Grantee Research Project Results
Final Report: Characterization of the Chemical Lability and Bioavailable Fraction of Heavy Metals in Natural Waters Using In-Situ Diffusion Gradient in Thin-Film (DGT) Probes.
EPA Grant Number: R828162Title: Characterization of the Chemical Lability and Bioavailable Fraction of Heavy Metals in Natural Waters Using In-Situ Diffusion Gradient in Thin-Film (DGT) Probes.
Investigators: Moffett, James W.
Institution: Woods Hole Oceanographic Institution
EPA Project Officer: Aja, Hayley
Project Period: June 1, 2000 through May 31, 2002
Project Amount: $224,949
RFA: Exploratory Research - Engineering, Chemistry, and Physics) (1999) RFA Text | Recipients Lists
Research Category: Water , Land and Waste Management , Air , Safer Chemicals
Objective:
The overall objective of this research project was to evaluate a new methodology for the determination of chemically reactive (bioavailable) concentrations of heavy metals in aquatic environments utilizing an in situ semipermeable membrane device. We proposed that diffusion gradient in thin film (DGT) probes may fill the void that currently exists in measurement capabilities by enabling inexpensive, rapid measurements of trace metals to be made, which will vastly expand the existing database for metal bioavailability, particularly during nonsteady state events. The specific objectives of this research project were to: (1) evaluate whether DGT probes can provide a direct measurement of the biologically available fraction of trace metals in aqueous environments; (2) investigate the use of probes to provide information about sources, sinks, and transport of heavy metals in complex systems (e.g., Boston Harbor) by making time-integrated measurements; and (3) explore modifications in gel chemistry to enable the probes to be used to determine total dissolved metals or free (inorganic) metal fractions.
Summary/Accomplishments (Outputs/Outcomes):
Our research was conducted at four different study sites. We conducted research at Boston Harbor, especially in the vicinity of the Prison Point Combined Sewage Overflow (CSO), at the mouth of the Charles River, a contaminated region subject to episodic inputs of metals during CSO discharge events, and at Vineyard Sound, MA, which is a pristine water body with low concentrations of metals. At Eel Pond, MA, we investigated a small yacht basin bordering Vineyard Sound with elevated metal concentrations, and at Childs River, MA, we conducted research at a low salinity, organic-rich site with elevated metal levels on the South Coast of Cape Cod.
The primary focus was Boston Harbor; the other sites were exploited primarily for methods development as they were close to our laboratory. Metal accumulation in the probes was studied in two ways:
· Deployment of the probes in situ for periods up to 72 hours at our study sites using simple mooring apparatus.
· Deployment in "mesocosms," which were 10 L polycarbonate carboys. These experiments were designed to evaluate the probes under conditions where we knew water chemistry did not change during the deployment.
The configuration of the probes was modified to optimize their ability to accumulate specific components of a given element's speciation. Cu, a primary focus of the project, can exist as inorganic complexes, organic complexes of varying strengths, inert complexes, and colloids. These species may comprise the 0.45 µm "total dissolved" fraction that is used by analysts and regulators alike. We are interested in probes that can be used to determine the total dissolved concentration, the inorganic fraction, and a labile fraction (corresponding to the total concentration less inert or extremely strong complexes). Three characteristics of the probes can be modified to render them more or less selective for specific fractions: (1) gel pore size (through crosslinking); (2) size cutoff filters between the gel and seawater; and (3) binding strength of the chelating agents on the metal binding resin. In this project, we tried various permutations of (1) and (2), but used only one metal binding resin, Chelex 100.
Three different types of gel were used: (1) agarose (AGE); (2) restricted (RGL); and (3) Lancaster (APA) gels. Diffusion coefficients for metal ions studied here were determined for each gel (see below). The most open gel was AGE, which also had the highest diffusion coefficient; this gel was used for probes that we hoped would yield estimates of total dissolved Cu.
RGL was the most highly crosslinked gel (cutoff point of >10,000 units of molecular weight [MW]), and was used for probes. We hoped they would only accumulate inorganic complexes, which are small and labile; however, in a previous study, we showed that RGL does not inhibit the uptake of organically complexed Cu by the cells. Therefore, a 500 MW cutoff (MWCO) filter also was added to the probe to further inhibit diffusion of organic complexes into the probe. We also investigated the use of a nafion layer instead of the 500 MWCO. Nafion is permeable to many analytes, but contains negatively charged functional groups that drastically lower the diffusion rates of negatively charged species such as naturally occurring dissolved organic matter.
Diffusion Coefficients. The AGE gel used has the same diffusion coefficients as the free ion in water. However, the RGL gel has a different diffusion coefficient, as this gel has a cutoff point of >10,000 MW and restricts the diffusion of metals bound to colloids and macromolecules. This has an effect on reducing the diffusion coefficient of the free ion. To determine the diffusion coefficients, a diffusion cell was manufactured to enable the diffusion coefficients of the different gels and MWCO membranes to be measured.
The diffusion cell was manufactured within Woods Hole Oceanographic Institution (WHOI). This enabled the diffusion coefficients for AGE, RGL, and APA gels to be determined for Cu and Cd. Subsequent coefficients then were calculated for Mn, Co, and Zn. Also, diffusion coefficients of the 500 MWCO membrane were determined.
Boston Harbor Sampling. Four fieldwork sampling trips to Boston Harbor were taken during 2001, and three sampling trips were taken in 2002. Ten different sites within the inner harbor have been sampled. The sampling sites were chosen due to their different environmental conditions such as salinity and proximity to CSOs, which are important sources of contaminants during storms. Two sites within the Charles and Mystic Rivers also were sampled. The sampling was conducted during March, May, July, and October 2001, and during March, June, and October 2002.
DGT probes were deployed using a plexiglass holder that held up to six probes. This enabled us to intercompare different probe types over the same deployment. The holder was attached to a polypropylene line with tie wraps and suspended 1 m below the surface. A small float was used for buoyancy, and plastic coated lead diving weights were used to anchor the mooring.
The DGT probes were deployed for a period of up to 42 hours in Boston Harbor and 72 hours in Vineyard Sound and Dyer Dock. After retrieval, the probes were dismantled in the clean laboratory, and the resin gel was eluted with 1 M nitric acid. The acid eluent then was analyzed (after dilution) using the Finnegan Element inductively coupled plasma-mass spectrometer (ICP-MS) at WHOI.
Water samples collected at the beginning and the end of the 24-hour gel deployments were filtered through a 0.2-mm filter and also through a 0.02-mm filter. These samples were analyzed directly (after dilution) on the ICP-MS. Collected samples also were acidified, UV irradiated, and analyzed for total dissolved Cu using the technique of competitive ligand exchange- adsorptive cathodic stripping voltammetry (CLE-ACSV).
Trace Metal Speciation-Free Metal Ion. Speciation measurements were conducted using CLE-ACSV (Moffett, 1995; Campos and van den Berg, 1994).
Three parameters were derived from these measurements: (1) free Cu2+; (2) inorganic Cu; and (3) total labile Cu (i.e., the total dissolved minus an operationally defined inert fraction [Kogut and Voelker, 2002] that is nonexchangeable even with large concentrations of added chelator). These parameters were directly compared with probe results.
DGT-AGE Cu versus UV total copper for the mesocosm data are shown in Figures 1 and 2. UV total was measured using CLE-ACSV with benzoylacetone as the competing ligand after filtration through a 0.2-µm filter followed by UV irradiation. The DGT-AGE Cu was obtained from a DGT unit with a 0.8-mm AGE gel diffusion layer and a 0.1-mm membrane (0.2 µm). There is consistency between the two; however, the difference is nearly 2:1, suggesting that some fraction of the Cu was inert to exchange.
Figure 1. Graph Showing DGT-AGE Versus "Exchangeable Cu" for the Mesocosm Data. The exchangeable is defined in Kogut and Voelker (2002).
Figure 2. Graph Showing DGT-AGE Versus SA Exchangeable for the Mesocosm Data. There is consistency between the exchangeable and the DGT-AGE, confirming our conclusion from Figure 1, that inert Cu does not react with the DGT.
Results suggest that the AGE gel only accumulates Cu, which is exchangeable with added ligands in CLE/ACSV.
We also compared DGT probes using Nafion and 500 MWCO filters with speciation data in waters from our study sites. Nafion data were inconclusive; they showed a large accumulation of metal, probably due to leakage of the probe or water ingressing around the Nafion. A redesign of the probe was undertaken, but leakage still occurred.
For the 500 MWCO DGT probes, consistency with inorganic Cu determined by anodic stripping voltammetry (ASV) was within a factor of four at all Boston Harbor stations. However, there was a wide disagreement between DGT results and CLE/ACSV in Vineyard Sound, where the probes overestimated inorganic Cu by more than a factor of 100. Differences are probably attributable to low MW organic complexes that can diffuse through the 500 MWCO filters. Two storm events were sampled at the Prison Point area in Boston Harbor in March 2001 and 2002. In March 2001, before the CSO activation, the DGT (AGE) Cu value over a 1-day deployment was 12.5 nM and the 0.2 mm total dissolved value was 33.8 nM. Twenty-four hours later, the CSO was activated, and the DGT probes were deployed for 36 hours during the activation. Cu concentrations increased to 25.6 nM for DGT and 74.8 nM for the 0.2 total dissolved fraction. In March 2002, during a shorter CSO activation, the Cu concentration for DGT (AGE) increased from 17.5 to 23.9 nM. The data from 2001 seem to highlight that the increase in Cu concentration due to the CSO activation is only detected by the total dissolved value and not by the DGT probes, thus indicating that a substantial amount of Cu is inert or "nonexchangeable," and not causing a serious toxicity problem. The CSO activation point is close the Charles River flood gates; large volumes of river water are introduced when CSO activation occurs and these waters are high in organic complexes, thus complexing any biologically available Cu into an inert form.
Massachusetts Water Resources Authority (MWRA) Chlorination Experiments. While collaborating with the MWRA, an opportunity arose to work together to solve a toxicity problem. Chlorination of their effluent was producing higher toxicity levels than they had expected. By using DGT with both AGE and RGL diffusion layers, we were able to show the robustness and usefulness of the gel probe.
Probe data showed a clear increase in the percentage accumulated by the DGT probes with an increase in the chlorination.
Conclusions:
DGT probes have considerable promise as a monitoring tool in complex chemical environments. Our results show that they have great potential in variable environments, particularly in waters where the presence of surfactants makes speciation measurements by voltammetry difficult or impossible. We have demonstrated a clear relationship between results using AGE as the gel and chemically exchangeable Cu determined by an independent method. Results suggest that during CSO discharge events, the spikes in bioavailable Cu over 24-hour probe deployments is relatively modest and probably does not pose a threat to organisms in the water column. Most likely, this reflects the high flow rates in the basins where the discharge occurs, because the outflows we studied were adjacent to rivers. However, we cannot use our results to derive mass balance estimates, because a significant fraction of the Cu is present as inert Cu and this is undetectable by our method. Zinc may be a more reliable diagnostic of plume input, because we anticipate that a smaller fraction of zinc is bound as inert complexes.
More work is needed to develop materials that will exclude organic complexes from accumulation; neither Nafion nor the 500 MWCO filters are effective in this regard.
Journal Articles:
No journal articles submitted with this report: View all 2 publications for this projectSupplemental Keywords:
water effects ratio, copper, silver, EPA Region 1, Boston Harbor, trace metal speciation, diffusion gradient in thin film, DGT., RFA, Scientific Discipline, Toxics, Waste, Ecosystem Protection/Environmental Exposure & Risk, Sustainable Industry/Business, Bioavailability, National Recommended Water Quality, cleaner production/pollution prevention, Environmental Chemistry, Sustainable Environment, Chemistry, Technology for Sustainable Environment, Monitoring/Modeling, Analytical Chemistry, Environmental Monitoring, New/Innovative technologies, Silver, lead, field monitoring, Zinc, measurement, sampling, copper, water quality, cadmium, heavy metals, membrane technologyProgress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.