Subscriber access provided by University of Newcastle, Australia
Article
Thermodynamic modelling of the solubility and chemical speciation of mercury and methylmercury driven by organic thiols and micromolar sulfide concentrations in boreal wetlands Van Liem-Nguyen, Ulf Skyllberg, and Erik Björn Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.6b04622 • Publication Date (Web): 01 Mar 2017 Downloaded from http://pubs.acs.org on March 8, 2017
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 free 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 accessible to all readers and 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.
Environmental Science & Technology 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 31
Environmental Science & Technology
1
Thermodynamic modelling of the solubility and
2
chemical speciation of mercury and
3
methylmercury driven by organic thiols and
4
micromolar sulfide concentrations in boreal
5
wetlands
6
VAN LIEM-NGUYEN§, ULF SKYLLBERG #*and ERIK BJÖRN§*
7 8 9
Department of Chemistry, Umeå University, SE-901 87 Umeå, Sweden
§
#
Department of Forest Ecology and Management, Swedish University of Agricultural
Sciences, SE-901 83 Umeå, Sweden
10
* Corresponding author phone: +46 (0)90-786 84 60; e-mail:
[email protected] 11
* Corresponding author phone: +46 (0)90-786 51 89; e-mail:
[email protected] 12
1 ACS Paragon Plus Environment
Environmental Science & Technology
Page 2 of 31
13
ABSTRACT
14
Boreal wetlands have been identified as environments in which inorganic divalent mercury
15
(HgII) is transformed to methylmercury (MeHg) by anaerobic microbes. In order to
16
understand this transformation and the mobility and transport of HgII and MeHg, factors and
17
conditions in control of the solubility and chemical speciation of HgII and MeHg need to be
18
clarified. Here we explore the ability of thermodynamic models to simulate measured
19
solubility of HgII and MeHg in different types of boreal wetland soils. With the input of
20
measured concentrations of MeHg, sulfide, eight low molecular mass thiols and thiol groups
21
associated with natural organic matter (NOM), as determined by sulfur K-edge X-ray
22
absorption near-edge structure (XANES) spectroscopy and Hg LIII-edge extended X-ray
23
absorption fine structure spectroscopy (EXAFS), the model could accurately predict
24
porewater concentrations of MeHg in the wetlands. A similar model for HgII successfully
25
predicted the average level of its concentration in the porewaters, but the variability among
26
samples, driven mainly by the concentration of aqueous inorganic sulfide, was predicted to be
27
larger than measurements. The smaller than predicted variability in HgII solubility is discussed
28
in light of possible formation of colloidal HgS(s) passing the 0.22 µm filters used to define
29
the aqueous phase. The chemical speciation of the solid/adsorbed and aqueous phases were
30
dominated by NOM associated thiol complexes for MeHg and by an equal contribution from
31
NOM associated thiols and HgS(s) for HgII.
2 ACS Paragon Plus Environment
Page 3 of 31
32
Environmental Science & Technology
INTRODUCTION
33
Mercury (Hg) is a major threat to wildlife and human health.1, 2 The element has both
34
natural and anthropogenic sources, and is spread globally via atmospheric processes and in
35
oceans. The most abundant organic form of mercury: methylmercury (MeHg) is neurotoxic
36
and is formed from divalent inorganic mercury (HgII) in various environments, linked to the
37
activity of phylogenetically diverse microorganisms.3-8 Methylmercury is more toxic than
38
inorganic mercury and it bioaccumulates readily in aquatic and terrestrial food-webs. The
39
solubility as well as availability for methylation, demethylation and bioaccumulation
40
processes of Hg in soils and waters are largely controlled by its chemical speciation.9
41
Mercury is classified as a soft type B metal, meaning it has a particularly high affinity
42
for inorganic and organic reduced sulfur (sulfide and thiol, respectively) ligands.10 Reactions
43
between HgII and sulfide may result in the formation of the solid phase metacinnabar, β-
44
HgS(s) in parallel to the formation of the aqueous complexes Hg(SH)20(aq), HgS2H-(aq) and
45
HgS22-(aq). Under low sulfidic conditions and neutral pH various types of polysulfides
46
HgSnSH-(aq) (n= 4-6) may also form.11 Thiol ligands (RSH) associated with natural organic
47
matter (NOM-RSH) have been shown to form two-coordinated complexes with HgII,
48
Hg(NOM-RS)2.12, 13 This structure is similar to complexes formed with low molecular mass
49
(LMM) thiol ligands, which also form three- and four-coordinated HgII-thiol complexes at
50
neutral and alkaline pH.14, 15
51
The often quoted review by Morel et al.16 in 1998 covered HgII chemical speciation in
52
aquatic and terrestrial environments, but without the inclusion of complexes with NOM and
53
LMM associated thiols. More recent studies have clearly demonstrated the importance of
54
NOM-RSH for the complexation of HgII 12, 17 and MeHg.18, 19 Yet, so far very few studies have
55
included the role of LMM-RSH in the chemical speciation modelling of HgII and MeHg in 3 ACS Paragon Plus Environment
Environmental Science & Technology
21
Page 4 of 31
56
natural environments.20,
57
determine ambient concentrations of LMM-RSH and because of the uncertainties regarding
58
stability constants of LMM-RSH complexes with HgII and MeHg. In early work LMM-RSH
59
concentrations occasionally have been reported to reach into the µM22 or even mM range23,
60
but more recent refined methods demonstrate that LMM-RSH commonly are present in the
61
range 1-200 nM in boreal wetlands24. In laboratory experiments using much higher
62
concentrations, LMM thiols have been shown to be important for the uptake and
63
transformation of HgII to MeHg by anaerobic bacteria.25, 26
This is mainly because of insufficient analytical methods to
64
In this study we select the most relevant thermodynamic constants dictating the
65
chemical speciation of MeHg and HgII in boreal wetland environments to test whether
66
thermodynamic models can predict the solubility of MeHg and HgII in wetland soil
67
porewaters. Thermodynamic reactions included the formation of complexes with LMM and
68
NOM associated organic thiols, inorganic sulfides and polysulfides in aqueous, solid and
69
adsorbed phases. The model thus allowed simulation of the chemical speciation of MeHg and
70
HgII in wetland soil – porewater systems including both aqueous phase and solid/adsorbed
71
phases. Total concentrations were used as input and the predictive power of the model was
72
evaluated against measured MeHg and HgII concentrations in the porewater.
73 74
MATERIALS AND METHODS
75
Sites description. Four of the eight boreal wetlands described in Tjerngren et al.27, 28
76
were included in this study. Briefly, four sites representative of northern and southern boreal
77
wetlands of Sweden were selected: Storkälsmyran (SKM), Kroksjön (KSN), Långedalen
78
(LDN) and Gästern (GTN). The northern wetlands SKM and KSN are of a poor fen type, the
79
latter with some open water, while the southern ones are characterized as a bog-fen gradient at
80
LDN and a mesotrophic wetland-lake at GTN. These wetlands cover a range of environmental 4 ACS Paragon Plus Environment
Page 5 of 31
Environmental Science & Technology
81
conditions typical for boreal wetlands, such as range in pH (4.0–6.1), C/N mass ratio (16–53),
82
dissolved organic carbon (DOC) concentrations (18-240 mg L-1), and H2S(aq) concentrations
83
( SKM (0.56 µmol g-1) > KSN (0.32 µmol g-1)
206
> GTN (0.28 µmol g-1).
207
The average total concentration of Hg in soils was similar for the SKM, KSN and GTN
208
sites, about 470 pmol g-1 (94 ng g-1), and almost the double at the LDN site with 900 pmol g-1
209
(180 ng g-1), Table S3. The average MeHg soil concentration was higher at LDN (90 pmol g-1;
210
18 ng g-1) and SKM (47 pmol g-1; 9.4 ng g-1), than at KSN (19 pmol g-1; 3.8 ng g-1) and GTN
211
(10 pmol g-1; 2.0 ng g-1). Concentrations of NOM-RSH in soils and in porewaters, as well as
212
LMM-RSH(aq) and S-II(aq) in porewaters, are compared in Figure S4.
213
The iron(II)sulfide mineral mackinawite, FeS(s), is frequently reported to be present in
214
anoxic sediment. Its formation is described by the reaction: FeS(s) + H+ = Fe2+ + HS-, having
215
a log K in the range -2.95 to -3.5,41-43 depending on the degree of crystallinity of the FeS(s)
216
phase. On average, 30% of total Fe measured in our wetland porewaters was in the form of
217
FeII. Even if we conservatively assume that 100% of measured total Fe concentration was
218
represented by FeII (due to possible underestimation by the Fe(II,III) speciation analysis method
219
used), all porewaters invesigated in this study were highly undersaturated in relation to even
220
the least soluble form of FeS(s) as shown by the calculation in Table S6. In agreement with
221
this model prediction, FeS(s) was not detected by S XANES in any of the examined soil
222
samples, as exemplified by two samples in Figure S3. It should be noted that while S XANES
223
data alone would not rule out FeS(s), since the detection limit for this well-resolved XANES
224
peak may be on the order of 2-3 % of total S, XANES data combined with the highly
225
undersaturated porewater provide independent information suggesting the absence of any
226
FeS(s)-phase in our wetland soils. This is an important conclusion since FeS(s) may act as an
227
adsorbing surface for HgII and MeHg, and HgII-FeS(s) surface complexes may serve as a
228
precursor for HgS(s) formation.44 The mineral pyrite (FeS2) was likley present in some of our 10 ACS Paragon Plus Environment
Page 11 of 31
Environmental Science & Technology
229
soils, but it is not expected to affect the chemical speciation of HgII in presence of excess
230
reduced S ligands.9
231 232
Solubility and chemical speciation model of MeHg. The fit of the MeHg model was
233
evaluated by comparing measured and modeled porewater concentrations. Soil water content,
234
total soil concentrations of MeHg and HgII, concentrations of NOM-RSH(ads), NOM-
235
RSH(aq), LMM thiols and S-II(aq) as well as pH were used as input to the model. Dyrssén and
236
Wedborg reported a log K for MeHgSH0(aq) of 14.5, with the pKa of H2S set to 6.88.10 In our
237
model, we used a pKa of 7.00 for H2S33 and consequently the log K for the formation of
238
MeHgSH0(aq) was corrected to 14.6. In the literature, reported log K for the formation of
239
MeHgSR-NOM complexes (reaction 2) varies between 15.6 and 17.5, and the pKa of the thiol
240
group varies between 8.5 and 10 (reaction 6).18, 19, 30 Therefore, during fitting, the log K of the
241
MeHgSR-NOM complex and the pKa of the RSH group were allowed to vary between the
242
above values. The best agreement between measured and modeled concentrations of
243
MeHg(aq) in the dissolved phase was achieved with a log K of MeHgSR-NOM(aq,ads) of
244
17.5 and a pKa of 9.0 for the thiol group (Figure 1). The linear relationship between modeled
245
and measured data has a slope of 0.73 (ideally it should be 1.0) and a coefficient of
246
determination (R2) of 0.75 (n=31, p
