.3 1±0.2 n.d.
WHO-TEQ 0.1±0.02 0.2±0.02 0.1±0.02 2.1±0.3 0.1±0.04 0.1±0.02 0.1±0.03 0.0
n.d.: under detection limit (<0.5 pg g-1)
- : could not be quantified due to disturbances during GC/HRMS analysis
As shown in Table S2 and S3 in the Supporting Information, the iron chlorides contained
PCDD/Fs, while the iron sulfates and the NaBH4 were practically free of them. The FeCl3-SA-I
and FeCl3-AA-I samples had the highest PCDD/Fs content, followed by FeCl3-VWR. In
general, PCDFs were present in higher concentrations than PCDD, and the highly-substituted
PCDD/Fs were more abundant than less substituted ones, except for MoCDD/F, which was
present at remarkable concentrations.
Moreover, in the FeCl3-SA-I sample, all toxic PCDF isomers were present, resulting in a WHO-
TEQ concentration of 2.1 pg g-1. Although the toxicities of the other chlorides were not as high
as those of the FeCl3-SA-I sample, their values, which were lower than a WHO-TEQ
concentration of 0.2 pg g-1, should not be neglected. Since a 5:1 mass relationship (between the
iron salt reagent and nZVI particles, respectively) is used for nZVI particle synthesis, PCDD/F
transfer from the raw materials to the nZVI could generate high concentrations of PCDD/Fs in
the nZVI particles, thus producing significant toxicity values for the nZVI particles
10
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
1819
The iron sulfates had negligible concentrations of PCDD/Fs. Only 4 of the 2,3,7,8-substituted
congeners could be detected, 1,2,3,4,6,7,8-HpCDD, OCDD, 1,2,3,7,8,9-HxCDF and OCDF, all
of which were present at concentrations below 2 pg g -1. Consequently, they possessed low
WHO-TEQ concentration values (see Table 3). As expected, PCDD/F concentrations in the
sodium borohydride was very similar to those in the iron sulfates; no PCDD/F content was
detected.
PCDD/F content varied considerably from one chloride to another, as mentioned above. The
differences might be attributed to several factors: purity of the compounds, method of
production, etc. An observation of the properties of each compound (Table 1) reveals that
compounds possessing higher PCDD/F content are not necessarily less pure or less costly.
Indeed, FeCl3-SA-I had the highest PCDD/F content and toxicity, intermediate purity (>98%),
and the highest price. Moreover, FeCl3-VWR, which had a higher PCDD/F content than FeCl3-
SA-II and FeCl3-AA-II, also had a higher purity than the others (>99% compared to 97%).
Consequently, the variation between the samples in total PCDD/F content could be due to the
specific method of production of these compounds, and not simply to their purity.
According to the literature, iron (III) chloride solutions are prepared by dissolving iron or iron
ore in hydrochloric acid and oxidizing the resulting iron (II) chloride with chlorine:
Fe3O4(s) + 8 HCl(aq) → FeCl2(aq) + 2 FeCl3(aq) + 4 H2O (5)
2 FeCl2(aq) + Cl2(g) → 2 FeCl3(aq) (6)
In a continuous, closed-cycle process, iron (III) chloride solution is reduced with iron, and the
resulting iron (II) chloride solution is re-oxidized with chlorine. Solid iron (III) chloride
hexahydrate is then crystallized by cooling a hot concentrated solution (Elvers, et al., 1989).
Both the reagents and operating conditions used in all the steps of the iron (III) chloride
manufacturing process, starting with the raw materials, are likely to create a favorable
environment for PCDD/F formation at any stage. At the industrial level, iron ore is initially
11
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
2021
subjected to a sintering process that converts iron ore fines into larger agglomerates. A mixture
of fine ores, coke, lime or limestone, and iron-bearing residues is heated at high temperatures
and sintered into a porous feedstock acceptable for the blast furnace (Buekens, et al., 2001).
Since chlorine is unavoidable in the feed materials, PCDD/Fs are generated as byproducts of the
process, as previous studies have shown (Ooi and Lu, 2011; Sun, et al., 2016).
Ryan and Altwicker (Ryan and Altwicker, 2004) investigated the formation of PCDD/Fs with
mixtures of black carbon and three iron chloride types: iron (II) chloride tetrahydrate, iron (III)
chloride hexahydrate, and iron (III) oxychloride. They concluded that iron chlorides are
important chlorinating agents and promoters of low-temperature carbon gasification. The
predominant PCDD/F isomer was OCDF, as we have observed in this study. Kuzuhara et al.
studied the influence of several metal chlorides, including FeCl3·6H2O, on de novo PCDD/F
formation and concluded that iron chloride activity is important for the formation of PCDD/Fs
(Kuzuhara, et al., 2003). Lin et al. analyzed the influence of inorganic and organic flocculants
on PCDD/F formation during sewage sludge incineration (Lin, et al., 2015). They concluded
that the presence of poly-ferric chloride and polyaluminium chloride increased the formation of
PCDD/F during combustion experiments, suggesting that metals promote PCDD/F formation.
However, to our knowledge, none of these authors analyzed the PCDD/F content of the iron
compounds used in their studies. This analysis would have been useful for more accurately
determining whether iron compounds are catalysts for the formation of PCDD/Fs, or sources of
PCDD/Fs themselves. However, as the homologue patterns of PCDD/Fs changed as a function
of the reaction conditions and not only with the iron chloride concentration, it is more
reasonable to suggest that iron chloride could be involved in the de novo formation of PCDD/Fs
rather than being the only source of PCDD/Fs.
Accounting for all of the above mentioned findings, iron chloride may be responsible for
transporting PCDD/Fs into the iron nanoparticles. It is also expected that the higher the PCDD/F
content in the iron chloride, the higher the resulting PCDD/Fs concentration in the nZVI
particles.
12
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
2223
3.3. Transport and transformation of PCDD/Fs from iron chloride to nZVI during its synthesis
through the borohydride method.
During nZVI particle synthesis through the borohydride method, iron chloride is first diluted in
water. Consequently, any PCDD/Fs present in the iron chloride will also be present in the
aqueous solution. Sodium borohydride is slowly added to the ferric solution to produce the
nanoparticles, and the synthesis takes 30 to 60 minutes. When nZVI particles are formed,
PCDD/Fs present in the solution are most likely adsorbed into the formed particles. During
nZVI particle synthesis, PCDD/Fs may be chlorinated or dechlorinated, thereby changing the
homolog profile and isomer pattern. We compared the PCDD/F pattern in nZVI particles
synthesized from an iron chloride with a pre-determined PCDD/F content (nZVI-B-ferric) to the
expected PCDD/Fs pattern that would result if all PCDD/Fs coming from the iron chloride
source were attached to the nZVI particle surface during the synthesis (nZVI-calc). This
comparison was performed to assess changes in the isomer pattern or homolog profile due to
chlorination or dechlorination of PCDD/Fs during nZVI particle synthesis.
FeCl3-SA-I (24.2 g) and NaBH4-AA (15.24 g) were used to synthesize nZVI particles (5 g).
Thus, PCDD/Fs were expected to be concentrated in the nZVI particles by a factor of 4.8.
13
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
2425
0.001
0.01
0.1
1
10C
(pm
ol/g
)
nZVI-calc nZVi-ferric
0.001
0.01
0.1
1
10
C (p
mol
/g)
nZVI-calc nZVI-B-ferric
Figure 1. Concentrations of 2,3,7,8-substituted congeners (a) and homolog concentrations of
mono- up to octa-chlorinated PCDD and PCDF (b) of nZVI-B-ferric and nZVI-calc. The nZVI-
B-ferric was synthesized using FeCl3-SA-I and NaBH4-AA.
14
a)
b)
286
287
288
289
290
2627
Table 4. WHO-TEQ concentration, total PCDD and PCDF content, and degree of chlorination
in nZVI-calc and nZVI-B-ferric.
nZVI-calc nZVI-B-ferricpg g-1 pmol g-1 pg g-1 pmol g-1
SUM PCDD 1680±300 3.95±0.72 1260±220 3.17±0.58SUM PCDF 4250±470 10.4±1.2 5390±680 13.0±1.7SUM PCDD/F 5930±770 14.3±2.9 6650±900 16±2.6WHO-TEQ 10.0±1 0.026±0.004 34.4±5.2 0.092±0.014Chlorine content in PCDD (pmol Cl g-1) 28±5.0 20±3.3Chlorine content in PCDF (pmol Cl g-1) 73±7.7 93±12Degree of Chlorination in PCDD1 (mol Cl mol-1 PCDD) 7.0±2.1 6.1±1.9
Average chlorine content in PCDF (mol Cl mol-1 PCDF) 7.0±1.7 7.1±1.9
1Degree of Chlorination = Σ(Homolog sum × No. of Cl)/ ΣPCDD
The results shown in Figure 1 and Table 4 reveal that PCDDs patterns were very similar for the
more highly chlorinated PCDDs in nZVI-calc particles and in nZVI-B-ferric particles, further
supporting the hypothesis that PCDD/Fs from the FeCl3 is transferred to the nZVI particles
during the synthesis. However, a larger difference in PCDF patterns between nZVI-calc and
nZVI-B-ferric particles was observed, which may imply a transformation between the different
congeners during synthesis. The total PCDD/F content in nZVI-B-ferric particles was a bit
higher than, but very similar to, that which was expected theoretically (nZVI-calc). Thus, we
could conclude that the synthesized nZVI particles probably adsorbed all, or almost all, of the
PCDD/Fs from the iron chloride. The difference between the actual and expected values was
lower than the standard deviation, hence our conclusion that both values were very similar.
Nevertheless, there was a remarkable change in total WHO-TEQ toxicity, since nZVI-B-ferric
particles were 3.5 times greater than the theoretical value. This change was attributed to a
change in the PCDD/F isomer pattern of the toxic congeners present in the nZVI particles. A
closer look at the 2,3,7,8-substituted congeners presented in Figure 1a shows that some of the
isomers that highly influence the WHO-TEQ concentration, such as 1,2,3,7,8-PCDD, 1,2,3,7,8-
15
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
2829
PeCDF, 1,2,3,4,7,8-HxCDF, etc, were present at higher concentrations in the nZVI-B-ferric
than expected.
The chlorine content was very similar in both cases (calculated and ferric nZVI particles), and
when accounting for the standard deviation of the values in Table 4, we could not attribute any
difference between those values to the average chlorine content. Thus, we conclude that during
nZVI synthesis, although the PCDD and PCDF contents are similar in the starting iron chloride
raw material and in the finished nZVI product, there is a change in the isomer pattern towards
formation of some of the most toxic congeners. The change in isomer pattern leads to a
significant increase in the WHO-TEQ concentration of the finished nZVI product.
3.4. Environmental implications
This study has shown that the use of iron chloride as a source of iron in nZVI particle synthesis
through the borohydride method can lead to the adsorption of PCDD/Fs onto the nZVI particles
during the synthesis procedure. We suggest that PCDD/Fs originate from the iron chloride raw
material due to the method of iron chloride production. Our results have remarkable
environmental implications since nZVI is used for environmental remediation of water or soils.
Under certain conditions, PCDD/Fs could be released from the nZVI particles and delivered to
drinking water or waste water. The PCDD/Fs could then end up in sewage sludge, thereby
increasing the PCDD/F content in the environment and resulting in negative effects. Such
transport of other contaminants, such as heavy metals and organic substances, has already been
observed in other studies (Brar, et al., 2010; Karn, et al., 2009) and could also occur for
PCDD/Fs. Thus, from an environmental perspective and with respect to PCDD/F content, we
recommend nZVI to be synthesized using iron sulfate as a starting material instead of iron
chloride, or that iron chloride is pre-treated to remove PCDD/Fs before use for nZVI synthesis.
We should also consider the method of production of the nZVI when using them for
dechlorinating contaminant compounds. If the nZVI are produced from iron chloride using the
sodium borohydride method, the nZVI will likely adsorb PCDD/Fs onto its surface. Applying
16
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
3031
these nZVI particles as a specific treatment for removing any pollutant could lead to higher
concentrations of some chlorinated micropollutants in the environment than were present before
nZVI treatment. These increased concentrations of chlorinated micropollutants may lead to
unexplained or misunderstood results after any evaluations of nZVI treatments.
It is also noteworthy that this work demonstrated that some iron chlorides have a significantly
greater PCDD/F content than others which, to our knowledge, has not previously been shown in
the literature. Although the PCDD/F content in iron chlorides is not much higher than in other
substances, it must be taken into account, because iron chloride, which is a known coagulator in
waste water treatment, could be another source of PCDD/Fs in the environment. Iron chloride
may be one of other sources that partially contribute to an explanation for the occurrence of
PCDD/Fs in sewage sludge. PCDD/F levels encountered in sewage sludge samples as well as
total TEQ toxicity, and both homolog and isomer PCDD/F patterns, vary considerably
depending on several factors including the origin of the sludge (rural/urban/industrial areas) and
waste water treatment (Balasubramani and Rifai, 2015). As a matter of interest, Pereira and
Kuch studied the content of PCDD/Fs, PCBs, and metals in three sewage sludge samples
(Pereira and Kuch, 2005). Their results show a strong correlation between the iron content in
each sample (9.1, 19.7 and 60.1 g kg-1) and the PCDD/F WHO-TEQ values obtained (2.3, 26.9
and 128.5 pg WHO-TEQ g-1). Obviously, this finding does not prove that PCDD/Fs come from
use of ferric chloride as a coagulant in waste water treatment, but in any case we think it would
be interesting to study this in further detail.
4. CONCLUSIONS
This study showed that, under certain conditions, PCDD/Fs can appear in nZVI synthesized in
the laboratory through the borohydride method when iron (III) chloride is the ferric source. We
propose that iron (III) chloride, used as a reagent in the process, is the source of the PCDD/Fs
present in nZVI particles. Iron (III) chloride can contain impurities, including PCDD/Fs, that
17
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
3233
may have formed during its manufacture. Among the PCDD/F isomers encountered, hepta- and
octa-chlorinated PCDD/Fs were more abundant both in ferric chloride and nZVI particles, with
much higher concentrations of PCDFs than PCDDs. There were slight differences in the
patterns between ferric chloride and nZVI particles that influenced their respective total TEQ
values.
None of the toxic 2,3,7,8-substituted PCDD/F isomers were detected in the commercially-
produced nZVI particles. Thus, if lab-synthesized nZVI particles produced through the
borohydride method using iron (III) chloride are to be used for environmental remediation of
contaminants, the likely presence of PCDD/Fs in nZVI should be meaningfully considered, and
a preliminary cleaning treatment should be performed before these nZVI are used. Although the
maximum PCDD/F WHO-TEQ concentration observed in nZVI, (35 pg g-1) was not especially
high, these values must be acknowledged since nZVI have proven to be very efficient transport
vectors of contaminants that are adsorbed onto their surface.
Finally, we recommend further investigation of, but not limited to, the following: (i) whether the
potential risk that PCDD/Fs are transferred into the environment from nZVI exists; (ii) broader
spectra of ferric chloride samples to analyze their PCDD/F content; and (iii) whether the
presence of PCDD/Fs in ferric chloride is relevant when applying this compound during
processes such as waste water treatment. Any correlation between PCDD/F and iron content in
sewage sludge samples relative to the various sources of iron (such as ferric chloride, ferric
sulfate, etc.) could provide insight for the latter recommendation in particular.
ACKNOWLEDGEMENTS
The study was performed at the Environmental Chemistry Laboratory, Umeå University and the
PCDD/F analysis were done at the Trace Analysis Platform, Umeå University. We would like to
acknowledge the University of Alicante (EEBB-UA2013) for financing this study and Per
Liljelind for his valuable help with the PCDD/F analysis.
REFERENCES
18
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
3435
References
Balasubramani, A., Rifai, H.S., 2015. Occurrence and distribution of polychlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans (PCDD/Fs) in industrial and domestic sewage sludge. Environ. Sci. Pollut. Res. 22, 14801-14808.
Boparai, H.K., Joseph, M., O'Carroll, D.M., 2011. Kinetics and thermodynamics of cadmium ion removal by adsorption onto nano zerovalent iron particles. J. Hazard. Mater. 186, 458-465.
Brar, S.K., Verma, M., Tyagi, R.D., Surampalli, R.Y., 2010. Engineered nanoparticles in wastewater and wastewater sludge – Evidence and impacts. Waste Manage. 30, 504-520.
Buekens, A., Stieglitz, L., Hell, K., Huang, H., Segers, P., 2001. Dioxins from thermal and metallurgical processes: Recent studies for the iron and steel industry. Chemosphere. 42, 729-735.
Crane, R.A., Scott, T.B., 2012. Nanoscale zero-valent iron: Future prospects for an emerging water treatment technology. J. Hazard. Mater. 211-212, 112-125.
Cundy, A.B., Hopkinson, L., Whitby, R.L.D., 2008. Use of iron-based technologies in contaminated land and groundwater remediation: A review. Sci. Total Environ. 400, 42-51.
Elliott, D.W., Zhang, W.-., 2001. Field assessment of nanoscale bimetallic particles for groundwater treatment. Environ. Sci. Technol. 35, 4922-4926.
Elvers, B., Hawkins, S., Ravenscroft, M., Schulz, G., 1989. Ullmann's Encyclopedia of Industrial Chemistry. A, 14, 5th ed.
Font, R., Gálvez, A., Moltó, J., Fullana, A., Aracil, I., 2010. Formation of polychlorinated compounds in the combustion of PVC with iron nanoparticles. Chemosphere. 78, 152-159.
Fu, F., Dionysiou, D.D., Liu, H., 2014. The use of zero-valent iron for groundwater remediation and wastewater treatment: A review. J. Hazard. Mater. 267, 194-205.
Han, Y., Yang, M.D.Y., Zhang, W., Yan, W., 2015. Optimizing synthesis conditions of nanoscale zero-valent iron (nZVI) through aqueous reactivity assessment. Front. Environ. Sci. Eng.
He, F., Zhao, D., 2007. Manipulating the size and dispersibility of zerovalent iron nanoparticles by use of carboxymethyl cellulose stabilizers. Environ. Sci. Technol. 41, 6216-6221.
Karn, B., Kuiken, T., Otto, M., 2009. Nanotechnology and in situ remediation: A review of the benefits and potential risks. Environ. Health Perspect. 117, 1823-1831.
Kim, J.-., Tratnyek, P.G., Chang, Y.-., 2008. Rapid dechlorination of polychlorinated dibenzo-p-dioxins by bimetallic and nanosized zerovalent iron. Environ. Sci. Technol. 42, 4106-4112.
Kuzuhara, S., Sato, H., Kasai, E., Nakamura, T., 2003. Influence of metallic chlorides on the formation of PCDD/Fs during low-temperature oxidation of carbon. Environ. Sci. Technol. 37, 2431-2435.
Li, S., Yan, W., Zhang, W.-., 2009. Solvent-free production of nanoscale zero-valent iron (nZVI) with precision milling. Green Chem. 11, 1618-1626.
19
388
389390391
392393
394395
396397398
399400
401402
403404
405406
407408
409410
411412413
414415
416417
418419
420421422
423424
3637
Li, X.-., Elliott, D.W., Zhang, W.-., 2006. Zero-valent iron nanoparticles for abatement of environmental pollutants: Materials and engineering aspects. Crit. Rev. Solid Statae Mater Sci. 31, 111-122.
Liljelind, P., Söderström, G., Hedman, B., Karlsson, S., Lundin, L., Marklund, S., 2003. Method for multiresidue determination of halogenated aromatics and PAHs in combustion-related samples. Environ. Sci. Technol. 37, 3680-3686.
Lin, X., Li, X., Lu, S., Wang, F., Chen, T., Yan, J., 2015. Influence of organic and inorganic flocculants on the formation of PCDD/Fs during sewage sludge incineration. Environ. Sci. Pollut. Res. 22, 14629-14636.
Lundin, L., Moltó, J., Fullana, A., 2013. Low temperature thermal degradation of PCDD/Fs in soil using nanosized particles of zerovalent iron and CaO. Chemosphere. 91, 740-744.
Ooi, T.C., Lu, L., 2011. Formation and mitigation of PCDD/Fs in iron ore sintering. Chemosphere. 85, 291-299.
Pereira, M.D.S., Kuch, B., 2005. Heavy metals, PCDD/F and PCB in sewage sludge samples from two wastewater treatment facilities in Rio de Janeiro State, Brazil. Chemosphere. 60, 844-853.
Ryan, S.P., Altwicker, E.R., 2004. Understanding the role of iron chlorides in the de novo synthesis of polychlorinated dibenzo-p-dioxins/dibenzofurans. Environ. Sci. Technol. 38, 1708-1717.
Shu, H.C., Jhy, W.Y., Hung, M.C., Li, Y.H., Kai, H.C., Moo, B.C., 2008. PCDD/F adsorption and destruction in the flue gas streams of MWI and MSP via Cu and Fe catalysts supported on carbon. Environ. Sci. Technol. 42, 5727-5733.
Stefaniuk, M., Oleszczuk, P., Ok, Y.S., 2016. Review on nano zerovalent iron (nZVI): From synthesis to environmental applications. Chem. Eng. J. 287, 618-632.
Sun, Y., Liu, L., Fu, X., Zhu, T., Buekens, A., Yang, X., Wang, Q., 2016. Mechanism of unintentionally produced persistent organic pollutant formation in iron ore sintering. J. Hazard. Mater. 306, 41-49.
Vallejo, M., Fresnedo San Román, M., Ortiz, I., Irabien, A., 2015. Overview of the PCDD/Fs degradation potential and formation risk in the application of advanced oxidation processes (AOPs) to wastewater treatment. Chemosphere. 118, 44-56.
Varanasi, P., Fullana, A., Sidhu, S., 2007. Remediation of PCB contaminated soils using iron nano-particles. Chemosphere. 66, 1031-1038.
Weber, R., Nagai, K., Nishino, J., Shiraishi, H., Ishida, M., Takasuga, T., Konndo, K., Hiraoka, M., 2002. Effects of selected metal oxides on the dechlorination and destruction of PCDD and PCDF. Chemosphere. 46, 1247-1253.
Wiesner, M.R., Lowry, G.V., Alvarez, P., Dionysiou, D., Biswas, P., 2006. Assessing the risks of manufactured nanomaterials. Environ. Sci. Technol. 40, 4336-4345.
Yan, W., Lien, H.-., Koel, B.E., Zhang, W.-., 2013. Iron nanoparticles for environmental clean-up: Recent developments and future outlook. Environ. Sci. Process. Impacts. 15, 63-77.
20
425426427
428429430
431432433
434435
436437
438439440
441442443
444445446
447448
449450451
452453454
455456
457458459
460461
462463
3839
Zhang, W.-., 2003. Nanoscale iron particles for environmental remediation: An overview. J. Nanopart. Res. 5, 323-332.
21
464465
466
4041