Grantee Research Project Results
Final Report: Variability in the Properties of Dissolved Organic Matter and its Influence on Speciation of Heavy Metals (Cd, Cu, Pb)
EPA Grant Number: R825395Title: Variability in the Properties of Dissolved Organic Matter and its Influence on Speciation of Heavy Metals (Cd, Cu, Pb)
Investigators: Benoit, Gaboury , Yu-Ping Chin, P. I. , Fisher, Nicholas S
Institution: Yale University , The Ohio State University , The State University of New York at Stony Brook
EPA Project Officer: Packard, Benjamin H
Project Period: December 5, 1996 through December 4, 1998 (Extended to August 31, 2000)
Project Amount: $253,769
RFA: Exploratory Research - Water Chemistry and Physics (1996) RFA Text | Recipients Lists
Research Category: Water , Land and Waste Management , Safer Chemicals
Objective:
Understanding and predicting trace metal behavior (bioavailability, toxicity, reactivity, transport, and fate) depends on knowledge of metal speciation, especially the concentration of free dissolved ionic species, such as Pb2+×xH2O and Cd2+×xH2O. The objective of this research was to determine factors that control metal speciation, with an ultimate goal of generating predictive relationships between speciation and easily measurable biogeochemical parameters (like pH, DOM quantity and quality, and ionic strength) across a broad range of environmental conditions. To accomplish this objective, our research plan called for: (1) collection of samples from multiple sites covering a range of background chemistries; (2) filtration and ultrafiltration of samples to determine heavy metal partitioning across particles of all size classes (macro-particulate, colloidal, and truly dissolved forms); (3) use of DPASV to delineate the speciation of metals; (4) confirmation of DPASV results by measurement of metal speciation via two independent analytical methods--;potentiometric titrations on preconcentrated samples, and uptake experiments with algae; (5) measurement of the abundance and characteristics of DOM by high temperature combustion, high pressure size exclusion chromatography, acidimetric titrations, and spectroscopic methods; and (6) generation of quantitative predictive relationships between metal speciation and readily measurable DOM characteristics. Four of the sites were to be sampled bi-monthly in order to investigate seasonal variation in metal speciation and DOM characteristics, while many others were to be sampled once to expand the range of conditions studied. The multi-method approach employed several independent analytical techniques, for qualitative and quantitative comparisons, to determine concentration, partitioning, and speciation of trace metals in the water column and along river courses. To avoid contamination and filtration artifacts, all metal sample collection and analysis were conducted according to strict clean protocols, and size fractionation was carried out in the field wherever practicable.
In addition to these original goals, most of which were achieved, two further goals were added in response to initial findings of the research:
- An important serendipitous discovery was that dissolved sulfides often comprise a substantial portion of metal species even in fully oxygenated waters, where sulfides were previously thought to be absent (because of rapid oxidation to sulfate). This is highly significant, since low levels of sulfide can potentially outcompete DOM for metals.
- Additional experiments were conducted on the effect of membrane filtration,
when it was discovered that small variation in the extent of filter loading
led to large variations in the quantity of metals that were retained by the
filters. This has critical significance since the distinction between so-called
"dissolved" and "particulate" forms of metals traditionally
depends on this nearly universal separation technology.
Summary/Accomplishments (Outputs/Outcomes):
Overview of Research. High temporal resolution samples were collected monthly from January 1997 through June 1998 from a total of 7 river sites.1-3 This is more sites (7 instead of 4) and dates (monthly instead of bi-monthly) than originally planned. Eleven additional sites were sampled once in summer 1999, with follow-up sampling in 2000. All sampling, treatment, and analyses were conducted with clean methods to ensure artifact-free results. Samples were filtered (0.45 mm) and ultrafiltered (3,000 nominal molecular weight) before analysis in order to distinguish macroparticulate, colloidal, and truly dissolved fractions.4 The most important measurements conducted were voltammetric determinations of speciation for Cu, Cd, and Pb under real world conditions (not elevated levels of DOM and metals, as is usually done). This involved measurement of each river sample by DPASV while it was titrated with metals. This method yields values for conditional stability constants and complexation site abundances of the DOM. We have found that in the detection window of DPASV (i.e., for those sites with strengths suitable for measurement by this method), the data can be explained by the existence of a single (in terms of its conditional stability constant) complexation site type at each sampling location. In other words, there is no need to posit multiple complexation site types.Total Cu, Cd, and Pb were measured on splits of these samples for a mass balance check. In addition, all samples were analyzed for ancillary chemistry, including pH, DO, temperature, conductivity, suspended particulate matter, acid neutralizing capacity, HCO3-, dissolved organic carbon, PO43-, NO3-, Cl-, F-, SO42-, and Fe, Al, Na+, Ca2+, Mg2+, and K+.
In a separate component of the project, speciation measured by DPASV was compared to results from potentiometric titrations with a Cu ion selective electrode (Cu-ISE).5 These tests to verify the DPASV method against an independent technique were conducted on natural river waters and standard humic substances. The intercomparison of DPASV and ISE methods showed that both techniques produce similar results. This is important, since it confirms the utility of DPASV for measuring metal speciation in the rivers we are examining. DPASV has very low detection limits, but measures a labile pool of metals that includes not only free metal but also weakly-bound metal organic complexes. ISE detects free metal uniquely, but its detection limit is too high for use under ambient conditions in real world surface waters. The two methods have a small concentration range where they overlap, and we have shown that in this window the two yield concordant results.
An unanticipated benefit occurred when additional voltammetric measurements revealed the common presence of metal-sulfide species at all of our study sites.6-8 Trace levels of sulfides have previously been detected in surface ocean waters, but not yet in fresh waters. These reduced species exist and persist even though the rivers are fully-oxygenated. Some direct voltammetric measurements of metal-sulfide complexes have been carried out9, as well as measurements of total acid labile sulfides by the classic methylene blue method. We lowered the detection limit of this method by orders of magnitude10 in order to reliably measure sulfides in the low nM range of importance in environmental samples.
Size distribution was a final important element of this research. Samples were discriminated into macroparticulate (>0.45 mm), colloidal (>3,000 MW, but <0.45 mm), and truly dissolved (<3,000 MW) fractions. It became apparent that substantial differences in the distinction between macroparticles and colloids occurred depending on the way in which filtration was effected.11 Extensive experiments revealed a consistent, predictable relationship between size discrimination and filter loading for a variety of environmentally important compounds.12
Binding of Metals by DOM. We have documented a seasonal cycle in the amount of each metal that occurs in the free form (that can be directly detected by DPASV). Lesser amounts are present during the growing season and concentrations increase in the winter. In contrast, total metal levels did not vary systematically over the year. Concomitant with the variation in free metals is a seasonal change in the amount of DOC. As would be expected, the variation is in the opposite sense (high in summer, low in winter) to free metal levels. When the DPASV-measured complexation site strength and abundance are used in thermodynamic calculations, the variation in free metal can be predicted.
It is noteworthy that the amount of DPASV-labile free metal goes from measurable to undetectably low amounts in the course of the annual cycle. This suggests that the detection window of DPASV quantifies a dynamic pool of complexation sites that are actively gaining and losing metal over the course of the year. We do not measure weaker sites that are never occupied by trace metals, nor stronger sites that are always filled.
Another important conclusion is that the abundance of binding sites is directly proportional to the concentration of DOC as it changes throughout the year, while conditional stability constants are nearly invariant at each site. Among other things, this means that it is possible to sample streams at any time of the year and get representative DOM. This inference was important for the remaining sampling strategy, where we tested many additional streams once, to expand total coverage.
Perhaps our most intriguing finding is that, for all of our sites, there is a constant ratio of the abundance of complexation sites (L) to the quantity of DOM. Moreover, this ratio appears to assume one of three values, resembling either standard tannic, fulvic, or humic acid. This relationship was first observed for the seven intensively sampled sites, and confirmed with the other eleven, though the lowest L:DOM ratio seems to predominate. We are in the process of seeking easily measurable characteristics of the DOM that might predict its trace metal complexation behavior. In particular, the ratio of E4:E6 absorbance shows promise. It is well known that the composition and structure of DOM is highly variable, thus the occurrence of only a few discrete kinds of DOM, as is indicated by our analyses, is somewhat surprising. It is important to realize that this simple categorization is only in terms of its interaction with trace metals at environmentally realistic concentrations (of both metals and DOM). The apparent occurrence of only a few types may be because complexation sites whose combination of strength and abundance make them suitable to dominate trace metal speciation constitute a subfraction of total complexation sites on the DOM (ranging from very weak to very strong). Only a few discrete functional groups may match these requirements, sites that are measured within the detection window of DPASV. Previous investigators may have missed this simple behavior by use of analytical methods with other detection windows.
Metal-Sulfide Complexes. An important discovery has been that metal-sulfide complexes are commonly a significant fraction of all metal species in rivers. Amounts vary from site-to-site, but sulfides can amount to half or more of the metals, and this is true both in the colloidal and truly dissolved size fractions. There are clearly many possible sources of reduced sulfur in rivers, ranging from suboxic ground waters and reducing sediments to wetlands and sewage treatment plants. (Notably, we have measured sulfides at all sites, including some that are not downstream from any STPs.) The metal sulfides may persist because of stability imparted by the strength of the complexes. Lab tests we conducted indicate that half-lives with respect to oxidation are several days, long compared to water residence times in most rivers. While much of the metal occurs bound to sulfides, we do not find substantial amounts of free sulfides, so the ability of these ligands to influence speciation of any added metal may be slight.
We have measured both specific metal-sulfide complexes (voltammetrically) and lumped acid-labile sulfides (by the methylene blue method). The detection limit of the latter method (> 10 nM) is close to sulfide levels we measure in many sites, making reliable quantitation difficult. Consequently we have developed a new method, isolating the dye by HPLC, and lowering detection limits to well below 1 nM. This has allowed us to begin to study the systematics of sulfide distribution in a variety of aquatic environments, even when concentrations are extremely low.
Filtration Artifacts. While conducting repeated filtrations of some samples we found wide variations in the proportion of metals in filter-retained and filter-passing fractions. We conducted an extensive series of filtration experiments with a variety of filter types and using real waters collected from several study sites. These tests revealed that filters can quickly clog, causing the effective pore-size to drop dramatically. When this happens, colloids that initially passed through the filters begin to be retained. For substances that occur substantially in the colloidal size fraction (e.g., DOM, Fe, Pb), this can lead to significant errors in assigning the metals to "dissolved" or "particulate" fractions, as is often done based on filtration data. Fortunately, overloaded filters quickly reveal themselves by declining filtration rates and increasing back-pressure.
Journal Articles on this Report : 6 Displayed | Download in RIS Format
Other project views: | All 12 publications | 6 publications in selected types | All 6 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Benoit G, Rozan TF. The influence of size distribution on the particle concentration effect and trace metal partitioning in rivers. Geochimica et Cosmochimica Acta, January 1999;63(1):113-127. |
R825395 (Final) |
Exit |
|
Morrison M, Benoit G. Filtration artifacts caused by overloading membrane filters. Environmental Science & Technology 2001;35(18):3774-3779. |
R825395 (Final) |
not available |
|
Mylon SE, Benoit G. Subnanomolar detection of acid-labile sulfides by the classical methylene blue method coupled to HPLC. Environmental Science & Technology 2001;35(22):4544-4548. |
R825395 (Final) |
not available |
|
Rozan TF, Benoit G. Geochemical factors controlling free Cu ion concentrations in river water. Geochimica et Cosmochimica Acta, October 1999;63(19-20):3311-3319. |
R825395 (Final) |
not available |
|
Rozan TF, Benoit G, Mash H, Chin Y-P. Intercomparison of DPASV and ISE for the measurement of Cu complexation characteristics of NOM in freshwater. Environmental Science & Technology 1999;33(10):1766-1770. |
R825395 (Final) |
not available |
|
Rozan TF, Benoit G, Luther III GW. Measuring metal sulfide complexes in oxic river waters with square wave voltammetry. Environmental Science & Technology 1999;33(17):3021-3026. |
R825395 (Final) |
not available |
Supplemental Keywords:
watersheds, chemical transport, adsorption, ecological effects, bioavailability, chemicals, toxics, effluent, discharge, dissolved solids, aquatic, environmental chemistry, limnology, analytical, measurement methods, northeast, NE, Atlantic coast, EPA Region 1., RFA, Scientific Discipline, Toxics, Geographic Area, Water, Ecosystem Protection/Environmental Exposure & Risk, National Recommended Water Quality, Water & Watershed, exploratory research environmental biology, Physics, Environmental Chemistry, Ecosystem/Assessment/Indicators, Chemical Mixtures - Environmental Exposure & Risk, Ecosystem Protection, Chemistry, State, Ecological Effects - Environmental Exposure & Risk, Ecological Effects - Human Health, Biology, EPA Region, Watersheds, Ecological Indicators, dissolved organic matter, predictive species model, differential pulse anodic stripping voltammetry, lead, chemical contaminants, metal release, colloids, chemical kinetics, aquatic ecosystems, water quality, Connecticut (CT), heavy metal contamination, cadmium, Region 1Relevant Websites:
http://pantheon.yale.edu/~gbenoit/ ExitProgress 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.