|
Final Report
Abstract
Monomethylmercury (MeHg) is the only form of mercury that
biomagnifies in aquatic foodwebs. Furthermore, since MeHg levels
within different watersheds and aquatic systems are not simply
proportional to their respective burdens of total Hg (Hg(-T)),
it is necessary to directly measure MeHg in these systems.
However, the scope of environmental research on MeHg is
restricted by the lengthy analytical procedures and high cost of
MeHg analysis by current methods. While total Hg (Hg(-T))
measurements can now be automated down to sub-parts per trillion
levels the same is not true of MeHg.
This project entailed two major tasks aimed at
improving our capabilities to measure MeHg in environmental
systems. The first task was to complete the development of a new
analytical method for measuring MeHg and inorganic Hg in
environmental samples. The new method, whose central ideas and
potential applications were laid out in our initial proposal,
has been refined to the point where we are now able to routinely
use it to make difficult environmental measurements of MeHg in
sediments and biota.
The second task was to use the system to analyze
MeHg in sediments of a contaminated urban watershed, i.e., the
Grand Calumet River watershed in northwestern Indiana. Using the
new method, we investigated spatial and seasonal variations in
MeHg levels in wetland sediments. Marked differences were found,
with late summer/early fall sediment samples having the highest
levels of MeHg. As has been observed elsewhere, MeHg becomes an
increasingly smaller fraction of the Hg(-T) as Hg(-T) increases.
Introduction
The major goals and objectives of this project were to: 1)
complete the development of and validate a new analytical method
for measuring MeHg, as well as inorganic Hg, in environmental
samples, 2) to identify sites of high methylating activity in a
contaminated watershed in the Lake Michigan drainage basin
(Grand Calumet River, IN), and 3) to test whether Hg methylation
and demethylation rates exhibit non-linear dependencies on Hg
contamination levels.
Narrative Report
A. Development of a Novel Ion Chromatographic Method for
Analyzing MeHg and Mercuric Hg in Environmental Sample Matrices
1. Introduction
The development of sensitive and selective methods for Hg
speciation analysis in gas, solid, and aqueous phases has
lead to a tremendous increase in our understanding of
environmental Hg contamination. Among the most important
findings of this research are that monomethylmercury (MeHg)
is the only form of mercury that biomagnifies in aquatic
foodwebs and that MeHg levels within different aquatic
systems are not simply proportional to the levels of total
Hg (Hg(-T)) deposited in or retained in these systems. As
important as it is to have these facts established, the
scope of environmental research remains limited strongly by
the lengthy time and high cost of MeHg analysis by current
methods. The same limitations profoundly affect the ability
of regulatory agencies to conduct sampling programs. While
total Hg (Hg(-T)) measurements can now be automated down to
sub-parts per trillion levels, MeHg analysis cannot. Thus,
it is imperative to develop new technologies for MeHg
analysis. As a part of this project, we completed the
development of and refined a novel liquid chromatography
system, which we call Hg-Thiourea Complex Ion
Chromatography, for mercury speciation analysis.
2. Methods
The analytical system (Figure 1) consists of: 1) a
high-pressure liquid chromatography system to separate
cationic thiourea complexes of MeHg+ and Hg2+ across an ion
chromatography column, 2) a low-pressure, flow injection
cold-vapor generation system with on-line photooxidation of
MeHg to HgII followed by reduction of HgII to Hg0, and 3) an
atomic fluorescence spectrometry (AFS) Hg detection system
that measures Hg0 transferred from the eluant into the
carrier gas stream. A custom-made thiol resin that is
relatively simple to fabricate is used as the trapping
column (TT). The system has a high throughput (one sample
every ~10 minutes) and uses external calibration.
Figure 1. System schematic. Hardware component
labels denote: [HPLC] High pressure, single-piston pump with
pulse-dampener; [CC] Cleaning column packed with sulfonated,
100%-DVB resin; [HPIV] High pressure injection valve with [SL]
100-μL sample loop and [TT] thiol preconcentration trap; [ICC]
ion chromatography column with mixed-mode resin; [UV-PCO] PTFE
tubing (~7 m) wound directly around an 8-W, low-pressure UV
bulb; [PP] Console drive peristaltic pump with mini-cartridge
pump head and Norprene™ pump tubes; [GLS] Thin-film diffusion,
gas-liquid separator (custom-built borosilicate glassware); [AFS]
Atomic fluorescence spectrometer with mass flow controller;
[A-D/PC] Analog-digital converter with output connected to a
personal computer running peak-integrating software; [NDT] a
48-in length Nafion™ gas drying tube. To stabilize its signal,
the AFS draws its power from an uninterruptible power supply
that was connected to a line conditioner. From Shade and Hudson
(in press). The key solutions used in the system include the
Eluant, which is acidic thiourea, the Oxidant (Ox), which
contains H2O2, the Antioxidant (AOx), which is ascorbic acid,
and the Reductant (Red), which is alkaline stannous chloride.
3. Results
There are several aspects of the system’s performance that
are worthy of note. First, it elutes MeHg before Hg2+, which
prevents Hg2+ peaks from interfering with the MeHg peaks.
This is important for analysis of samples with high levels
of Hg2+. Second, it completely oxidizes MeHg, so that the
system sensitivity is identical both for MeHg and Hg2+. This
makes calibration simpler for simultaneous detection of both
Hg species. Finally, the system can sustain a sample
throughput rate of at least one sample every ten minutes.
Certified reference materials – both sediments and biota –
made from environmental samples were analyzed using the
system (Table 1) and all results were within the precision
of the measurements. Analysis of calibration curve data
(Fig. 2) shows that the precision of the method is
excellent, with an RMS error of <5% above 10 pg absolute.
The absolute detection limit is < 1 pg Hg, which is
comparable to the standard method for MeHg analysis (EPA
Method 1630).
Table 1. Comparison of MeHg and Hg(-T) levels
measured in digested and extracted reference materials with
certified values (CV). Samples were injected either directly via
the sample loop (SL) or concentrated on the thiol trap (TT).
From Shade and Hudson (in press).
Figure 2. Residuals (actual - predicted)/si for 17 MeHg
calibration curves (N = 95). Lines represent modeled coefficient
of variation (si/Ai) of peak area. RMS error is 5% at > 10 pg.
From Shade and Hudson (in press).
B. Sediment MeHg Extraction Methods Development
Wetland sediments are among the most difficult environmental
samples to analyze for MeHg because their high levels of reduced
sulfur compounds and organic matter strongly bind MeHg to the
matrix. A part of our effort on this project involved testing
and developing methods of extracting MeHg from polluted, highly
organic wetland sediments collected from marshes in the Grand
Calumet River watershed.
The new method employs 4 M HNO3 with lower
concentrations of HCl and CuSO4 to attack the sediment matrix
and release MeHg as a MeHgCl complex during a 1-h incubation
with shaking. This neutral complex then partitions into an
organic solvent, toluene, which is also shaken with the
sediment/leaching solution mixture for 1-h. The MeHg is
back-extracted from the toluene into an aqueous thiourea
solution. The resulting sample solution is preconcentrated
on-line over a thiol resin and eluted into the Hg-Thiourea
Complex IC-FIA CVAFS system for analysis.
The new extraction method was compared with the
widely-used H2SO4/KBr/CuSO4 leaching followed by solvent
extraction. Both methods perform equivalently with the reference
marine sediment examined, the HNO3 method gives ~95% extraction
efficiency with the more difficult wetland sediments compared
with ~80% for the standard H2SO4-based method. Based on a
chemical consideration of the nature of the extractants, we
suggest that the efficacy of the HNO3-based extraction is due to
the oxidative effect of the acid on the sediment matrix since
the coordinative strength of the ligands remains similar in the
two extractants.
Results of spike recoveries from wetland
sediments using the conventional leaching recipe and the new
nitric acid recipe support our interpretations of the above
experiments (Table 2). The fact that both leaching methods yield
good results for reference materials, but vary widely in results
from fresh, highly-organic sediments indicates that reference
materials alone may not be enough to prove efficacy of sediment
MeHg extraction methods, especially for very difficult samples
such as these. Note that the new method has the capability to
adapted to quantitatively extract the sediment without drying, a
step that adds a degree of uncertainty about species
transformation.
Table 2. MeHg measurements and spike recoveries for certified
reference materials and Roxana Marsh sediments using new method
(in blue), HBr/CuSO4 leaching, and conventional extraction (in
grey). N = 3 for all measurements, except DOLT (N=2) as noted.
From Shade (2005).
C. MeHg and Hg in Wetland Sediments of a Contaminated Urban
Watershed
1. Introduction
The Grand Calumet River is located at the southern tip of Lake
Michigan. Its watershed contains a steep gradient in
contamination, ranging from i) highly-contaminated areas that
bear the legacy of pollution from the intense industrial
activities in the cities of East Chicago, Hammond, and Gary to
ii) much less contaminated parkland in the Indiana Dunes
National Lakeshore. Total Hg concentrations in the water
discharged by the Grand Calumet River into Lake Michigan are
elevated ~10-fold relative to lakewater and are the second
highest of all rivers discharging into the lake (Hurley et al.,
1998). Within its watershed, fish consumption advisories due to
high Hg levels have been instituted (IDEM, pers. comm.) and
sediment Hg levels up to 20 ppm have been measured in the West
Branch of the river (Cahill et al., 1999), although MeHg levels
are much more modest. Due to the presence of these high levels
of Hg and other contaminants, the Grand Calumet is currently the
subject of programs to develop a Remediation Assessment Plan
(RAP) and a Total Maximum Daily Load (TMDL). The largest repository of Hg in contaminated systems invariably
lies in their soils and sediments. The fate of this Hg depends
on the biogeochemical processes that take place there. In
particular, since sulfate reducing bacteria are the main
methylators of MeHg in freshwater systems, it is wetland soils
and sediments where the greatest production of MeHg occurs (St.
Louis et al., 1995). This MeHg can have ecotoxicological impacts
through its hydrologic transport to surface waters or through
its biomagnification in food webs in the wetlands themselves.
Our purpose here is to investigate the distribution of MeHg and
total Hg (Hg(-T)) in wetland sediments of the Grand Calumet
watershed and identify geochemical controls on their spatial
distribution and temporal variations.
2. Methods
Several sites in wetlands were selected roughly in an east-west
transect across the Grand Calumet watershed (Table 1). Samples
were collected on 3 different dates: June 2003, November 2003,
and September 2004. A handheld GPS unit was used to guide field
workers to the same location on each sampling date. Triplicate
cores were collected by inserting PVC-tubes (2-in. ID ´ 6-in. L)
into wetland soils and capping the ends before pulling them out.
Once removed, the bottom end was immediately sealed and the core
preserved on ice. Back in the UIUC lab, the cores were extruded,
sectioned into intervals of 0-2 cm, 2-5 cm, and 5-10 cm, placed
in sample jars, and frozen until analysis. MeHg in sediment samples was analyzed using the methods
described above in B) for the latter two sampling dates. Samples
from the June 2003 date were analyzed using the conventional KBr/Cu/H2SO4
extraction procedure (also see B) above). Total mercury analyses
were conducted using a LECO Hg analyzer, courtesy of Prof. Gabe
Fillipelli (Dept. of Geology, IUPUI).
Table 1. Wetland Sites Selected for Grand Calumet River
Watershed MeHg Study.
Sedimentary concentrations of AVS and CRS plus water content
were determined in sub-samples taken from core sections. AVS and
CRS were measured using a diffusion-based technique (Canfield et
al.1986; Brouwer and Murphy, 1994; Mulvaney and Khan, 2001).
Sediment samples (~1-g) taken from the small sub-cores were
placed in vessels containing a nitrogen atmosphere (Miller and
Wolin, 1974) and sealed. Upon injection of 1 M HCl (~10 mL) into
the vessel, the H2S formed from FeS-containing sediments
diffused to a sulfide trap (~3 mL of sulfide anti-oxidation
buffer) suspended above the acidified sediments inside the
vessel. Sulfide accumulated in the trap was analyzed using an
ion selective electrode to determine AVS. After renewing the
sulfide trap, CRS was measured by addition of a strongly acidic
CrII solution and measuring the amount of additional sulfide
generated. 3. Results
To date, we have conducted the following numbers of MeHg
analyses: i) June 2003: 25 samples, ii) Nov. 2003: 40 samples,
and iii) Sep. 2004: 46 samples. In addition, 147 measurements of
Hg(-T) have been made. The results of MeHg and Hg(-T) analyses
obtained so far are consistent with our expectations; Roxanna
Marsh sediments have the highest levels of both Hg(-T) and MeHg
and there is an east-to-west increasing gradient in Hg(-T), at
least up until near the end (Table 1). Note however that MeHg at
Roxanna increased less in proportional to Hg(-T) than would be
suggested by the MeHg/Hg(-T) ratio at other sites.
While we have not yet finished analyzing the data from the 3
sampling trips, the set of data from the June 2003 trip do show
intriguing relationships between sulfur fractions and
MeHg/Hg(-T) ratios (Fig. 3).
Figure 3. A) AVS as a function of TRS for June 2003 data. B)
depends less on Berit for doing little jobs.
Table 1. Average MeHg and Hg(-T) Concentrations in Wetland
Sediments from Sites Selected for Grand Calumet River Watershed
MeHg Study. Average is over 3-replicates and 3-depths.
Potential Applications or Benefits
The most significant applications of the new analytical methods
developed under this part project will be in the area of
environmental management. In making decisions about how to
remediate contaminated lake or river sediments, local data
concerning the long-term fate and transport of the hazardous
materials are needed. That is, environmental managers must
decide whether it is best to leave the contaminants undisturbed,
to seal them from surface exposure to immobilize them, or to
remove contaminated materials. In cases of mercury
contamination, the primary concerns are the production of
methylmercury (methylation), its release to nearby waterbodies,
and its subsequent bioaccumulation in food webs. It is important
to know whether highly mercury-laden sediments, as would be
found near old metal smelters in the Grand Calumet basin or in
contaminated sediments at the mouths or rivers draining to the
great lakes, provide a significantly larger source of
methylmercury than do sediments with more moderate or background
Hg loads. Such information would help determine what kind of
remediation will be most effective and the extent of clean up
needed.
Keywords
Methylmercury,
Mercury speciation analysis,
Wetlands,
Urban,
Watershed
International Implications
Mercury contamination is a global concern. Ultimately the
methodological advances from this project may receive widespread
application. Media Coverage
None yet, but a UIUC press release is pending.
Partnerships with other institutions/individuals initiated or
continued by your project
The main partnership established as a result of this study is
with Prof. Holger Hintelmann of Trent University, Peterborough,
ON. Publications
Accepted:
Shade, C.W. and R.J.M. Hudson, 2005, Determination of MeHg in
environmental sample matrices using Hg-Thiourea Complex Ion
Chromatography with on-line cold vapor generation and atomic
fluorescence spectrometric detection (Hg(-T)U/IC-FI-CVAFS),
Environmental Science and Technology, accepted Feb 2005.
In preparation:
Shade, C.W. and R.J.M. Hudson, Extraction procedures for
quantification of sediment methylmercury in contaminated wetland
sediments by Hg-Thiourea Complex Ion Chromatography, Analytical
and Bioanalytical Chemistry.
Graduate Students Supported
1) This project provided significant support for the Ph.D.
thesis project of:
Christopher W. Shade
Title: Hg-Thiourea Complex Ion Chromatography with On-line Cold
Vapor Generation and Atomic Fluorescence Spectrometric Detection
(Hg(-T)U/IC-FI-CVAFS): The Development of a Novel System for
Analysis of Monomethyl and Inorganic Mercury and of Compatible
Matrix-Specific Digestion Approaches for Environmental Samples.
Date Defended: January 14, 2005
Department of Natural Resources and Environmental Sciences
University of Illinois at Urbana-Champaign
2) This project provided partial support provided for the M.S.
thesis work of:
Hongbo Zhang
Title: Modeling copper complexometric titration of natural water
samples.
Date Defended: August 2004
Department of Natural Resources and Environmental Sciences
University of Illinois at Urbana-Champaign
3) This project also provided partial support provided for the
M.S. thesis work of:
Wade Wimer
Title: Effect of Environmental Mercury Contamination on the
Recruitment of
Trachemys scripta elegans.
To be completed: Fall 2005
Department of Natural Resources and Environmental Sciences
University of Illinois at Urbana-Champaign
Related Projects
Based on the methods developed in this project, we obtained
grants from the Great Rivers Environmental Research and
Education Center for 2003-2005 ($29,400 from a project with
Hudson as PI and $15,000 from a project with Soucek et al. as
PI’s) to study mercury in the Mississippi River floodplain and
tributaries.
Graphs, Figures and/or Photos
Figure 1 illustrates the analytical system that was refined for
this project. |
