Final Report: Natural Mercury Isotopes as Tracers of Sources, Cycling, and Deposition of Atmospheric MercuryEPA Grant Number: R830603
Title: Natural Mercury Isotopes as Tracers of Sources, Cycling, and Deposition of Atmospheric Mercury
Investigators: Odom, A. Leroy , Landing, William , Salters, Vincent
Institution: Florida State University
EPA Project Officer: Chung, Serena
Project Period: October 2, 2002 through December 31, 2006
Project Amount: $827,147
RFA: Mercury: Transport, Transportation, and Fate in the Atmosphere (2001) RFA Text | Recipients Lists
Research Category: Mercury , Air Quality and Air Toxics , Safer Chemicals , Air
This project’s goal was to demonstrate variations in the isotopic composition of mercury in nature and to utilize these variations as a new way of investigating natural and anthropogenic emissions of mercury into the atmosphere and of the atmospheric processes that affect transportation and deposition. We began with as plan to determine the isotopic composition of atmospheric mercury (1) thought to be dominated by far-field sources, (2) influenced by both far field and regional or local sources, and (3) influenced by known point sources. Early on in this project we encountered unanticipated technical difficulties that seriously delayed achieving all of the original goals. Along the way, however, came a serendipitous discovery that portends entirely new avenues to investigate sources and the atmospheric transportation and deposition of mercury and to impose stringent constraints on models. It is shown that the relative abundances of mercury isotopes in nature do vary not only because of differences in nuclear mass, but also because of differences in nuclear spin and nuclear volume.
Isotopic measurements were made with a Thermo-Finnigan Neptune inductively coupled plasma mass spectrometry (IPCMS) using a standard-sample-standard bracketing technique. Mercury isotope ratios of samples are expressed as permil deviations from values in the bracketing standard NIST SRM 3133, such as
δAHg = [(AHg/202Hg)sample/(AHg/202Hg)NIST3133] -1 x1000 o/oo (1)
where A represents the mercury 198Hg, 199Hg, 200Hg, 201Hg, 202Hg and 204Hg.
At the beginning we had anticipated finding small variations in mercury isotope ratios that were caused by their mass differences. Figure 1 shows an example of mass dependent fractionation in a sample, here a relative enrichment in the lighter isotopes.
Figure 1. Example of sample showing only mass dependent fractionation
There are theoretical and some experimental bases for suspecting that effects due to differences in nuclear volume (nuclear charge densities) (ref) and nuclear magnetic moments (ref) could produce a partial separation of odd neutron number isotopes from even-N number isotopes.
Figure 2: Example of sample showing mass independent fractionation and enrichment of odd-N isotopes.
It can be seen in Figure 2 that the degree of fractionation of those isotopes with an even neutron number is proportional to mass number. However the odd-N isotopes indicate some mass-independent effect. Figure 2 shows an enrichment of odd-N isotopes relative to the even-N isotopes. Many other samples exhibit a depletion of odd-N isotopes. Such isotope effects are important because they give significant clues to the chemical, physical chemical, and biochemical processes that produce them. Nuclear volume effects result from a change in the nuclear charge distribution with neutron number and the effect this has on the ground state energies of electrons. Isotope separation factors depend on the electron density at the nucleus, thus chemical bond type, and differences in nuclear volume (mean squared nuclear charge radii). Schauble has shown that the nuclear volume effect is a function of the oxidation state of the Hg chemical species: Hg0, and methylmercury, exhibit relative enrichment of the odd-N isotopes, and oxidized species (e.g., Hg+2) exhibit relative depletions. Of the stable mercury isotopes only 199Hg and 201Hg have non-zero values of the nuclear spin quantum number; these are 1/2 and 3/2 respectively. In reactions involving appropriately long-lived, radical-pair intermediates, triplet-singlet and singlet-triplet conversions can be influenced by nuclear-electron hyperfine coupling to the extent that a nucleus possessing a magnetic moment can influence kinetics and paths of recombination and therefore reaction rates. In the process it can achieve a degree of separation from isotopes with zero or lesser magnetic moments. This effect has been described as the magnetic isotope effect.
Figures 2 exhibits two features that are present in most of the samples analyzed. 1) The d values of the even-N number isotopes 198Hg, 200Hg, 202Hg (fixed at δ202Hg = 0), and 204Hg are proportional to the mass numbers. 2) The δ values of odd-N isotopes 199Hg and 201Hg plot either above or below the mass dependent fractionation (MDF) line and exhibit a mass-independent fractionation (MIF), Δ199Hg and Δ201Hg, where
Δ199HgMIF = δ199HgTOTAL – δ199HgMDF (2)
and Δ201HgMIF = δ201HgTOTAL – δ201HgMDF.
The values of Δ199Hg and Δ201Hg are measures of the degree of MIF. They can have either positive (indicating enrichment of odd-N isotopes) or negative (depletion) values. Values of Δ199Hg we have measured in samples range from – 1.19 (atmosphere deposited Hg) to + 3.15 (Hg in fish tissue). Ratios of Δ201Hg/Δ199Hg encountered range from approximately 0.55 to 1.43. Such a range in Δ201Hg/Δ199Hg ratios indicate that more than one mass-independent effects has been involved in the isotopic fractionation observed.
Fractionation of mercury isotopes due to only the magnetic isotope effect would produce an enrichment (or depletion) of the odd-N mercury isotopes in a Δ201Hg/Δ199Hg ratio of 1.11, the ratio of the nuclear magnetic moments. Figure 3 includes many of the atmospheric mercury samples that we have measured. The isotopic composition of all of these samples plot along a line passing through zero and having a slope essentially identical to the theoretical slope for isotope fractionation due to the magnetic isotope effect. It would seem therefore that the mercury has been involved in chemical reactions involving intermediate radical pairs. There are a few possibilities that cannot be uniquely identified at present but include photolytic reduction or reduction with organic radicals and evasion from oceans and lakes, or atmospheric oxidation of gaseous mercury by radicals. It is hoped that current research will reveal the mechanism and the extent of regional and global variations in atmospheric isotopic composition of mercury.
Figure 3: Examples of atmospheric mercury showing distinct magnetic isotope effect
The data point shown in red are representative of those samples of flue gas that we have measured form coal burning power plants. The coal samples have an isotopic composition of the flue gas and, not surprisingly similar to peat material Samples of ombrotrophic peat (not shown) also plot along the MIE line with Δ199Hg values between -0.2 and -0.6.
It was considered that if the odd-N depletion by the magnetic isotope effect observed in the atmospheric samples was due to the role of radicals produced by photo-reduction in waters and evasion of gaseous mercury, there might be an odd-N enriched reservoir found in the oceans and lakes. At the end of this project period we had collected seawater samples but not completed analyses. While not part of the proposed research, we analyzed fish tissue, in part because the high mercury concentrations made it quite easy. Figure 6 shows a Delta-Delta plot of these data.
On Figure 4 lines defining the Δ199Hg and Δ201Hg values produced by the effects of nuclear spin (magnetic isotope effect) and nuclear volume are drawn for reference.
Figure 4: Isotopic data for samples of fish tissue, includes marine and fresh water fish
All of data for methylated mercury in fish plot within analytical uncertainty either on the magnetic isotope effect (MIE) line or between that and the line representing the nuclear volume effect (NVE). Of the total Δ199Hg of 3.14 permil in the example illustrated, approximately half can be accounted for by nuclear-spin fractionation and half by nuclear-volume fractionation. It can be seen from Figure 6 that for most samples the MIE has had the more prevalent role in mass independent fraction.
There does exist natural variations in the isotopic composition of mercury in environmental materials. This was suspected, but not demonstrated at the beginning of this project. It has been demonstrated that these variations result from effects of nuclear mass, nuclear volume, and nuclear spin on reaction kinetics, chemical exchange reactions, and some physical processes. Nuclear volume effects depend on oxidation state and the nature of chemical bonds involved and vary with temperature. The magnetic isotope effect involves the hyperfine coupling of nuclear and electron spins and is manifested in reactions that involve intermediate products of free radical pairs. These can be chemically produced organic radicals or photolytically produced radicals.
The isotopic composition of mercury is complicated by mass, volume, and spin dependent fractionation. The absolute magnitude of each of these effects and their relative magnitudes as measured in a sample of environmental mercury provide information about the physical, chemical, and biochemical history of that history. The challenge ahead is for researchers to learn how to extract this information. As surely as the discovery of multiple isotope effects in oxygen and sulfur have led to new advances in planetary science, atmospheric chemistry, climatology, chemical physics, bioproductivity, and the evolution the Earth’s atmosphere and oceans, isotopic effects in mercury will provide new paths for understanding source, dispersion, and deposition of atmospheric mercury.
Based only on the information at hand, consider the following as one example. The relative abundance of 199Hg and 201Hg (after removal of mass dependent effects) deposited from the atmosphere (Figures 4 and 5) is dominated by the magnetic isotope effect. Assume that there are only two (obviously not correct) possible sources of atmospheric mercury 1) evasion from the oceans and two 2) emissions form coal burning. Emissions from coal burning that we have measure do indicate isotopic fractionation by the magnetic isotope effect, however the magnitude of this (~ 0.2 Δ199Hg) is much less than that observed for atmospheric samples (Δ199Hg = 0.5-1.2). If evasion of mercury from the ocean and other waters are a source of atmospheric mercury, the process should have left behind mercury with a magnetic isotope effect that is enriched in the odd-N isotopes to balance the depletion observed for atmospheric samples. Photolytic reduction of mercury in waters could produce radicals that can drive the magnetic isotope effect. Since at the end of the project period, we had not characterized water, we presently turn to fish tissue samples. If the mercury therein is representative of what is in the water column left, it does contain the odd-N isotope enrichment (Figure 6), perhaps the compliment to the odd-N isotope depletion of the atmosphere.
The significance of this research has been to reveal the power of isotopic investigations to provide new understanding of the sources, pathways and fate of atmospheric mercury and stringent new constraints on models. In this project we have only scratched the surface, but through the scratch, substantial future outcomes can be seen.