1998 Progress Report: Light Induced Mercury Volatilization from Substrate Mechanism(s) Responsible and In situ OccurrenceEPA Grant Number: R825249
Title: Light Induced Mercury Volatilization from Substrate Mechanism(s) Responsible and In situ Occurrence
Investigators: Gustin, Mae Sexauer
Institution: University of Nevada - Reno
EPA Project Officer: Chung, Serena
Project Period: December 15, 1996 through December 14, 1999 (Extended to December 14, 2000)
Project Period Covered by this Report: December 15, 1997 through December 14, 1998
Project Amount: $288,645
RFA: Exploratory Research - Air Chemistry & Physics (1996) RFA Text | Recipients Lists
Research Category: Engineering and Environmental Chemistry , Air
Objective:The objectives of this project are to: (1) determine the mechanisms responsible for light enhanced mercury volatilization from substrate using controlled laboratory experiments, and (2) investigate the effect of ambient light conditions on mercury flux in situ using micrometeorological and field chamber techniques.
Progress Summary:Speciation of Mercury in Soil. To determine the mechanism responsible for light enhancement of mercury emissions, the form of mercury in the soil must be assessed. This means determining if it is elemental mercury, loosely or strongly bound oxidized mercury, and/or mercury bound up in a mineral phase. As part of this project, multiple collaborations have been established to assess methods applied to determine the speciation of mercury in soils. The collaborators include: Harald Biester, Institute of Environmental Geochemisty, Heildelberg, Germany; Paul Lechler, Nevada Bureau of Mines and Geology, Reno, NV; and Chris Kim, a Ph.D. student of Jim Rytuba of the U.S. Geological Survey (USGS), Menlo Park, CA. We chose to utilize and modify sequential extraction procedures developed by other researchers, while the other groups mentioned above are applying a pyrolytic method for determination of mercury species in soils, a sequential extraction procedure, and synchrotron radiation based x-ray absorption spectroscopy, respectively. Samples used for this intercomparison include: synthetic samples created by adding 100 mg of HgS, HgCl2 , elemental mercury (Hgo) and HgO to 1 kg of autoclaved beach sand and pulverized quartz, and samples from the Carson River Superfund site, Clear Lake Superfund site, New Idria and Ivanhoe mining districts, and selected environmental samples from Germany.
In the process of testing previously published sequential extraction methods, we have found that some methods do not achieve good recoveries. For example, previous studies suggested the use of water to remove HgCl2. We achieved very poor recoveries with this method. Other researchers determined Hgo by heating the sample, assuming the mercury lost from the sample total was Hgo, and calculating Hgo based on concentration differences. We are finding, through the direct measurement of Hgo lost from the sample, that not all the Hgo is removed with heating. In addition, heating is affecting the other mercury species in the sample resulting in poor recoveries of the other mercury species amended to the soil as well. We currently are focusing on developing and testing alternative methods for sequentially extracting elemental mercury, HgCl2, and HgS from several soil types. We chose these species because it would allow us to assess which species is most important in the light enhancement mechanism, and it appears that enhancement is associated primarily with substrate containing Hgo and HgS. In addition, knowledge regarding the presence of these species will provide a means for assessing the environmental risk associated with a soil sample. For example, if the mercury species is elemental mercury, HgCl2 or HgS, this will affect whether it may be volatilized from a system, mobilized in solution, and/or remain fairly stable in the environment, respectively.
Harald Biester has completed his analysis of sample splits for the intercomparison, and some of these results of the pyrolytic method are published in the two manuscripts cited below. The pyrolytic method determines mercury species based on thermal decomposition and subsequent release of mercury at specific temperatures with continuous heating (Biester, et al., 1997). The method allows for a qualitative assessment as to whether there is Hgo, HgS, and HgO in substrate. However, there is overlap of mercury release curves for HgCl2 and organically bound species, and the efficiency of the substrate to adsorb mercury can influence release.
Laboratory Chamber Investigation of Emissions. Fluxes of mercury vapor from a variety of substrates were measured as a function of incident light using a laboratory gas exchange chamber. Field samples were obtained from four natural mercuriferous areas?the Steamboat Springs Geothermal Area, Reno, NV; the Clyde Forks mineralized fault zone, Ontario, Canada; a Proterozoic black shale obtained from a quarry just west of Thunder Bay, Ontario, Canada; the Ivanhoe Mercury Mining district, NV (from two Superfund sites?the Sulfur Bank Mercury Mine Superfund site, CA, and the Carson River Superfund site, NV). In addition, emissions as a function of incident light were also determined from the synthetic samples described above in the speciation discussion.
The laboratory chamber consists of a glass chamber constructed so that environmental variables impacting emissions are precisely controlled (c.f. Gustin, et al., 1997). Flux is determined with the following equation:
where F is the flux in ng/m2h, Co and Ci are the mercury concentration measured at the chamber inlet and outlet, respectively, A is the area from which emissions are being measured, and Q is the flow of air through the chamber. A light induced enhancement of emissions was determined for all field samples, except the Sulfur Bank Superfund sample and the Thunder Bay black shale sample. The Sulfur Bank sample was roasted cinnabar ore and although the mercury concentration of this sample was the same as the Carson River sample, the roasting process may have resulted in only strongly bound mercury being retained. The Thunder Bay shale sample had a low concentration of mercury (<1 ppm). It is hypothesized that the mercury in this sample was either tightly bound in the carbonaceous shale matrix or the dark black color of the rock resulted in most of the light energy being absorbed. Enhanced mercury emissions were observed from a weathered piece of cinnabar ore from the Ivanhoe Mercury mining district, but not from cinnabar ore that was unweathered. The most significant light enhancement of mercury emissions, normalized to total mercury concentration, occurred from samples from the Carson River Superfund site, the Clyde Forks site, and the Steamboat Geothermal Area. The percent enhancement of Hg flux, relative to that occurring in the dark at the same temperature, was >20 percent for those samples showing an effect. Almost two orders of magnitude enhancement was measured for the Carson River samples. Substrate amended with HgS and Hgo exhibited light enhanced emissions, while substrate amended with mercury oxide and mercury chloride did not. The presence of light enhanced emissions also were determined for mineral and organic soil horizons developed at the Clyde Forks site. There were light enhanced emissions for both the mineral and organic horizons with the exception of the low mercury concentration organic horizon sample. Slopes determined for plots of mercury flux from the same sample as a function of temperature in the light and dark are steepest for the light experiments. These data prove that incident light enhances emissions above that occurring in the dark. Data developed suggest that the enhancement of emission is not directly correlated with concentration, but is a function of the chemistry of the substrate.
In Situ Measurement of the Light Effect on Mercury Emissions. Dr. Robert Keislar, Desert Research Institute, is collaborating on this component of the project. Obtaining the micrometeorological system to be applied in this project was delayed until after a field intercomparison of methods used to measure mercury emissions was held in Reno, NV (Gustin, 1998; Gustin, et al., in review). As a result of this project, it was decided that the Modified Bowen Ratio method (MBR) would give the resolution we needed to assess the effect of light during the day-to-night transition. Reconnaissance of field sites was conducted in early summer and equipment was purchased. Collaboration with Frank Marsik, who is with Gerald Keeler's group at the University of Michigan, resulted in rapid set up of the equipment. Data were collected in August, September, and October using the micrometeorological system and field flux chambers. Sites visited included tailings in the Carson River Superfund site, NV; the Steamboat Springs Geothermal Area, NV; the New Idria Mercury Mining District; and the Ivanhoe Mercury Mining District. The Clear Lake Superfund was originally proposed as a field site; however, we decided to select other sites because laboratory chamber experiments have demonstrated that light does not enhance emissions from roasted ore from Clear Lake.
To determine the effect of light on the enhancement of Hg emissions flux, in situ field measurements were made during the dark to light transition in the morning and using high intensity halogen lamps during the night. Two field methods were applied?micrometeoro-logical methods and field flux chambers. The least invasive technique for measuring air-surface exchange of Hg is the application of a micrometeorological approach. Although flux chambers are useful for measuring Hg emissions from small surface areas (<0.1 m2), micrometeorological techniques integrate fluxes over much larger areas (~50 to 200 m2). The MBR technique involves the simultaneous measurement of both the flux and gradient of passive scalars (e.g., CO2, H2O) in the surface layer to derive an empirical turbulent transfer coefficient, Kz. With the advent of sensitive and accurate thermocouples (0.01oC), measurements of temperature gradients can be combined with measurements of water vapor gradients to obtain sensible (QH) and latent (QE) heat fluxes and the transfer coefficient Kz using the classic Bowen Ratio. The Kz determined from these approaches is applied to Hg gradients measured at the same heights as the other parameters to calculate the Hg flux. The relationship used to calculate the Hg flux is given by:
where FHg is the flux of Hg, FS is the flux of a passive scalar (e.g., CO2, H2O, heat), and CHg and CS are the concentration gradients of Hg and the passive scalar, respectively, measured at the same heights above the surface. To derive Kz, a variety of meteorological measurements are required: wind speed and direction; CO2/H2O concentrations; temperature gradient at two to four heights above the surface; Hg vapor concentration gradient at two to four heights; global and net radiation; soil temperature and soil heat flux; relative humidity; and barometric pressure. In addition, to derive Kz from the MBR, turbulent fluctuations of vertical velocity (w') and H2O concentrations (c'H2O) must be measured with fast response instruments (i.e., a sonic anemometer for w' and a krypton hygrometer for c'H2O). The covariance, w' c'H2O, provides the moisture flux in the MBR method, which is considered an improvement over using latent heat flux in the classic Bowen Ratio method because QE depends on the surface energy budget balance; the MBR method is a more direct way to measure moisture flux.
Two types of field chambers (FC) were applied during the field season. The type of FC used at the Carson River Superfund site was shared with us by Steve Lindberg, Oak Ridge National Laboratories (ORNL). It consisted of a Teflon chamber (surface area = 0.12 m2; volume = 24 L) supported by a metal frame. The one used at New Idria and Steamboat Springs was a circular chamber made of polycarbonate (UNR PCFC). The UNR PCFC covers a surface area of 0.038 m2 and has an internal volume of 5.7 L). The smaller volume of the UNR PCFC enables a shorter air residence time within the chamber or turnover time (~0.5 minutes) than the ORNL and UNR Teflon FC (~2.5 minutes) at a flow rate of 10 L/minute. Experiments conducted by ORNL indicated that a shorter turnover time provided more accurate flux measurements (Gustin, et al., in review). The inlet for the UNR PCFC consists of 4 holes (diameter = 1.25 cm) evenly distributed 2 cm from the bottom of the chamber with a single outlet located at the top of the chamber. Also, the UNR PCFC has a sharp bottom edge with no skirt, which allows the chamber to be pressed into the soil without covering the soil outside the chamber. This provides a less invasive approach to measuring fluxes.
Field Chamber Results. Field chamber measurements at the Carson River Superfund site reveal a 100-fold increase in mercury emissions at dawn as incident radiation first struck the tailings with no change in surface temperature (dark emissions = -4 ng/m2h and light emissions = 93 ng/m2h). At New Idria, emissions doubled at first light with no increase in soil surface temperatures (Figure 1). These almost real-time measurements of emissions were possible only with use of a Tekran mercury analyzer.
Micrometeorological Results. Figure 2 shows an example of light-enhanced mercury emissions from the Carson River Superfund site. Ambient mercury concentrations at two levels (25 cm and 175 cm) have pronounced peaks at first light, as defined by the rise in the global radiation curve. Note that soil temperature (displayed as x30 on the graph) is near its daily minimum, indicating that the peak in ambient concentration is not a temperature effect. The classic Bowen Ratio results were very noisy due to small and fluctuating moisture gradients and are not presented here.
Figure 3 shows the light enhancement of mercury emissions as the sun rose at the Steamboat Springs Geothermal Area. The soil temperature did not increase; however, mercury emissions increased with the inception of first light.
Figure 1. Effect of incident light on mercury emissions measured with a field flux chamber at the New Idria Mining District. There was no increase in soil surface temperature during the increase in memory flux.
Figure 2. Time series of mercury concentrations measured at 175 cm ("High") and 25 cm ("Low") at the Carson River Superfund site. Also shown are the global radiation and the soil temperature. Note that the highest concentrations are observed at first light yet the soil temperature is near minimum.
Figure 3. These figures show the increase in mercury flux measured with increasing global radiation but no change in soil temperature at the Steamboat Springs Geothermal Area, Nevada.
Accomplishments and Research Results. Significant research results are presented in the discussion above. Data developed thus far have led to the two working hypotheses on light enhanced emissions:
from the soil surface due to incident light energy.
Kothny (1991) and Gustin, et al. (1997) have demonstrated that cinnabar is emitting Hgo to the air (c.f. Kothny, 1991; Gustin, et al., 1997). Wood (1974) suggested that microorganisms in soil might play an important role in breaking down cinnabar in soils. Cinnabar is the primary mercury mineral found in enriched substrate. If it is constantly breaking down to form Hgo, this mercury would migrate to the soil surface where it would adsorb to electrostatically charged mineral grains as well as to organic matter. Incident light on the soil surface could cause Hgo to vaporize and move away from the soil surface just like the evaporation of water. This surface layer would constantly be recharged by mercury vapor moving upward towards the soil air interface. An alternative hypothesis is:
of photo reactive species.
These would include such things as iron oxide or sulfur-hydrogen complexes, both of which have been suggested to be photoreactive (Stromberg, et al., 1991; Lin and Pehkonen, 1998).
Not only is this project developing data on the mechanisms responsible for the light enhancement of mercury emissions from substrate, but the data developed in this project are helping to build our database on the emissions of mercury from diffuse sources and furthering our understanding of parameters controlling Hg fluxes. Currently, there is not a good database from which we may estimate Hg fluxes from these areas. Constraining emissions from these nonpoint sources is a critical need if we are to develop realistic regulatory controls. Data from this project also are being used to develop collaborations with other scientists and other research projects. Two proposals will be submitted in response to the Mercury in Watersheds Science to Achieve Results (STAR) grant solicitation, and both of them draw upon data developed in this project and utilize equipment purchased in this project.
Biester H, Scholz C. Determination of mercury binding forms in contaminated soils: mercury pyrolysis versus sequential extractions. Environmental Science and Technology 1997;31:233-239.
Gustin MS, Taylor GE, Maxey RA. Effect of temperature and air movement on the flux of elemental mercury from substrate to the atmosphere. Journal of Geophysical Research 1997;102:3891-3898.
Gustin MS. Nevada mercury emissions project: mercury flux measurements?an intercomparison and assessment. Electric Power Research Institute, 1998 (published report).
Gustin MS, Lindberg S, Marsik F, Casimir A, Ebinghaus R, Edwards G, Fitzgerald C, Kemp J, Kock H, Leonard T, London J, Majewski M, Owens J, Piolote M, Poissant L, Rasmussen P, Schaledlich F, Schneeberger D, Schroeder W, Sommary J, Turner R, Vette A, Wallschlaeger D, Ziao X. The Nevada storms mercury flux methods intercomparison. Journal of Geophysical Research of Atmospheres 1998 (in review).
Kothny EL. Three phase equilibrium of mercury in nature, in trace metals in the environment. Symposium Volume for the Water, Air, and Waste Chemistry Meeting of the American Chemical Society, Washington, DC, September 15, 1971.
Lin C, Pehkonen SO. Aqueous free radical chemistry of mercury in the presence of iron oxides and ambient aerosol. Atmospheric Environment 1997;31:4125-4137.
Stromberg D, Stromberg A, Wahlgren U. Relativistic quantum calculations on some mercury sulfide molecules. Water, Air, and Soil Pollution 1991;56:681-695.
Wood JM. Biological cycles for toxic elements in the environment. Science 1974;183:1049-1052.