|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 COUPLED BIOGEOCHEMICAL CYCLING OF MERCURY AND IRON: IMPLICATIONS FOR MERCURY REMOVAL FROM AQUEOUS EFFLUENTS AND BIOTRANSFORMATION IN SEDIMENTARY ENVIRONMENTS By JULIANNE D. VERNON A DISSERTATION PRESENTED TO THE GRADUATE SC HOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2010
2 2010 Julianne Vernon
3 To my loving husband
4 ACKNOWLEDGMENTS I thank my advisor, Dr. JeanC laude J. Bonzongo, for his guidance, support, and humor that has helped me through my journey at the University of Florida. I thank my commi ttee members, Dr. Angela S. Linder Dr. Joseph J. D elfino, and Dr. Roy Rhue, for their valuable expertise, advice, and support. My special thanks go to NSF Graduate Assistance in Areas of National Needs, HDR Engineering Inc., Progress Energy Foundation and the UF Graduate Minority Office for supporting me in my research. I thank all members of my research group for t heir help and support Finally, special thanks go to my family and friends. I greatly appreciate their love, support, and encouragement. Lastly I thank m y husband for his encouragement, patience, and support.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................. 4 LIST OF TABLES ............................................................................................................ 8 LIST OF FIGURES .......................................................................................................... 9 LI ST OF ABBREVIATIONS ........................................................................................... 12 ABSTRACT ................................................................................................................... 14 CHAPTER 1 INTRODUCTION .................................................................................................... 17 2 M ERCURY IN AQUATIC SYSTEMS AND CURRENT REMEDIATION TECHNIQUES ........................................................................................................ 22 Mercury ................................................................................................................... 22 Mercury in Aquatic Systems ................................................................................... 22 Health Impacts ........................................................................................................ 23 Common Techniques Used in Remediation of Mercury Contaminated Aqueous Effluents ............................................................................................................... 25 Phytoremediation ............................................................................................. 25 Microbial Bioremediation .................................................................................. 26 Constructed Wetlands ...................................................................................... 27 Sorption Materials ............................................................................................ 29 Activated carbon adsorption ....................................................................... 29 Ion exchange resins ................................................................................... 29 Other low cost adsorbents ......................................................................... 30 Zero Valent Iron (ZVI) ....................................................................................... 31 Zero valent iron and volatilization of dissolved mercury through reduction to elemental mercury ............................................................... 33 Adsorption on iron oxyhydroxides .............................................................. 34 Precipitation of dissolved merc ury as sulfide complexes ........................... 35 Effect of pH, competitive ions, and mercury binding ligands ...................... 36 Nano Zero Valent Iron (nZVI) ........................................................................... 36 nZVI and nitrate interaction ........................................................................ 39 nZVI and arsenic interaction ...................................................................... 40 nZVI and chromium interaction .................................................................. 41 3 INVESTIGATION OF MERCURY AND IRON INTERACTIONS IN AQUEOUS SYSTEMS: IMPLICATIONS FOR WATER REMEDIATION ................................... 4 3 Introduction ............................................................................................................. 43
6 Materials and Methods ............................................................................................ 45 Iron Particle Characterization and Chemistry of Water Used in Laboratory Experim ents .................................................................................................. 45 Corrosion of Iron Particles and Temporal Changes in Specific Surface Areas: ............................................................................................................ 46 Effects of Water Chemistry ............................................................................... 46 Volatilization of Dissolved Mercury by Iron Particles in Closed Batch Reactors ........................................................................................................ 47 Effect of Dissolved Natural Organic Matter on Mercury Vol atilization ............... 48 Hg Sorption onto Zero Valent Iron and Nano Zero Valent Iron Particles .......... 50 Statistical Analysis ............................................................................................ 51 Results and Discussion ........................................................................................... 51 Particle Characterization and Water Chemistry ................................................ 51 Kinet ic Change of Specific Surface Area of Zero Valent Iron and Nano Zero Valent Iron in Deionized Water and Wastewater ........................................... 53 Mercury Volatilization by Metallic Iron Particles: Effect of Particle Size ............ 55 Effect of Dissolved Natural Organic Carbon on Volatilization of Aqueous Mercury ......................................................................................................... 60 Mercury Removal from Solution by Zero Val ent Iron and Nano Zero Valent Iron through Sorption .................................................................................... 65 Conclusion .............................................................................................................. 68 4 REMOVAL OF MERCURY FROM WASTEWATER EFFLUENT USING FLOW TH ROUGH COLUMNS PACKED WITH NANO ZERO VALENT IRON PARTICLES ............................................................................................................ 71 Introduction ............................................................................................................. 71 Materials and Methods ............................................................................................ 74 Particle Characterization .................................................................................. 74 Column Preparation and Setup ........................................................................ 74 Water Used in Column Studies and Experimental Protocol .............................. 75 Optimization of Mercury Removal in Columns Containing Nano Zero Valent Iron: Effects of Particle Mass, Flow Rate, and Water Chemical Composition .................................................................................................. 77 Statistical Analysis ............................................................................................ 78 Results and Discussion ........................................................................................... 79 Effects of Iron Particle Size on Mercury Removal from Wastewater Effluent ... 79 Effect of Nano Zero Valent Iron Mass on Mercury Removal ............................. 81 Effect of Flow Rate on Mercury Removal ......................................................... 82 Hydrogen Peroxide Treatment of the Influent and Oxidation of Dissolved Organic Compounds to Improve Mercury Removal from Solution ................ 84 Mercury Removal Using Column in Series and Effect of Zinc and Cadmium as Competitive Cations ................................................................................. 91 Scanning Electron Microscopy Characterization of the Reactive Media ........... 94 Conclusion .............................................................................................................. 98
7 5 IRON MERCURY INTERACTIONS: EFFECTS ON MERCURY BIOAVAILABILITY AND METHYLMERCURY PRODUCTION IN FRESHWATER SEDIMENTS ............................................................................... 100 Introduction ........................................................................................................... 100 Materials and Methods .......................................................................................... 103 Sediment and Water Samples Used in Hg Methylation Studies ..................... 103 Determination of Total and Methyl mercury .................................................. 104 Speciation of Solid Phase Mercur y in Collected Sediment Samples .............. 105 Investigation of the Effect of Changing Sulfate Concentrations on the Biotransformation of Mercury in Sediment Slurries Containing Fixed Iron Masses ........................................................................................................ 106 Investigation of the Effect of Changing Iron Masses on the Biotransformation of Mercury in Sediment Slurries Containing Sulfate Concentrations ............................................................................................ 108 Effect of Iron on the Biotransformation of Newly Added Hg into Sediments ... 108 Statistical Analysis .......................................................................................... 109 Results .................................................................................................................. 109 Discussion ............................................................................................................ 115 Speciation of Mercury and its Methylation in Aquatic Sediments ................... 115 Effect of Specific Experimental Parameters ................................................... 117 Sulfate ...................................................................................................... 117 Organic carbon ........................................................................................ 118 Effect of iron addition to sediment slurries ............................................... 119 Methylation and Demethylation ...................................................................... 119 Conclusion ............................................................................................................ 120 6 CONCLUSION AND RECOMMENDATIONS ....................................................... 121 APPENDIX A STATISTICAL ANALYSIS DATA .......................................................................... 124 LIST OF REFERENCES ............................................................................................. 125 BIOGRAPHICAL SKETCH .......................................................................................... 142
8 LIST OF TABLES Table page 2 1 Site remediation using nano zero valent iron: List of recent projects (Li et al. 2006b) ................................................................................................................ 37 3 1 Chemical composition of water samples collected along the Suwannee River from headwaters to the river delta ...................................................................... 49 3 2 Chemical composition of used wastewater effluent. Major ions were determined by ion chromatography.. .................................................................. 53 3 3 Hg loading capacity of ZVI and nZVI for Hg spiked wastewater effluent at pH 5.3 and 8.3. ......................................................................................................... 66 3 4 Pseudofirst order adsorption rate constants for Hg on ZVI and n ZVI based on Hg spiked deionized water and wastewater effluent ...................................... 67 4 1 Chemical composition of wastewater effluent used in column studies ............... 76 4 2 Ionic radii and electronegativities (Krauskopf and Bird 1995). ............................ 93 5 1 Selective extraction fractions of mercury. Adapted from Bloom et al., 2003 ..... 106 5 2 Experimental Design for Hg Methylation in Sediment slurries containing Fixed Iron Amounts and Changing Sulfate Concentrations .............................. 107 5 3 Experimental Design for Studies on Hg Methylation in Sediment Slurries with Fixed Sulfate Concentrations and Changing Iron Masses ................................ 108 5 4 Experimental Design for Hg Methylation Studies in Sediment Slurries Spiked with Hg(NO3)2 in Addition to Iron and Sulfate ................................................. 108 5 5 Average metal concentrations found in sediment and site water determined by ICP AES. ..................................................................................................... 110 5 6 Average major ions found in sediment (water soluble fraction) and used natural water, analyzed by ion chromatography (ND = not determined). .......... 110
9 LIST OF FIGURES Figure page 1 1 Interaction of zero valent iron with pollutants in aqueous solutions. Me=metal, RCl=chlorinated organic pollutant (Li et al. 2006b). ............................................ 19 2 1 Schematic illustration of how nanoparticles could be used for in situ remediation of polluted ground waters (Zhang 2003). ........................................ 39 3 1 Trends of dissolved organic carbon (DOC) (adapted from Gao et al., 2009). ..... 49 3 2 Trends of ionic strength (I) in waters of the Suwannee River (adapted from Gao et al., 2009) ................................................................................................. 50 3 2 Effect of pH on zeta po tential of nanozero valent iron (nZVI) p articles suspended in DI water and Wastewater Effluent .. .............................................. 52 3 3 Effect of Water Matrix on Iron Surface Area. ...................................................... 54 3 4 Temporal trends of Hg volatilization and sorption in DI water with a concentration of 1.0 g Hg/mL and 0.04 g of zero valent iron (ZVI) particles. .... 57 3 5 Simplifie d schematic representation of Hg interaction with metallic and corroded iron particles in aqueous solutions along a corrosion gradient. ........... 58 3 6 Temporal trends of Hg volatilization and sorption in wastewater effluent with a concentration of 1 g Hg/mL and 0.04 g of iron particles. ............................... 60 3 7 Effect of natural organic carbon (SR1) on Hg volatilization (300 g/L) ............... 61 3 8 Quin ones: hydroquinone (QH2), fully oxidized quinone (Q), semiquinone radical ( QH ). ..................................................................................................... 62 3 9 Effect of dissolved natural organic carbon on Hg reduction/volatilization ........... 63 3 10 Effect of Hg concentration on its volatilization from SR1 waters ......................... 64 3 11 Effect of initial Hg concentration dissolved in either Deionized water (DI water) or wastewater effluent (WW) on Hg adsorption by ZVI ............................ 68 4 1 Picture of the experimental setup used in described column studies. ................ 76 4 2 Comparison of adsorption profiles of Hg onto ZVI and nZVI. ............................. 79 4 3 Trends of Hg removal from wastewater influent containing150 g Hg/L in columns packed with nZVI at 0.1% and 1% per mass basis. ............................. 81
10 4 4 Effect of flow rates on adsorption profile of nZVI. Influents are Hg spiked wastewater effluent (150 g Hg/L). ..................................................................... 83 4 5 Cross sectional representation of the columns reactive media .......................... 84 4 6 Effect of influent pretreatment on adsorption profile of n Z VI. Influent for column was Hg spiked wastewater effluent (150 g Hg/L). ................................ 85 4 7 Effect of hydrogen peroxide (H2O2) treatment on dissolved organic matter present in wastewater used as influent in column studies. ................................. 86 4 8 Organic constituents present in wastewater effluents ........................................ 87 4 9 Excitation emission peaks of dominant soluble organic matter compounds based on a literature review publ ished by Chen et al (2003) ............................. 87 4 10 Excitation emission spectra to determine dominant organic species present. (A) column influent untreated, (B) column influent treated with 1% hydrogen perox ide, and (C) column effluent. ...................................................................... 89 4 11 Volatilization of Hg in wastewater effluent (WW) and WW treated with 1% hydrogen peroxide (H2O2) with an incubation time of 20 days ........................... 91 4 12 Mercury removal from wastewater (WW) effluent in a single column and two columns used in series. ...................................................................................... 92 4 13 Effect of cadmium (Cd) and zinc (Zn) on the removal of Hg from a wastewater used as influent in a study using two 0.1% n ZVI packed columns in series. ............................................................................................................. 94 4 14 (A) SEM image of pure nZVI at 600x. (B) EDS spectrum of pure nZVI. The presence of a large oxygen peak is highlighted by red box. ............................... 95 4 15 SEM image of the columns reactive media in series. (A) column 1 resolution of 400x, (B) column 2 resolution of 2500x .......................................................... 96 4 16 SEM image of the columns reactive media in series (C) red box zoomed in of Figure A at a resolution of 4000x. ................................................................... 97 4 17 Energy dispersive spectra analysis of column 1 (A) and column 2 (B) in the column series ..................................................................................................... 98 5 1 Map of East Fork Poplar Creek in Big Turtl e Park Greenway in Oakridge,TN. 104 5 2 Methyl mercury produced in Hg contaminated sediments with sulfate addition after 25 days of incubation. .............................................................................. 111
11 5 3 Effect of particle size with a sulf ate concentration of 500 M treatment added to sediments. .................................................................................................... 112 5 4 Effect of different iron masses on MeHg production in sediment slurries containing 1 mM of sulfate. ............................................................................... 113 5 5 Methyl Hg produced in sediment slurries spiked and nonspiked with Hg ........ 114 5 6 Percentage of Hg associated with different sediment fractions based on sequential selective extractions using a method adapted from Bloom et al. (2003). .............................................................................................................. 116 5 7 Sulfate concentration range for optimal mercury methylation rates in sediments (Gilmour and Henry 1991) ............................................................... 117 6 1 Hg iron interactions investigated in this study. ................................................. 121
12 LIST OF ABBREVIATION S AC Activated carbon As Arsenic BET Braunauer Emmett Teller CMC Carboxyl methyl cellulose Cr Chro mium DI Deionized water DOC Dissolved organic carbon Hg Mercury HgS Mercury sulfide IRB Iron reducing bacteria MeHg Methyl mercury NM Nano materials nZVI Nano zero valent iron OM Organic matter PRB Permeable reactive barriers SR1 Suwannee R iver 1 SR2 Suwannee R iver 2 SR3 Suwannee R iver 3 SRB Sulfate reducing bacteria SSA Specific surface area THg Total mercury US EPA Unites States Environmental Protection Agency WW Wastewater effluent VOC Volatile Organic Carbon
13 ZVI Zero Valent Iron
14 Abstract of Dissertatio n Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy COUPLED BIOGEOCHEMICAL CYCLING OF MERCURY AND IRON: IMPLICATIONS FOR MERCURY REMOVAL FROM AQUEOUS EFFLUENTS AND BIOTRANSFORMATION IN SEDIMENTARY ENVIRONMENTS B y Julianne Vernon December 2010 Chair: Jean Claude J. Bonzongo Major: Environmental Engineering The toxic effects of mercury ( Hg ) on the environment and human health have led the US reg ulatory agencies to set stringent guidelines for Hg levels in gaseous and aqueous waste effluents. With regard to Hg contaminated waters which is the focus of this proposed research, Hg levels in the concentration range of 1215 ng/ L, p arts p er t rillion ( ppt ) are targeted for wastewat er effluents. Several treatment methods for Hg remediation of contaminat ed waters and sediments/soils exist. Unfortunately, the abovem entioned low ppt levels set forth by federal and state agencies remain out of reach by any of the treatment technologies that are currently available. In addition to these remediation issues, the conversion of inorganic Hg into methyl mercury (MeHg) compounds, primarily in anoxic aquatic compartments has led to c oncerns over the cycling o f Hg in the environment due to the toxicity of MeHg as it readily accumulates in living tissues and biomagnifies in food chains. In this study, interactions between metallic iron and dissolved aqueous Hg were investigated with the ultimate goal to evaluate the potential for use of zerovalent iron
15 (ZVI) in the treatment of Hg contaminated effluents on one hand, and for controlling MeHg formation in sediments on the other. First, the increase in surface area and the decrease in kinetic reaction time fav ored by nanosize particles could allow the removal of Hg from aqueous effluents through the combination of (i) ionic Hg (Hg2+) reduction to elemental Hg and volatilization, and (ii) Hg adsorption onto the oxyhydroxide coating that develops over time around the iro n particles. To verify the above hypothesis, laboratory batch and column experiments were conducted to investigate the interaction between either bulk ZVI or nanosized ZVI ( nZVI ) and Hg dissolved in aqueous solutions ; and the resulting efficiency for Hg r emoval The results of these investigations can be summarized as follows: nZVI particles have a much higher maximum adsorption capacity and a faster adsorption rate than ZVI The removal of dissolved Hg under dynamic conditions using column studies show tha t flow rate, Hg speciat ion, and mass of iron particles used play a significant role in the efficiency of Hg removal by iron particles. Further studies are needed to obtain conditions that favor the removal of Hg down to ppt level. Second, the methylation of Hg in aquatic sediments depends on the interaction of a wide variety of indigenous microorganisms and the intimate coupling of such actions to ke y geochemical factors such as Hg speciation and availability temperature, pH, the quantity and types of org anic compounds available. However, previous research points to dominant microbial processes and Hg speciation and availability as the main criteria for Hg methylation, with microbial sulfate reduction being the most important microbial catalyzed geochemical reaction linked to Hg biomethylation. Accordingly, the presence of metallic iron as a source of dissolved Fe(I I) under anoxic conditions in a sulfate rich
16 anoxic sediments would lead to Hg sequestration through coprecipitation with iron sulfide mineral s, and therefore, a reduced fraction of bioavailable Hg. Microcosm studies using anoxic sediment slurries were conducted by varying the Fe to sulfate ratios in Hg rich sediments in the absence (controls) or presence (treatments) of metallic iron particles The main findings can be summarized as follows: The relationship between the different solid phase Hg fractions (easily exchangeable, humic acid soluble, organo chelated, elemental Hg, and mercuric sulfide) and MeHg production show that Hg speciation plays a significant role in controlling Hg methylation. MeHg production i s impacted by two main mechanisms, (i) adsorption of inorganic Hg on oxyh y droxide layers and/or (ii) coprecipitation of inorganic Hg iron sulfide complexes. This reduces the availabil ity of ionic Hg and its subsequent methylation. Overall, the results obtained from th is study establish the ability of ZVI and nZVI to remove Hg from aqueous effluents and interfere with ionic Hg methylation in sediments. Therefore, the obtained results c onstitute a foundation for further resear ch on validating the use of nZVI in the r emediation of Hg contaminated systems.
17 CHAPTER 1 INTRODUCTION In a paper entitled A Silent Epidemic of Environmental Metal Poisoning published over two decades ago, Nriagu (1988) rang the bell on potential health effects of increasing levels of heavy metals introduced into natural systems by anthropogenic activities. The outcry was driven primarily by the increasi ng pressure on natural resources and waste generation associated with the exponential growth of human populations. Since, a large numbers of papers focusing on the investigation of both the environmental fate and impacts of heavy metals and the remediation of metal contaminated systems have been published (e.g. see review by (Cundy et al. 2008; Mulligan et al. 200 1; Noubactep 2008) Heavy metals are introduced to aquatic systems from both anthropogenic and natural sources. However, unlike organic pollutants, metals released to the environment are not biodegradable and their persistence, transformation, and transf er to the food chain lead to negative effects on living organisms. Although much is now known on the biogeochemistry of several trace metals including mercury (Hg) in water and soil/sediment, research on the development of cost effective and environmental l y friendly remediation techniques remains challenging. Current methods for remediation of metal contaminated liquid and solid matrices include physical separation, thermal processes, biological decontamination, phytoremediation, electrokinetics, washing, stabilization, and solidification techniques. Unfortunately, only a few of these techniques have been tested commercially and their use remains limited due to factors such as their prohibitive costs and more recently, their inability to meet some of the n ew limits imposed by federal and state regulatory agencies. Overall, the remediation of metal -
18 contaminated environments remains one of the most intractable problems of environmental restoration and this is both a national and international issue. It also r equires the development of remedial approaches that remove or immobilize metals while avoiding adverse effects on treated systems. Concerns over the cycling of Hg in the environment have been driven primarily by the toxicity of methyl Hg (MeHg) as it accum ulates in living tissues and biomagnifies in food chains. Since anthropogenic activities (e.g. combustion of fossil fuels, incineration of wastes, industrial activities, gold mining by Hg amalgamation techniques, etc) constitute the primary cause of anomal ously high Hg levels found in aquatic systems. T he recently developed Clean Air Mercury Rule requires control of Hg emissions by electric power generators by 2020 which would have an estimated annual cost of about $1 billion annually by 2020 (US.EPA 2005) However, despite the above projected high costs, the benefits of the proposed preventive/remedial measures can not be fully guaranteed. The reason is that the link between atmospheric emission of Hg and the bioaccumulation of MeHg in aquatic biota involves a complex series of biogeochemical processes This includes the poorly studied response of soil/sediment storedHg within watersheds to land use changes, the delivery of inorganic Hg species to areas that act as primary loci for Hg methylation, and the presence of both biological and geochemical parameters necessary for MeHg production and accumulation. In addition, despite t he overwhelming literature on Hg in aquatic systems, several gaps exist in our current knowledge of the different processes involved in insitu transformation of inorganic Hg species to MeHg. Due to sitespecific changes in both geochemical parameters and the microbial community composition, the assumption of sulfur reducing bacteria (SRB) as
19 major Hg methylators in all aquatic systems could lead to erroneous model predictions of Hg bioaccumulation in fish and risks for human exposure. Therefore, an approac h that takes into account all key environmental parameters in a simultaneous manner and focuses on Hg naturally present in the sediment/soil is needed. The first objective of this proposed research program is to comparatively investigate the potential use of zero valent iron (Z VI) and nanozero valent iron (nZVI) as sorbents for Hg removal from aqueous effluents and investigate the mechanisms involved in Hg immobilization. The interaction of ZVI with other metals and organic pollutants can involve several m echanisms including (i) reduction, (ii) sorption on oxidized iron surfaces and (iii) coupled reduction and sorption as seen in Figure 11 (Cundy et al. 2008; Li et al. 2006a; Mulligan et al. 2001; Noubactep 2008) Figure 11. Interaction of z ero valent iron with pollutants in aqueous solutions. Me=metal, RCl=chlorinated organic pollutant (Li et al. 2006b) It is hypothesized that: t he increase in surface area and the decrease in kinetic reaction time favored by nanosized iron particles will allow a faster removal of Hg from
20 aqueous effluents through the combination of (i) ionic Hg reduction to elemental Hg and volatilization, and (ii) Hgn+ adsorption onto the oxyhydroxide coating layer that develops over time around the solid particles. This study investigated the efficiency of Hg r emoval by both ZVI and nanoZVI and assessed their potentials to bring Hg levels from contaminated effluents to levels that are targeted by regulation agencies such as the EPAs 12 to 15 ng/L range In natural aquatic compartments such as sediments, the pr esence of solid iron particles could, depending on redox conditions, impact the cycling of metal and organic pollutants as well. The refore, besides the above described three types of interactions that could lead to the remediation of Hg contaminated efflue nts the secon d objective of this study was to determine the e ffect of iron on Hg availability and methylation by sediment microorganisms. In contrast to the common practice that consists of spiking sediments with inorganic or organic Hg compounds to deter mine potential rates of Hg methylation and MeHg degradation, this study investigated the effect of iron particle addition to sediments historically contaminated with Hg (i.e., not spiked with Hg in the laboratory) in comparison with sediments from the sour ce but spiked with Hg salt. This approach used here intends to determine the fate of old versus newly added Hg in sediments containing iron particles. It was hypothesized that the addition of iron particles in Hg contaminated sediments would limit Hg bioavailability through a combination of its sorption onto oxyhydroxide layers and coprecipitation with iron sulfide species, depending on redox conditions The methylation of Hg in aquatic sediments depends on the activity of a wide variety of indigenous microorganisms primarily the sulfate reducing bacteria (SRB), and key geochemical factors such
21 availability of specific terminal electron acceptors (TEAs), the quantity/quality of organic substrates, temperature, pH and Hg speciation. However, the availability of Hg to methylating agents constitutes the primary limitation for methylmercury (MeHg) production, even in Hg contaminated systems. Accordingly, iron particles could depress MeHg production in sediments. In this dissertation, a review of relevant literature on Hg in the environment its environmental and health implications, and different t echniques that are used for the remediation of Hg contaminated are discussed in Chapter 2. This review is then followed by chapters investigating the interactions of iron particles with aqueous Hg (Chapter 3), as well as the potential of such particles to remove Hg from wastewater effluents using column studies (Chapter 4). The potential impact of iron particles on Hg present in sediments with regard to its bioav ailability and transformation to MeHg is presented in Chapter 5. Finally, Chapter 6 gives a summary of the major contributions of this research effort to current knowledge on the biogeochemistry of Hg and its interaction with Fe.
22 CHAPTER 2 MERCURY IN AQUA TIC SYSTEMS AND CURRENT REMEDIAT ION TECHNIQUES Mercury Mercury (Hg) introduced into the environment by most anthropogenic activities are primarily in the inorganic form (e.g., Hg0 and Hgn+). When released into the atm osphere, Hg can be subject to long rang e transport leading to the contamination of remote and often pristine environments through both dry and wet precipitations onto terrestrial and aquatic systems. In addition to Hg introduced directly into aquatic systems, Hg deposited on soils can later reach aquatic systems through infiltration (contamination of groundwater) and/or via surface runoff (contamination of rivers, lakes, and coastal waters). Mercury in Aquatic Systems I norganic Hg undergoes a number of biochemical and physical transformations in aquatic systems For instance, the oxidation of metallic mercury (Hg0) produces ionic species which are then methylated by sedimentary microorganisms (e.g., (Bonzongo et al. 1996; Compeau and Bartha 1985; Gilmour and Henry 1992) Methyl mercury (MeHg ) that accumulates in the environment and in living organisms comes primarily from the net balance of the concurrent processes of microbial inorganic Hg methylation and demethylation of produced MeHg (US.EPA 1997) Indeed, nearly all Hg found in biological tissues tend to be present as MeHg (Bloom 1992; Kim 1995; Watras et al. 1995) Unfortunately, exposure to MeHg and other Hg species lead to adverse effects on living organisms including a reduced reproductive capacity, an i mpaired growth and development, behavioral abnormalities and death. These impacts of Hg on human health depend on several pathways related to toxicokinetic mechanisms of its major
23 chemical forms present including Hg0, inorganic Hg salts (e.g., HgCl2) and organic Hg compounds such as MeHg (W.H.O. 1990; W.H.O. 1991) in different environmental m edia. Depending on the chemical form of Hg, the combination of these toxicokinetic mechanisms (absorption, distribution, metabolism and excretion) will determine the risk associated with human exposure to Hg and its compounds. For most of the abovementioned adverse ef fects, the process of Hg methylation appears to be an important part of the contamination and response processes (US.EPA 1997) Health Impacts MeHg is classified as a possible human carcinogen. It is rapidly and extensively absorbed through the gastrointestinal tract (Aberg et al. 1969; Boffetta et al. 1998; Boffetta et al. 1993; US.EPA 1997; W.H.O. 1990) Epidemics of Hg poisoning following exposure to MeHg in Japan and Iraq hav e demonstrated that neurotoxicity is the health effect of greatest concern when a developing fetus is exposed to MeHg (US.EPA 1997) The disaster in Minamata, Japan, where mass poisonings involving Hg attracted the attention of the world scientific community in late 1950s is still fresh in memories. The inhabitants of fishing villages along Minamata Bay suffered an epidemic of neurological disorders, visual constriction, brain damage, impairment of speech and hearing, numbness of extremities, impairment of gait, and sever al death cases The above aforementioned epidemic was afterward all attributed to Hg poisoning (now known as Minamata disease) due to fish consumption from the bay. In the past two decades, there have been reports of symptoms of the Minamata disease in Brazil, due to the use of Hg0 in artisanal gold mining (AGM) (Hylander et al. 2006) In this case, miners who burn gold (Au) Hg amalgams show signs of mercurialism due primarily to the inhalation of Hg vapor released during AuHg
24 amalgam burning In addition fish eating people living within the mining impacted areas show high Hg concentrations in blood and other tissues (Hinton et al. 2003; Hylander et al. 2006) Hg0 in the lungs oxidizes to ionic Hg and forms complexes which are quite soluble in body fluids and lipids, allowing for rapid diffusion through cell membranes and reaching vital tissues such as those of t he brain. A chronic exposure to Hg vapor results in symptoms like depression, exaggerated emotional response, gingivitis, muscular tremors and ultimately death, while acute exposure produces dysfunction of kidneys and urinary tract, vomiting, and potential ly death too (Hinton et al. 2003; Valenzuela and Ftyas 2002) High levels of Hg are also detected in the blood and other tissues of indigenous people from the Ar ctic Region. Here, atmospheric deposition of Hg transported fro m lower latitude industrialized regions is the main source of Hg in both terrestrial and aquatic syst ems. In fact, the Ar ctic is a recognized sink of Hg released to the atmosphere in lower latitude and warmer regions (Boening 2000; Givelet et al. 2004; Steffen et al. 2008; Ullrich et al. 2001) Diet, primarily the consumption of fish and other aquatic organisms, is the main exposure pathway for the indigenous Arctic people. Over time, the accumulation of Hg in the Arctic has resulted in incre ased average Hg levels in the c ord blood of newborns and widespread child development problems (e.g. speech, walking, and loss of IQ) (Hylander et al. 2006) The above listed effects of Hg on humans have led the US regulatory agenci es to set stringent guidelines for Hg levels in waste gaseous and aqueous effluents. With regard to Hg contaminated waters which is the focus of this proposed research, Hg
25 levels in concentration range of 1215 ng/L are targeted for wastewater effluents (Hanlon 2007) Common Techniques Used in Remediation of Mercury Contaminated Aqueous Effluents Several t reatment methods for Hg removal from water and Hg removal from and immobilization in sediments/soils exist (Atwood and Zaman 2006; Wang et al. 2004) These treatments include phytoremediation, microbial remediation, constructed wetlands and sorption materials. A brief description is given below. Phytoremediation It is d efined as the use of living g reen plants to remove pollutants from environmental compartments or to render them harmless. Phytoremediation has emerged as a promising, cost effective and environmental ly friendly alternative to most engineering based remediation techniques. This technology can be applied to either organic and inorganic pollutants present in soil, sludge, sediment, water, or t he air. Phytoremediation is divided into the following five subcategories: (i) p hytoextraction, which uses hyper accumulators to transport and concentrate contaminants from contaminated media to the aboveground parts for subsequent harvesting and removal from the site (Adams et al. 2000; Schnoor 1997) ; (ii) p hytofiltration relies on plant root s or seedlings to sorb contaminants; (iii) phytostabilization uses plants to reduce the mobility and bioavailability of contami nants in the environment; (iv) phytovolatilization where plants volatilize contaminants; and (v) phytodegradation where plants an d associated microorganisms are used to degrade organic contaminants. In addition to problems associated with the disposal of metal loaded plants, one of the main disadvantages of this remediation approach is the long time that is required in
26 comparison to the other available methods (Alkorta and Garbisu 2001; Mulligan et al. 2001; Zavoda et al. 2001) Different plants have been used in experimental settings to remove Hg from contaminated waters. Kamal et.al (2004) found that plants such as the parrot feather, creeping primrose, and water mint could reduce Hg in an aqueous system to levels as low as 0.15, 1.3, and 0.02 g/L respectively in a 21 day remediation experiment. Bennicelli et. al (2004) found that the fern Azolla Caroliniana could reduce the dissolved Hg levels to 20 g/L in a12 day experiment. Finally, Skinner et.al (2007) reported on the ability of four different aquatic plants, namely, the water hyacinth, the water lettuce, the zebra brush and the taro to lower Hg from aqueous solutions, but the analytical technique used in this study limited the true assessment of the efficiency of these plants as the instrument detection limit was in mg/L range. Overall, plants are low cost materials. However, the long time required for the removal of the pollutant down to acceptable levels and the fate of Hg loaded plants are two of the main disadvantages of this technology. Microbial Bioremediation Biodegradation refers generally to the breakdown of organic contaminants by microbial population. Metals cannot be biodegraded; however, they can be biotransformed into compounds w ith different chemical speciation to reduce the mobility, bioavailability and/or toxicity. Microbialdriven remediation are dependent on site specific and environmental conditions such as pH, temperature, oxygen, nutrients, and soil moisture, (Alvarez and Illman 2006) Some of the limitations are related to the fact that certa in microorganisms need specific conditions to be effective, unpleasant odors
27 can be produced due to the rel ease of volatile organic carbon compounds and reduced sulfur containing gas es. Furthermore, nutrient additions to sustain bacterial growth could lead to water contamination via surface runoff and/or infiltration. With regard to Hg, its removal through bioremediation occurs by either microbial catalyzed conversion of ionic Hg to elemental mercury which is then vaporized into the air. Or by precipitation of ionic Hg through binding to reduced sulfur produced by sulfur reducing bacteria ( SRB ) under reducing conditions. The genetically engineered microorganism, Pseudomonas putida KT 2442::mer 73, reduced Hg2+ from 200 mg/l to 2.4 mg/L via conversion to elem ental Hg (Leonhauser et al. 2006) The natural isolate of Pseudomonas putida SP3 reduced Hg2+ from 192 mg/L to 3.1 mg/L (Leonhauser et al. 2006) The removal of Hg through this method is not sufficient to meet the new requirements that are in the range of ng/L. Also volati lization of Hg does not solve the pollution problem because vaporized mercury will undergo deposition and contaminate pristine systems at both regional and global scales. Constructed W etlands Wetlands are low l ying ecosystems where the water table is near or at the surface. Constructed wetlands are manmade and are used for a variety purposes. Some are used as rehabilitating areas, for wastewater treatment, as buffer zones to protect dow nstream aquatic systems, treatment of metals, etc. These wetlands have been effective in the removal of metals from wastewater and acid mine drainage areas (Hawkins et al. 1997) Some of the principles used for removal of pollutants in wetlands are microbial bioremediation, phytoremediation, adsorption, sedimentation and precipitation. King et al. (2002) found that constructed wetlands can adequat ely remove low level Hg to produce an effluent concentration in the range of 18.0 28.8 ng/L, but
28 ove r a one year period. Gustin et al (2006) also found that four experimental designs of c onstructed wetlands proved to remove Hg. All the designs contained vegetation of 70% cattails (Typha sp.) and the remaining comprised of rushes (Juncus sp.) and duckweed (Lemna sp.). The designs were (1) Hg contaminated water and Hg contaminated sediment s, (2) Hg contaminated water and clean sediments, (3) clean water and Hg contaminated sediments, and (4) clean water and clean sediments. The inflow of designs (1) and (2) came from Steamboat Creek, Nevada. The inflow of designs (3) and (4) came from the T ruckee Meadows Water Reclamation Facility, Nevada USA. The mean concentration of total Hg for the inflow of design (1) and (2) was 71 ng/L and the outflow was 32 and 25 ng/L, respectively. The mean concentration of total Hg for the inflow of design (3) an d (4) was 6.4 ng/L and the outflow was 7.2 and 5.0 ng/L, respectively (Gustin et al. 2006) Design (3) gave a higher total Hg concentration in the outflow than the inflow because there were natural levels pr esent in the wetland. It has been shown that wetlands can be constructed at low cost and require little maintenance. However, if rigorous experimental examination is lacking, then predicting the effluent characteristics of the wetlands is extremely diffic ult. This is due to the diversity and complexity of these systems. Wetlands were built to reduce phosphorus loading runoff from the Everglades Agricultural Area and downstream eutrophication. The major concern was that the wetlands would unintentionally worsen the mercury problem. Cell one of three had surface water mercury concentrations reach 32 ng/L of total Hg and 20 ng/L of methyl Hg from about 4 ng/l and 0 ng/l, respectively. By controlling the flow rate and water depth mercury methylation may be reduced. Cell one
29 met the requirements of not being significantly greater than the average inflow, 1.34 ng/L of total Hg and 0.29 ng/L of Me Hg, after four months (Rumbold and Fink 2006) Sorption Materials Activa ted carbon a dsorption Activated carbon (AC) has proven to remove a wide range of pollutants in aqueous systems (Yin et al. 2007) This is due to a number of characteristics such as a high specific surface area which ranges fro m 500 to 1500 m2g1, an internal microporosity configuration, and the presence of diverse surface functional groups (Chingombe et al. 2005) Modified AC with sulfur has been used to remove Hg from aqueous effluents, with an increased adsorption capacity of approximately 1.4 times higher than the one obtained with the use of virgin AC (Gomez Serrano et al. 1998) The basic principle may be explained by Pearsons principle of hard and soft acids and bases (Pearson 1963) Based on this classification, Hg, a soft acid would react readily with sulfur, a soft base to form highly covalent bonds. Overall, the use of both virgin and modified AC allows Hg removal to reach values in the range of 460 150 mg/g from 750 mg/g (Nabais et al. 2006) Unfortunately, these values are far above the target action limits mentioned earlier. Ion exchange resins Ion exchange resins are insoluble structures created from organic polymer beads. In this technology, the trapping and removal of pollutants from the aqueous phase is done through ion exchange. The resin type a nd fabrication methods are very diverse and allow resins to be manipulated to remove specific ions. Resins without specified functional groups have been shown not to remove targeted metals. For instance, thiol functionalized resins, have been designed for specific removal of ionic Hg species,
30 decreasing Hg levels in aqueous eff luents from approximately 10,000 g/L to concentrations below 5 g/L (Dujardin et al. 2000) The use of mercaptanamine chelating resins showed a Hg removal capacity of 621 g/ kg (Atia et al. 2005) A cation exchange resin made from banana stems, an abundant lignocellulosic biomass waste, was able to remove Hg from aqueous solutions to about 70 g/L, but the removal efficiency seemed to decrease with the increase in initial solution concentration (Anirudhan et al. 2007) Although this technology is effective to reduce Hg levels from aqueous effluents, it is apparent that its ability to reduce Hg level s below the target level of 12 ng/L is so far out of reach. It is worth to note that with regard to Hg removal by ionexchange resins, ongoing research focuses on: (i) understanding the removal mechanisms, (ii) increasing the efficiency o f Hg removal in the presence of competitive ions; and (iii) developing efficient recycling methods to guarantee the reuse of exhausted resins. Other low cost adsorbents A wide variety of natural materials are classified as low cost adsorbents due to their low cost and local availability. These low cost adsorbents are potential candidates to replace expensive adsorbents such as AC. Examples include chitosan and certain waste products from industrial or agricultural operations. Chitosan, a product of the deac e tylation of chitin which is the structural element in the exoskeleton of crustaceans, has excellent metal bi nding capacity. Penichecovas et al (1992) found that chitosan has the ability to remove Hg with an adsorption capacity of 430 mg /g. On the other hand, DiNatale et al. (2006) found that three natural materials (i.e. char of South African coal, po zzolana, and yellow tuff) had higher ratios of Hg captured to
31 amount of adsorbent when compared to AC. Most of these low cost adsorbents are st ill in the investigative phase and more research is needed to establish their efficiencies, and ultimately, their use in full scale water treatment operations. Zero valent iron (ZVI), which is the focus of this study, has been used to treat organic and inorganic contaminates from a variety of effluents like wastewater, storm water runoff, industrial, and groundwater. A detailed review of the use of ZVI for remediation is given below. Additionally the extension of this technology is seen in the use of nano s ize zero valent iron (nZVI), which has been of interest recently due to the theoretical increase in remediation power due to the increase in SSA. Zero Valent Iron (ZVI) ZVI has been in use for a number of years as an alternative to the pump and treat method in groundwater remediation. It is readily available, cost effective and has been used in the treatment of aquatic systems contaminated with recalcitrant or ganic pollutants, primarily volatile organic carbons (Gillham and Ohannesin 1994; Matheson and Tratnyek 1994) Studies have shown that ZVI could remove over 90% of DDT [1,1, 1 trichloro 2,2 bis(p chlorophenyl)ethane] from contaminated aqueous systems (Boussahel et al. 2007; Sayles et al. 1997) In addition, ZVI also removes over 95% of DDD [dichlro diphenyl dichloroethane] and DDE [dichlorodiphenyl dichloroethy lene] the products of natural transformations of DDT (Sayles et al. 1997) When used in combination with ethylene diamine tetraacetic acid ZVI accelerates the decomposition of different organic solvents reaching removal efficiencies of 100% for 2chlorophenol, 85% for phenol, 70% for ocresol, 67% for aniline, and 28% for pnit r ophenol (Sanchez et al. 2007) In wastewater treatment, ZVI has been used to lower levels of certain toxic
32 contaminants such as hydrocarbon compounds, dyes, pesticides, and herbicides (Junyapoon 2005) In addition to its use in the remediation of aqueous systems contaminated with organic pollutants, ZVI has also been used in remediation of some metal contaminated waters (Blowes et al. 2000; Blowes et al. 1997; Cantrell et al. 1995; Powell et al. 1995) For instance, chromium (Cr) VI can be reduced to Cr (III) through the r eaction shown in equations 1 and 2, leading to the precipitation of Cr (III) oxyhydroxides, and therefore to its immobilization (Astrup et al. 2000; Blowes et al. 2000; Blowes et al. 1997; Bostick et al. 1990; Powell et al. 1995) 24CrO + Fe0 + 8H+ Fe3+ + Cr3+ + 4H2O (2 1) (1 x) Fe3+ + (x)Cr3+ + 2H2O Fe(1 x)CrOOH + 3H+ (2 2) The efficiency of ZVI has been tested in experimental settings for the remediation of arsenic (As) contaminated waters, and for both As(V) and As(III), ZVI was able to lower initial high arsenic concentrations (mg/L range) to values below the current US EPA action limit of 10 g/L (Farrell et al. 2001; Lien and Wilkin 2005b; McRae et al. 1997) Besides the above two oxyanionforming elements, the literature is quite abundant with research on the removal of several other metal cations from aqueous effluents usi ng ZVI ( e.g. (Bartzas et al. 2006; Blowes et al. 2000; Gu et al. 1998; Khudenko and Garciapastrana 1987) However, unlike the widespread use of ZVI, research focusing on the use of nanoZVI (nZVI) particles in remediation of metal contaminated systems is still in its early stages Preliminary findings fr om studies dealing with As, Se, and Cr suggest that the efficiency of metal removal from aqueous
33 solutions c ould be significantly improved (Kanel et al. 2006; Li et al. 2006a; Mondal et al. 2004; Ponder et al. 2001; Ponder et al. 2000) Despite the extensive use of ZVI in remediation studies, published research on the remediation of Hg contaminated systems is still scarce. Wilkin and McNeil (2003) used ZVI to remove several metal cations i.e. Fe, Al, Hg, As, Cd, Cu, Mn, Ni, and Zn from synthetic acid mine solutions. In this study, initial Hg concentrations of about 3.1 mg/L were lowered to levels <0.07 mg/L. In a column study, Weisener et al (2005) used a ZVI reaction medium to treat a contaminated groundwater with an in itial total Hg concentration of about 40 g/L but this initial Hg level was decreased only to values much higher than t he EPAs suggested 12 to 15 ng/L for discharged wastewater effluents. Based on empirical knowledge, dissolved Hg can be removed from aqueous solutions containing ZVI through a combination of mechanisms that depend upon key factors such as Hg s peciation, pH, redox conditions and competitive cations. Such mechanisms include but are not limited to (i) loss by volatilization following th e re duction of ionic Hg to Hg0, (ii) adsorption onto solid oxyhydroxides as the ZVI undergoes oxidation, and (iii) removal through formation and precipitation of the highly insoluble mercury sulfide complexes under anaerobic conditions. Zero valent iron and v o la ti lization of dissolved mercury t hrough r eduction to elemental mercury When in contact with ZVI or nZVI dissolved Hg2+ undergoes reduction to form Hg0 as shown in Equation 3 Hg2+(aq) + Fe0(s) Hg0(g) + Fe2+ (2 3)
34 In this case, formed Hg0 will very q uickly partition between the aqueous and gaseous phases due to its very low solubility in water, with reported KH values ranging from 376 to 391 mol atm L at 20oC (Clever et al. 1985; Lin and Pehkonen 1998; Loux 2004; Sanemasa 1975; Schroeder et al. 1992) In addition to the volatilization through direct Hgn+ reduction by ZVI/nZVI, Hg0 may also form when the produced Fe2+(aq) interacts with ionic Hg as shown in equations 4 and 5 (Raposo et al. 2000; Zhang and Lindberg 2001) Hg2+ + Fe2+ Hg+ + Fe3+ (2 4) Hg+ + Fe2+ Hg0(g) + Fe3+ (2 5 ) Additionally, in natural aquatic systems, ionic Hg can be reduced to its elemental form by reacting with organic matter (Schluter 2000) sunlight (Zhang and Lindberg 2001) and through bacterial formation of Hg0 (Gabriel and Williamson 2004; Leonhauser et al. 2006; Lin and Pehkonen 1999) One of the limita tions of Hgn+ reduction to Hg0 would be its availability for these different reactions. For instance, Hg2+ adsorption onto iron oxyhydroxides that form on the surface of corroded ZVI/nZVI would limit its availability and interactions in the above listed r eactions. Adsorption on iron o xyhydro xide s Iron particles can react with water under both aerobic and anaerobic conditions to form Fe2+ and Fe3+ as shown in equations 6 through 9 (Biernat and Robins 1972; Gu et al. 1999; Kenneke and McCutcheon 2003; Majewski 2006; Sayles et al. 1997) u nder anaerobic conditions: Fe0 + 2H2O Fe2+ + H2 + 2OH(2 6)
35 2+ 3+ 221 Fe+HOFe+H+OH 2 (2 7) u nder aerobic conditions: 2 Fe0 + O2 + 2 H2O 2 Fe2+ + 4 OH(2 8) 4Fe2+ + O2 + 2H+ 4Fe3+ + 2OH(2 9) Iron oxyhydroxide layer s formed on th e surface of ZVI/n ZVI as corrosion products are made of one or more of the following amorphous minerals: lepidocrocite ( FeOOH), goethite ( FeOOH), akaganeite ( FeOOH), magnetite (Fe3O4), green rust ( Fe ( II) Fe ( III) hydroxyl salts), siderite (FeCO3), de pending on the system (Farrell et al. 2000; Gu et al. 1999; Huang and Zhang 2005; Huang et al. 2003; Phillips et al. 2000; Rangsivek and Jekel 2005; Ritter et al. 2003) Once form ed these oxyhydroxides adsorb Hg2+ which becomes encapsulated into oxyhydroxide inner layers as observed by scanning electron microscopy (SEM) in a column study (Weisener et al. 2005) Pre cipit ation of dissolved mercury a s sulfide complexes The sulfate ion, 24SO is usually present in natural waters and under favorable conditions (e.g. low pH), it can react with ZVI as shown in equation 10 (Weisener et al. 2005) 4Fe0(s) + 24SO + 10H+ H2S (g) + 4Fe2+ + 4H2O (2 10) The produced sulfide can then react with ionic Hg to produce solid mercury sulfide (HgS) complexes (Hepler and Olofsson 1975; Hsu Kim and Sedlak 2005; Paquette and Helz 1997; Weisener et al. 2005) Also the analysis of HgS precipitates fo und in such experiments have shown Hg and S in 1 to 0.85 ratio, suggesting HgS could form in low sulfide environments thus removing Hg from aq ueous solutions (Weisener et al. 2005)
36 In addition to HgS, other Hg sulfur complexes (e.g. 0 2Hg(SH) 2Hg(SH) and 22Hg(SH) ) may form, as pr edicted by thermodynamic models (Benoit et al. 2003; Miller et al. 2007) Effect of pH, competitive ions, and m ercury binding l igands Several cations can compete for the same adsorption sites on solid surfaces or in chemical reactions with binding ligands present in solution. A study investigating the role of both competitive cations and pH on Hg adsorption onto k aolinite, Sarkar et al (2000) found that the addition of chloride would shift the pH at which 50% of Hg becomes adsorbed onto kaolinite from 3.4 to 7. Similarly, the pH at which the maximum amount of Hg adsorbed onto kaolinite shifted when nickel (Ni ) was added to the mixture. However, no significant changes occurred when 24SO 34PO and Pb were added individually to the system. In contrast, Bartzas et al. (Bartzas et al. 2006) noticed an adverse effect related to the formation of sulfate green rust as an intermediate product of ZVI corrosion in sulfate treated systems. In natural systems there may be several different apparent redox levels In these systems, the change in redox adjusts the tendency for key reactions to take place (Stumm and Morgan 1996) It is well known that natural systems cycle redox changes over time. Accordingly, the reaction mechanisms under which Hg is removed from aqueous systems by ZVI may change with varying redox levels and some of the above discussed reaction mechanisms may occur simultaneously in natural systems. N ano Zero Valent Iron (nZVI) The use of nano materials (NM) has been increasing over the years. The higher efficiencies while using a smaller amount of mat erial is one of the main reasons for
37 investigating the use of NM. The general characteristic of NM is the increase in surface area, which theoretically can increase reaction rates The reason is that remediation is dependent on the surface interaction of t he contaminant. A review by Narr et al. (2007) outlining the potential use of nanotechnology in wastewater/water treatment to use NM in pollution prevention, t reatment and remediation. The use of NM can reduce the consumption of expensive chemicals and high energy consuming processes, like ultra violate light. New detection methods need to be created for these NM and environmental models are essential. However, life cycle assessments and environmental impacts of these NM need to be investigated. In addition, human health impacts of NM are required prior to the release or usage into the environment. Table 21. Site remediation using nano zero valent iron: List of recent projects (Li et al. 2006b) Site Location Phoenix Goodyear Airport (Unidynamics) Phoenix, AZ Defense Contractor Site CA Jacksonville dry cleaner sites, pilot tests using nZVI (several FL State Lead Site ID Groveland Wells Superfund Site Grove land, MA Aberdeen site MD Sierra Army Depot NV Pharmaceutical plant, Pilot test Research Triangle park, Industrial site Edison, NJ Picattiny Arsenal Dover, NJ Shieldalloy plant NJ Manufacturing Site Passaic, NJ Klockner Road Site Hamilton Towns hip, NJ Manufacturing Plant Trenton, NJ Naval Air Engineering Station Lakehurst, NJ Confidential site, Pilot test Winslow Township, NJ
38 Table 21. Continued. Site Location Confidential site, Pilot test Rochester, NY Nease superfund site, Pilot test OH Former Electronics Manufacturing Plant PA Rock Hill, pharmaceutical plant, full scale using nZVI SC Memphis Defense depot TN Grand Plaza Dry Cleaning Site Dallas, TX Industrial plant, pilot test Ontario, Canada Public domain, Pilot test Quebec, Canada Solvent Manufacturing Plant, Pilot test Czech Republic Industrial Plant, Pilot test Czech Republic Industrial Plant, Pilot test Germany Industrial Plant, Pilot test Italy Brownfields, Pilot test Slovakia nZVI have been used in pilot scale and ful l scale facilities throughout the US and internationally, as shown in Table 21. Proposed usage of nZVI as a reactive media in permeable reactive barriers is shown in Figure 22. It illustrates how a contamination source leached into the groundwater, and nZVI particles injected directly into the contaminated plume. The contamination plume may be treated to remove organic (i.e. organic solvents, pesticides, and fertilizers) and inorganic contaminants (i.e. heavy metals). The fate, transport, and toxicity of nZVI have not been fully understood. In addition the reaction mechanism of these nano particles under specific conditions has not been investigated.
39 Figure 21 Schematic illustration of how nanoparticles could be used for in situ remediation of polluted ground waters (Zhang 2003) nZVI and nitrate interaction Synthesized nZVI with a diameter range of 1100 nm and a BET (Brunauer Emmett and Teller) specific surface area of 31.4 m2/g was used to remove 50, 100, 200, 400 mg/l of nitrate in deionized water with complete conversion to nitrogen gas in only 30 minutes (Choe et al. 2000) The effect of pH on nitrate removal was investigated with nZV I ( particle size range of 5080 nm) and a BET SSA of 37.83 m2/g (Yang and Lee 2005) The isoelectric point was found to be at pH 7.3 for nZVI and at pH 5.4 the particles have the highest stability due to the large repulsive force act ing on the particles. The removal rate of nitrate at low pH was highest. This is consistent with the results of having a more stable particle which means less aggregation and optimal surface sites for remediation (Yang and Lee 2005) Zhang et al. (2006) also confirmed t hat at lower pH values ni trate reduction has a faster removal. The difference with the two works is that the latter used nZVI on supported graphite and the BET SSA for the
40 highest iron loading of 20% (by mass) was 6.18 m2/g while the other did not use any support material In all studies, as the iron content increased the amount of nitrate removed also increased (Choe et al. 2000; Yang and Lee 2005; Zhang et al. 2006) Biotic applications using n ZVI to remediate nitrate showed that removal was highest when nZVI and biota was present (Oh et al. 2007) nZVI and a rsenic interaction n ZVI with a BET SSA of 24.4 m2/g and a s ize distribution of 10 to 100 nm was used to treat arsenic from aqueous media. First order rate constants ranged from 0.07 to 1.3 min1 for n ZVI. These rate constants are 1000 times higher than reported rate constants when micron ZVI was used (Kanel et al. 2005) The isoelectric point of nZVI was determined to be at a pH of 7.8, which is similar to what Yang et al (2005) found when removing nitrate. O utside the pH range of 4.5 10, arsenic adsorptions decreased (Kanel et al. 2005) Simila r particle characteristics of nZVI with a BET SSA of 33.5 m2/g and a size distributi on of 50 to 100 nm was seen by Kanel et al (2005) (Yuan and Lien 2006) The isoelectric point was determined to be at pH 4.4, which is significantly different from that obtained by Kanel et al. (2005) a nd Yang et al. (2005) (pH of 7.8 and 7.4 respectively). The difference might be due to the addition of NaClO4 to the solution matrix while the others only used pure deionized water. Kanel et al (2006) gives an overview of nZVI versus ZVI on removing arsenic. The n ZVI rates constants, as stated by Kanel et al. earlier in 2005, are about 1000 times greater than ZVI However, the mass of arsenic removed for both ZVI and nZVI wa s similar (~ 100% ) So using n ZVI instead of ZVI suggests a faster rate but not an
41 increase in mass adsorbed. The usual time frame for remediation may take years with other adsorbents but nZVI shows promise as an adsorbent with less remediation time. nZVI and c hromium i nteraction n ZVI has also been used to treat chromium (Cr) contaminate in aqueous solutions. Ponder et al (2001; 2000) used supported nZVI. In sup ported nZVI the nZVI particles are dispersed in a fluid or gel to stabilize the nano particles. The support material may be a resin, silica gel or sand (Ponder et al. 2000) The BET SSA of the material used by Ponder et al ( 2000) w as 24.4 m2/g. The rate constant for the resin supported material was highest, 1.18 h1, but for unsupported n ZVI the rate constant was 1.16 h1. The supported material may not be needed (Ponder et al. 2000) The rate constants of granular ZVI were significantly less than that of the NM (Ponder et al. 2001; Ponder et al. 2000) Niu et al (2005) also used supported nZVI. Starch stabilized nVZI was used along with pure nZVI and ZVI powder for comparison purposes. Starch nZVI showed the highest Cr(VI) removal rate followed by pure nZVI, ZVI powder, and ZVI filling respectively (Niu et al. 2005) Theoretically the surface area decrease s in the same fashion. The initial pH affected the removal rate as well. As the pH decreased so did the rate of removal for Cr(VI) (Niu et al. 2005) Xu et al (2007) use d a cellulosebase d support for nZVI called sodium carboxy m ethyl cellulose (CMC). CMC with out iron di d not remove any of the chromium, whi ch eliminates the e ffect of the support material. Column experiments showed that Cr(VI) was not found in the effluent even after six pore volumes but total chromium did appear in the first pore volume and decreased to zero after that (Xu and Zhao 2007)
42 CM C has not been the only support material for nZVI; carbon black has been used as well Carbon black was us ed as the support material for nZVI to remove chromium. The BET SSA was found to be 130 m2/g (Hoch et al. 2008) which is the highest seen in this review. However, carbon black by itself has a high BET SSA of 80 m2/g (Hoch et al. 2008) as well. Additionally, when nZVI is synthesized, there are several starting solutions that can be used which results in different SSA. The four iron solutions and the surface area were ferric nitrate (130 m2/g), ferric oxalate (64 m2/g), ferric citrate (95 m2/g) and ferrous acetate (38 m2/g) (Hoch et al. 2008) The nZVI supported ferric nitrate removed Cr(VI) at a rate of 1.2 h1m2, which is similar to what has been seen previously (Hoch et al. 2008) Synthesized nZVI with BET SSA of 24.4 m2/g was used to investigate the reaction product of Cr(VI) (Manning et al. 2007) Li et al (Li et al 2008) found that the removal capacity for n Z VI ranged from 180 to 50 mg Cr/g n ZVI while for macro ZVI the range was typically less than 4 mg Cr/g ZVI. The data from Li and Manning suggests that Cr(OH)3 precipitates are the products formed when nZVI is used to remove chromium (Li et al 2008; Manning et al. 2007)
43 CHAPTER 3 INVESTIGATION OF MER CURY AND IRON INTERACTIONS IN AQUEOUS SYSTEMS: IMPLICATION S FOR WATER REMEDIAT ION Introduction Heavy metals are introduced to aquatic systems from both anthropogenic and natural sources, but unlike organic pollutants, metals released to the environment are not biodegradable and can undergo transformations that affect their potentials for bioaccumulation in food chains and toxicity to living organisms. Although much is now known on the biogeoc hemistry of several trace metals such as mercury (Hg), research on the development of cost effective and environmental friendly remediation techniques remains rather challenging. The now well established effects of Hg on aquatic organisms and human health have led US regulatory agencies to contemplate stringent guidelines for Hg levels in waste gaseous and aqueous effluents For instance, low Hg concentrations ranging from 12 to 15 ng/L are now being targeted for waste water effluents (Hanlon 2007) Un fortunately, current methods for remediation of metal contaminated environmental matrices often fail to meet the action limit levels imposed or targeted by federal and/or state regulatory agencies. In addition, the commercial use of available techniques remains limited due to several factors such as their prohibitive costs. A wide variety of sorbents have been tested in studies focusing on the removal of pollutants from aqueous solutions, including zero valent iron particles (ZVI), which have been in use as an alternative to the pump and treat method for groundwater remediation. ZVI is both readily available and inexpensive. In fact, the current increasing trend in the use of zerovalent iron nanoparticles ( referred to herein as nZVI) finds its origin in past intensive and still ongoing use of ZVI in the remediation of groundwater
44 contaminated with recalcitrant organic pollutants (Gillham and Ohannesin 1994; Matheson and Tratnyek 1994) The use of granular ZVI in permeable reactive barrier (PRBs) started in the 1990s; (Gillham an d Ohannesin 1994; Gu et al. 1999; Reynolds et al. 1990) and has been found effective for the treatment of many organic pollutants such as volatile organic carbons in polluted ground waters (Gillham and Ohannesin 1994; Matheson and Tratnyek 1994) Besides organic pollutants, ZVI has also been used to treat metal contaminated waters (Blowes et al. 2000; Blowes et al. 1997; Cantrell et al. 1995; Powell et al. 1995) In a laboratory study by Wilkin and McNei l (2003) the efficiency of ZVI to remove several trace metals, i ncluding Hg, from synthetic acid mine solutions resulted in significant decrease of Hg levels from initial concentrations of ~ 3100 g/L to values <70 g/L. Additionally, ZVI packed columns have also been used as reaction media to treat a Hg contaminated g roundwater with an initial totalHg ( THg) concentration of about 40 g/L, resulting in a THg concentration of 0.168 g/L in the treated effluent (2005) The use of nZ VI in remediation can therefore be seen as an extension of the above ZVI technology. However, the use of nanosize particles could provide several advantages related to physicochemical characteristics specific to nanoparticles including the distinctive cat alytic and chemical properties associated with the large surface to volume ratio characteristics of nanosize materials. The latter can lead to interesting and some time surprising surface and quantum size effects. nZVI is anticipated to be used as alternative or supplement to the conventional ZVI PRB technology. For instance, injections of nZVI slurries targeting heavily contaminated
45 source areas or hot spots could add to the efficiency of traditional ZVI PRBs that function as barriers to contain the dis persion of contaminants. This study focuses on the potential of nZVI to remove Hg from waste water effluents. I n fact, the interaction of metallic iron and Hg in aqueous solutions results in Hg removal from water through a combination of mechanisms that d epend upon key factors such as Hg speciation, pH, oxidationreduction potentials, the presence and types of binding ligands, as well as the presence of competitive cations. Such mechanisms include but are not necessarily limited to (i) loss by volatilizati on following the reduction of ionic Hg to Hg0, (ii) adsor ption onto solid oxyhydr oxides as the ZVI surfaces undergo oxidation, and (iii) removal through formation and precipitation of highly insoluble Hg sulfide species in sulfide rich anaerobic systems. B ased on these Hg removal mechanisms, one could speculate that the efficiency of Hg removal by ZVI particles can be improved by increasing the surface area, and therefore, the reactivity of used particles. Accordingly, the currently emerging nanotechnology and the production of n ZVI particles offer the opportunity to test the above hypothesis. In this study, ZVI and nZVI are used comparatively in laboratory experiments investigating their ability to (1) volatilize aqueous Hg and (2) act as sorbents for Hg di ssolved in waters of different solution chemistries. Materials and Methods Iron Particle Characterization and Chemistry of Water Used in Laboratory Experiments ZVI particles with a diameter size range of 1 to 2 mm were obtained from Alfa Aesar (PA, USA). The n ZVI particles were purchased from Quantum Sphere, Inc (CA, USA), and had a particle size range of 15 to 25 nm as reported by the vendor. The N2-
46 BET specific surface areas (SSA) of these particles were determined using a QuantoChrome NOVA 1200. The particle size distribution (PSD), zeta potential, and density of n ZVI particles were determined using a Brookhaven Zeta Plus equipment and a Quanto Chrome Ultra Pycnometer 1000. Two types of waters were used in the batch experiments conducted in this study The first type of water used was laboratory Nanopure water, referred to herein as DI water. The second water was a filtered (0.45 m) wastewater effluent (WW) collected from a wastewater treatment facility located on the campus of the University of Flor ida. For the latter, major ions and dissolved organic carbon (DOC) were determined by ion chromatography (Dionex IC320) and a Tekmar DohrmannApollo 90000, respectively. The carbonaceous biological oxygen demand (CBOD) was obtained from the wastewater treatment plants monitoring record. Mercury solutions used in different experiments were prepared by spiking each of the above waters with aliquots volumes of a stock 1000 ppm solution of Hg(NO3)2 obtained from Fisher Scientific, USA. Corrosion of Iron Part icles and Temporal C ha nges in Specific Surface A reas: Effects of Water C hemistry E xperiments were conducted using Hg free DI water and WW to investigate temporal changes in iron particles SSA as a function of solution chemistry. Briefly, for each of the i ron particle type (i.e. ZVI and n ZVI), treatments consisted of a total of 10 containers, each containing either 1g of n ZVI or 5g of ZVI. Next, for each iron particle type, DIwater was added to five of the containers, while the other five were filled with WW to a final volume of 1.1 L. The containers were sealed and left at room temperature on a laboratory bench. At time 2, 4, 6, 8, and 10 days, selected individual containers
47 were removed and the oxidized Feparticles harvested after discarding the liquid p hase. The harvested particles were then dried and out gassed at 150oC for 2 hours prior to analysis for SSA. Volatilization of Dissolved Mercury by Iron P ar ticles in C lose d Batch R eactors B atch experiments were conducted by mixing a preweighed mass of iron particles with a fixed volume of water containing a known amount of Hg. For these experiments, 50 mL serum vials precleaned by soaking in 10% trace metal grade HNO3 for 24 hours and rinsing with DI water were used as batch reactors. Hg spiked waters wer e equilibrated with known amount of iron particles. T o account for the effect of temperature, both sample vials and vials containing elemental Hg used as standard for calibration purposes were maintained in a large bath at a fixed temperature. Serum bottle s were incubated comparatively under light and dark conditions, with dark conditions achieved by fully wrapping each vial with aluminum foil. All treatments were prepared in triplicates. Over time, gas samples (100 L) were withdrawn from the vials headspace with a gas tight syringe, and injected into a stream of ultra high purity helium for the determination of Hg0 content (0 gasHg), using cold vapor atomic fluorescence spectrometry (CV AFS, Tekran2500, Ontario, Canada). Levels of Hg0 in the corresponding aqueous phase ( 0 aqHg ) were calculated using Henrys Law. This approach allowed for the determination of the total amount of Hg0 ( 0 THg ) produced in the reactor at any sampling time (i.e. 000 TaqgasHg=Hg+Hg ), and therefore the possibility to express the data as percent of initial Hg(II) in solution converted to Hg0.
48 Effect of Dissolved Natural O rga nic Matter on Mercury V olatilization Experiments were also conducted to investigate the effect of dissol ved organic carbon (DOC) on the interaction of iron with dissolved Hg. For these experiments, water samples with different organic matter content and types were collected from the Suwannee River (SR) basin. The SR system contains three linked hydrologic units, each providing distinct hydrological characteristics and gradients in DOC and ionic strength (I). Briefly, in the upper watershed of the SR system, confinement of the Floridian aquifer provides surface drainage and sources of organic carbon from wetla nds (e.g. the Okefenokee Swamp in southern Georgia). The boundary between the upper confined and middle unconfined watershed is a geomorphic feature called the Cody Scarp, below which ground water returns to the surface from many large springs, increasing ionic strength of surface waters. Finally, the river delta leads to the Gulf of Mexico and provides sites for collection of water samples with lower DOC and higher salinity. Ongoing studies in our laboratory (Gao et al. 2009) have shown that water samples collected from the headwaters and referred to herein as SR 1, the river midsection ( SR 2) and off the Suwannee River delta (SR 3) showed contrasting chemical compositions Waters were filtered (0.45 analyzed by a Tekmar DohrmannApollo 9000 for DOC, ion chromatography (Dionex DX 320) for major ions (e.g. Na+, K+, Mg2+, Ca2+, Cl-, 3NO and 24SO ), and phenolphthalein titration to pH 3.7 for total alkalinity. Table 31 and Figure 31 show trends of the major chemical parameters in these water samples. Only sample SR1 (dominated by allochtonous organic matter) and SR2 (site impacted by nutrients with a high production of autochtonous organic matter) were used i n this study.
49 Batch experiments were conducted by mixing a preweighed mass of iron particles with a fixed volume of either SR1 or SR2 water samples containing a known amount of Hg. Table 31. Chemical composition of water samples collected along the Suwannee River from headwaters to the river delta Water Sample pH Alkalinity (mg/L as CaCO3) Chloride (mg/L) TOC (mg/L) SR 1 4.70 6.0 6.2185 45.71 SR 2 7.15 88.0 6.9797 10.23 SR 3 7.56 132.0 13608.0 Not determined For these experiments, 50 mL serum vials pre cleaned by soaking in 10% trace metal grade HNO3 for 24 hours and rinsing with DI water were used as batch r eactors. Hg spiked waters, 25 mL, were equilibrated with known amount of iron particles, 0.5 g. The incubation procedure and analysis is the same as the one stated in the previous section. Figure 31: Trends of dis solved organic carbon (DOC) ( adapted from Gao et al., 2009)
50 Figure 32 Trends of ionic strength (I) in waters of the Suwannee River ( adapted from Gao et al., 2009) Hg S orpt ion ont o Zero Valent I ron and Nano Z ero V alent I ron P articles T he maximum adsorption capacities of tested iron particles for Hg dissolved in DI water or WW were determined by equilibration of prepared Hg solutions of known and increasing Hg concentrations (ranging from 100 to 500 mg/L) with a fixed amount of iron particles (i.e. 1.0 g for ZVI or 0.04 g for n ZVI). All treatments were run in triplicates, and based on preliminary sorption kinetic results ( data not shown), 1 mL aliquot of the aqueous phase was collect ed on the 10th day of the equilibration process (plateau region of the adsorption curve) and analyzed for Hg remaining in solution. Hg was analyzed by the SnCl2reduction technique after sample oxidation by bromine monochloride and prereduction by hydroxy lamine, followed by detection by cold vapor atomic fluorescence spectrometry (Tekran Model 2600). Details on used analytical procedures and QA/QC criteria (blanks and Hg standard solutions) have been
51 described in several of our earlier publications (Hovsepyan and Bonzongo 2009; Warner et al. 2005; Warner et al. 2004) Following the water particle mixing approach described above, the effect of pH on aqueous Hg removal by either ZVI or n ZVI was also investigated. In this case, the pH of prepared mixtures was adjusted to 5, 7, or 8 using HCl or NaOH solutions. All treatments were run in triplicates, and the disappearance of Hg from solutions in batch reactors monitored over time through analysis of withdrawn aliquot samples by the SnCl2reduction technique and detection by CV AFS. Finally, the effe ct of initial Hg concentration on rates of Hg removal from aqueous phase by the two types of tested iron particles was investigated using Hgsolutions with a fixed pH of 7, while the range of Hg concentrations in used waters varied from 0.05 to 5 mg/L. For these experiments, the disappearance of Hg from aqueous solutions was monitored over time and analyzed as described above. Statistical Analysis To evaluate the significance of differences observed between treatments, a simple t test for equal or unequal v ariance was used at a confidence level of 95%. Analysis was performed using the Microsoft Office Excel 2003 statistical data analysis tools. Results and Discussion Particle Characterization and Water C hemistry The measured SSA of particles used in these experiments were 0.116 m2/g and 30.5 m2/g for ZVI and n ZVI, respectively. The particle size distribution (PSD) of n ZVI ranged from 345 to 558 nm with an average diameter of 492 nm when dispersed in DI water and analyzed by dynamic light scattering technique (DLS). The PSD of n ZVI dispersed in tested WW effluent ranged from 469 to 651 nm with an average value of
52 564 nm. Trends of zeta potential for n ZVI particles suspended in DI water and WW effluent as function of pH are shown in Figure 32 The point of zer o charge (PZC) for particles suspended in DI water occurs at a pH of ~4.15, a PZC value that is quite similar to that reported previously by Yuan et al. (2006) In WW effluent, the PZC dropped to a pH of ~2.1. pH 0 2 4 6 8 10 12 14 Zet aPotential (mV) -60 -40 -20 0 20 40 60 DI-water WW Figure 32 Effect of pH on zeta potential of nanozero valent iron ( n ZVI) par ticles suspended in DI water ( ) and Wastewater Effluent ( ). Plotted values are averages of 5 measurements and in all cases, error bars are very small and do overlap with symbols. While DI water us ed in our experiments can be considered as simple synthetic water, the WW effluent had complex solution chemistry ( Table 32 ), including a background total Hg concentration of 55 g/L (or ppb ) and a relatively high dissolved organic carbon content (8 mg C/L). Therefore, the difference in solution chemistry offered by these two waters allowed for the assessment of the effect of increasing water chemical complexity on interactions between Hg and metallic iron particles.
53 Table 32 Chemical composition of used w astewater effluent. Major ions were determined by ion chromatography. *DOC = dissolved organic c arbon; **CBOD = carbonaceous biological oxygen demand. Analyzed parameters Conc entration (mg/L) Cl 125.5 F 0.583 0 34P-PO 7.079 3N-NO 4.589 K + 10.93 Mg 2+ 33.62 Ca 2+ 55.94 Na + 60.23 24SO 201.5 DOC* 8.000 CBOD** 3.070 Total Hg 0.055 0 Kinetic C hange of S pecific S urface A rea of Zero V alent I ron an d Nano Z ero V alent I ron in D eionized W ater and W astewater T o gain i nsight on the temporal effect metallic iron particles have on the two selected waters, measurements of SSA of particles exposed to Hg free waters were conducted over time The results are p resented in Figure 33. From the raw materials ( corresponding to the number 0 in Figure 33 ), it can be seen that exposure to water leads to increased SSA over time. However, there is a clear difference in the significance of values measured for ZVI (Fig. 3 3A) and nZVI (Fig. 3 3B). ZVI particles soaked in DI water showed no significant change in SSA even after 10 days of exposure while contact with WW resulted in an order of magnitude increase from 0.116 m2/g in raw particles to >2 m2/g. A similar overall trend was obtained with n ZVI, except that n ZVI exposed to WW reached SSA in excess of 100 m2/g from the initial 30.5 m2/g.
54 Figure 3 3. Effect of Water Matrix on Iro n Surface Area. The increase in SSA is likely due to the formation over time of oxyhydroxide coating layers on metallic iron particles, and the observed differences driven by solution chemistry. This is a well studied phenomenon, and depending on the chem istry of water, hydr(oxide) layers formed on the surface of ZVI and n ZVI can include amorphous minerals such as FeOOH), magnetite (Fe3O4), green rust ( FeIIFeIII hydroxyl salts), siderite (FeCO3) (Farrell et al. 2000; Gu et al. 1999; Huang and Zhang 2005; Huang et al. 2003; Phillips et al. 2000; Rangsivek and Jekel 2005; Ritter et al. 2003) These oxyhydroxides would provide adsorption sites for metal cations such as n+ 2xHg(HO) which would then become
55 encapsulated into the oxyhydroxides inner layers as observed by others using scanning electron microscopy (Weisener et al. 2005) Mercury Volatilization by M etallic Iron Particles: Effect of Particle S ize I n contact with metallic iron, dissolved ionic Hg can undergo reduction to form elemental mercury (Hg0) as shown in equation 1. Hg2+ (aq) + Fe0 (s) Hg0 (g) + Fe2+ (3 1) In this case, formed Hg0 would quickly partition between the aqueous and gaseous phases due to its very low solubility in water and reported KH values ranging from 376 to 391 L.atm.mol1 at 200C (Clever et al. 1985; Lin and Pehkonen 1998; Loux 2004; Sanemasa 1975; Schroeder et al. 1992) On the other hand, meta llic iron particles can react with dissolved oxygen (DO) and water in corrosion reactions by which iron is oxidized as illustrated in equations 2 and 3. 2Fe0 ( s ) + 4H+ ( aq) + O2 ( aq ) 2Fe2 + ( aq) + 2H2O (3 2) Fe0( s) + 2H2O Fe2 + ( aq) + H2 ( g ) + 2OH ( aq) (3 3) In addition, Hg0 may also be produced from the product of metallic iron oxidation (Fe2+) as illustrated in the stepwise equations 4 and 5 (Raposo et al. 2000; Zhang and Lindberg 2001) Hg2+ + Fe2+ Hg+ + Fe3+ (3 4) Hg+ + Fe2+ Hg0 (g) + Fe3+ (3 5) Metallic iron particles can react with water under both aerobic (Eq. 6 and 7) and anaerobic (Eq. 8 and 9) conditions to form Fe2+ and Fe3+ as shown in the following equations (Biernat and Robins 1972; Gu et al. 1999; Kenneke and McCutcheon 2003; Majewski 2006; Sayles et al. 1997) 2Fe0 + O2 + 2H2O 2Fe2+ + 4OH(3 6)
56 4Fe2+ + O2 + 2H+ 4 Fe3+ + 2 OH(3 7) Fe0 + 2H2O Fe2+ + H2 + 2OH(3 8) Fe2+ + H2O Fe3+ + 2 1 H2 + OH(3 9) These last four reactions bring about the possibility of oxyhydroxide coating layers on iron particles, and therefore, the presence of sorption sites for metal pollutants such as Hg as discussed earlier in relation with results presented in Figure 3 3. Using closed batch reactors and Hg containing DI waters mixed with metallic iron particles, trends of volatilized Hg0 were obtained from aliquot gas samples taken from the vials headspace. The obtained res ults are presented in ( Figure 3 4). In the presence of ZVI, Hg volatilization shows an increasing trend during the first 5 days of incubation, followed by a brief plateau, and then a progressive decrease over time to reach Hg levels below our analytical detection limit. In contrast, the interaction of dissolved Hg with n ZVI resulted in instantaneous Hg reduction with peak volatilization in the first few minutes of contact, followed immediately by a sharp loss of Hg0 from the gas phase. As illustrated in F igure 35, the observed behavior of Hg throughout the above volatilization experiment is bracketed between dominant Hg reduction and Hg adsorption endmembers. The observed peaks correspond to exposure periods during which rates of Hg reduction by metallic iron particles is significantly higher than that of metallic iron corrosion by both dissolved oxygen and water molecules.
57 Figure 34 Temporal trends of Hg volatilization and sorption in a closed batch reactor containing 30 mL of DI water with a concentration of 1.0 g Hg/mL and 0.04 g of zero valent iron (ZVI) particles. For each sampling point, levels of Hg0 in the aqueous phase were calculated using Henrys Law, and the total amount of produced Hg0 determined as 000 aqgasHg=Hg+Hg Results are plotted as ratio s of the above quantified Hg0 to the initial total Hg concentration (Y axis) versus time. However, as corrosion increases, the removal of Hg from solution through adsorption drives the depletion of Hg0 (g) from the reactor s headspace ( Figure 35). This Time (days) 0 5 10 15 20 25 30 (Hg 0 )/(Hg II) 0.00 0.02 0.04 0.06 0.08 0.10 Time (days) 0 5 10 15 20 25 30 (Hg 0 )/(Hg II) 0.00 0.02 0.04 0.06 0.08 0.10 ZVI nZVI
58 is possible due to the metallic Hg reducing Fe3+ based on the serial arrangement of electrode potentials (i.e. electromotive series). However, results presented in Figure 34 show the difference in kinetic rates of these reac tions when ZVI and n ZVI are used. These results point to the importance of physicochemical characteristics specific to n ZVI in impacting the rate of reactions between metallic iron and ionic Hg on one hand, and metallic iron and dissolved oxygen or water on the other. Figure 35 S implified s chematic representation of Hg interaction with metallic and corroded iron particles in aqueous solutions along a corrosion gradient Hg reduction dominates initially, and as iron corrosion reactions evolve, hydr(oxide) layers are formed on metallic iron surfaces and progressively, Hg adsorption becomes the dominant pathway for Hg removal from solution Overall, Hg volatilization under used experimental conditions, and regardless of the type of metallic iron particles used, was rather minimal, representing <10% of total Hg concentrations in the reactors. When WW effluent instead of DI water was used in experiments similar to the one discussed above, different volatilization patterns were obtained ( Figure 36). Unlike the experiment with DI water, here we tested the effects of pH change on Hg volatilization 2e 0Hg 0Fe 0 2nFe+nHOFe(OH) 2+ nnHg+Fe(OH)FeHgOH) Dominant Hg Reduction End Member 2+Fe 2+Hg Fe 0 Fe 0 Fe(OH) n Fe(OH) n Dominant Hg Adsorption End Member
59 by metallic iron particles under either light or dark conditions First, volatilization was faster in the presence of n ZVI, regardless of light/dark exposure conditions. Second, the volatilization step was followed by a plateau of Hg0 concentration in WW n ZVI mixtures. On the other hand, ZVI WW mixtures behaved differently, in that, at pH 5 very little Hg was reduced to Hg0, and light condit ions seemed to favor Hg reduction and volatilization. In these reactors, obtained trends suggest that Hg volatilization can persist for a long time. At pH 7, WW containing ZVI showed a spike in Hg0 concentration, but instead of a plateau as seen with n ZVI, a slow but decreasing trend similar to that described in figure 3 is observed. Volatilization experiments at pH 8 gave Hg0 levels and trends similar to those obtained with WW at pH 7 (data not shown). It is obvious that in this complex water matrix, Hg s peciation and the presence of dissolved organic compounds play a role in the fate of Hg present in water. The use of Visual Minteq and parameters listed in Table 1 showed that for the used WW effluent, Hg would occur predominantly as HgCl2 (70%) at pH 7, w hile the other species would be present at much lower proportions (e.g. 25% HgClOH, 3% HgCl 3 and 2% Hg(OH)2). In contrast, Hg speciation in used DI water at pH 7 was dominated by Hg(OH)2 (99%). It appears that in addition to particl es properties, Hg chemical speciation and the type of ligands present in solution would play a role in controlling the extent of Hg volatilization, probably through a combination of several mechanisms. This is because in natural aquatic systems, ionic Hg2+ can be reduced to its elemental form through abiotic reactions with organic matter (Schluter 2000) sunlight (Zhang and Lindberg 2001) and through microbial catalyzed formation of Hg0 (Gabriel and Williamson 2004; Leonhauser
60 et al. 2006; Lin and Pehkonen 1999) Overall, the observed volatilization phenomenon impacted only a small fraction of dissolved Hg. pH 7.0Time (days) 0 10 20 30 40 ng Hg Volatilized / g iron 0 2000 4000 6000 8000 10000 12000 pH 7.0 0 10 20 30 40 pH 7.0 Light Conditions pH 7.0 Dark Conditions Zero Valent Iron 0 2000 4000 6000 8000 10000 12000 Nano Zero Valent Iron pH 5.0 pH 5.0pH 5.0 Light Conditions pH 5.0 Dark Conditions Figure 36. Temporal trends of Hg volatilization and sorption in closed batch reactors containing 25 mL of wastewater effluent with a concentration of 1 g Hg/mL and 0.04 g of iron particles. For each sampling point, levels of Hg0 in the aqueous phase were calculated using Henrys Law, and the total amount of produced Hg0 determined as 000 aqgasHg=Hg+Hg Results are plotted as ratios of the above quantified Hg0 to the mass of iron particles used Effect of Dissolved Natural Organic Carbon on Volatilization of A queous Mercury Results obtained from experiments investigating the effects of naturally occurring dissolved organic matter (DOC) on Hg ZVI inter actions are summarized in Figure 37. C ontrol experiments (i.e. reactors with no ZVI addition) run in parallel with ZVI spiked reactors show that the presence of iron inhibits Hg volatilization in DOCrich waters
61 collected from the Suwannee River (site SR 1 with a DOC concentration of 45.71 mg C/L) (Figure 3 7). 0 200 400 600 800 ZVI Control ZVI Control Time (days) 0 10 20 30 40 ng Hg 0 Volatilized 0 200 400 600 800 ZVI Control 0 10 20 30 40 ZVI Control Hg spiked DI-water Hg spiked SR1 Light Light Dark Dark Figure 37. Effect of natural organic carbon ( SR1) on Hg volatilization (300 g/L) The reduction and subsequent v olatilization of Hg occurs when conditions are favorable to electron transfer to t he Hg cation. Such electrons c ould come from several sources when dealing with natural water and its complex chemistry For instance, in the presence of light, photocatalytic reactions can produce free radicals that can reduce Hg2+ to Hg0 (Deng et al. 2009; Ravichandran 2004) However, Figure 3 7 shows however that Hg volatilization in these experiments could be quite independent of photocatalyzed reactions, as the latter may not be the dominant pat hway for induced Hg reduction. There are three different interactions that can occur, (i) DOC and Hg, (ii) DOC and ir on, and (iii) Hg and iron
62 Another type of electron donor is natural organic matter. It is known that Hg may be reduced directly by the hu mic and fulvic fractions of natural organic matter see equation Equation 10 (Alberts et al. 1974; Ravichandran 2004; Skogerboe and Wilson 1981) Alberts et al. (1974) suggested that quinones (Figure 38) present in large organic molecules could act as electron donors for Hg reduction. Figure 38. Quinones: hydroquinone (QH2), fully oxidized quinone (Q), semiquinone radical ( QH ) (Uchimiya and Stone 2009) Uchimiya et al. (200 9) proposed that electrons are transferred when hydroquinones (QH2) is converted to Q, the fully oxidized quinone. The effect of DOC can be seen in Figure 37 when comparing the different iron treatments to the control samples. In the control samples, it i s hypothesized that when the DOC is oxidized, Hg2+ in solution is reduced to Hg0 and then subsequent volatilization. In addition Hg can be adsorbed by DOC present in the natural water. Hg binds to the acid sites of DOC. These acid sites consist of carboxy lic acids, phenols, ammonium ions, alcohols and thiols (Ravichandran 2004) 2+ 0Hg+2DOCHg+DOC (3 10) 2+ nnHg+nDOC(DOC)Hg (3 11)
63 This will decrease Hg availability to take part in other reactions. However based on the results shown in Figure 37, Hg adsorption by DOC is not d ominant because in control samples Hg volatilization is higher than in the presence of iron. However, when ZVI is present, there may be other reactions occurring. For insta nce, ZVI may be favored more than Hg for reduction. Based on standard electrochemical potentials iron (E0= 0.44V (Noubactep 2008) is thermodynamicall y preferred tha n Hg2+ (E0=0.79V (Atkins 1994) ). Additionally, DOC can be removed fr om solution by adsorption onto the ZVI surface thus decreasing irons reactivity (Hagare et al. 2001) It is apparent that iron introduction into SR waters inhibits the production of Hg0. Hg and iron interaction can yield Hg volat ilization and adsorption. ZVI bare surface can reduce Hg2+. Overtime iron oxyh y d r oxides can form and Hg can be removed f ro m solution through adsorption onto these surfaces. Time (days) 0 2 4 6 8 10 12 14 16 18 20 ng Hg 0 Volatilized 0 20 40 60 80 ZVI and SR1 ZVI and SR2 Control SR1 Control SR2 Figure 39. Effect of dissolved natural organic carbon on Hg reduction/volatili zation SR1 and SR2 water samples which are characterized not only by different DOC concentrations (45.71 and 10.23 mg C/L, respectively), but probably by the type of
64 predominant organic compounds (with allochthonous organic carbon being dominant in SR1 wh ile SR2 contains a significant authochtnonous fraction) were used comparatively Figure 39 shows that as the organic carbon concentration is increased from sample SR2 to SR1, reactors without ZVI addition (controls) show an inverse trend in the amount of Hg volatilized. It is likely that the Hg/DOC ratios as well as the Hg binding capacity of organic compounds present in these samples do control the reduction and subsequent volatilization of Hg. In contrast to these control samples, ZVI containing reactors exhibited much lower volatilization of Hg, but with SR2 (water with the smaller DOC concentration) leading to the highest Hg volatilization, comparatively. Therefore, Hg reduction/volatilization is limited in the presence of ZVI, but DOC concentration and probably the type of organic carbon compounds present play a role in determining the quantity of Hg volatilized. Time (days) 0 2 4 6 8 10 12 ng Hg0 Volatilized/g iron 0 500 1000 1500 2000 2500 3000 3500 40 ug/L, Dark 40 ug/L, Light 300 ug/L, Dark 300 ug/L, Light 1000 ug/L, Dark 1000 ug/L, Light Figure 310. Effect of Hg concentration on its volatilization from SR1 waters Figure 310 shows that when Hg concentration is increased fro m 40 to 1000 g Hg/L the amount of Hg volatilized increases as well. This is due to increase in Hg for
65 reduction in the system whether it be by ZVI or DOC, as explained above. At the higher concentrations light and dark conditions trends separate. This is caused by the presence of light in natural waters producing reactive oxygen species that may interact with Hg to form Hg0 (Lalonde et al. 2004; Lin and Pehkonen 1999) Mercury Removal from S olution by Zero V alent I ron and Nano Z ero V alent I ron through S orption Hg loading capacity of both ZVI and nZVI at pH of 5.3 and 8.3 were measured using 100, 250, and 500 mg/L at a predetermined equilibration time of 10 days. Table 3 3 shows that nZVI has a higher Hg loading capacity than ZVI at different pH and concentrations. At pH 5.3 oxyhydroxide layers seemed to have formed at a faster rate thus removing approximately twice as much Hg from solution than at pH 8.3. However, when using the Langmuir adsorption model (R2=0.99 for all treatments) the maximum adsorption capacity for ZVI and nZVI was determined to be 109 and 833 mg/g at pH 5.3 respectively At pH 8.3 the maximum adsorption capacity of ZVI was higher, 120 mg/g, but lower for nZVI, 286 mg/g. Equation 3 12 shows the Langmuir model used where b is the adsorption maxima (mg/g), q is the amount adsorbed (mg/g) ,Ceq is the equilibrium solution chemistry (mg/L) and KL is adsorption constant (intensity). (3 12) Based on preliminary studies on the effect of pH (5.0, 7.0 and 8.0) on sorption kinetics a pH of 7.0 was chosen for all adsorption studies. Figure 311 gives the results of Hg removal from solution by either ZVI or n ZVI mixed with DI water or WW with different initial Hg concentrations. T he removal cur ves for both ZVI and n ZVI in the eqeq LCC 1 =+ qbbK
66 batch experiments showed similar removal trends when the initial Hg concentration varied. Table 33 Hg loading capacity of ZVI and nZVI for Hg spiked wastewater effluent at pH 5.3 and 8.3. pH 5.3 pH 8.3 Initial Hg Concentration ( mg Hg/L ) mg Hg adsorbed / g iron mg Hg adsorbed / g iron ZVI n ZVI ZVI n ZVI 500 195.0 448.4 96.1 239.0 250 169.8 373.3 53.1 139.3 100 51.3 130.9 22.8 63.3 ZVI has been shown to remediate inorganic contaminants through co precipitation or reductive precipitation (Blowes et al. 2000; Su and Puls 2001; Weisener et al. 2005) Weisener et al (2005) showed that ZVI could interact with aqueous Hg present in groundwater to form Hg sulfide complexes. These removals from aqueous solutions w ere found to take place on inner sphere complexes with the iron oxides (Fendorf et al. 1997; Lien and Wilkin 2005a; Su and Puls 2001) Despite the similarity in observed Hg removal trends, the removal rates determ ined by assuming pseudo first order reactions produced rate constants of different orders of magnitude (Table 3 3 ). Similar to the results reported by Sarathy et al (2010) the rate constants are ex pressed as per mass (km) and as per specific surface area (ks). Particles used were not suspended either by physical or chemical means therefore in these experiments the particles form ed aggregates. The rate constants for ZVI are higher than that of its n anosize counterpart when rate constant values are expressed per surface area units. For example, Su et al (2001) express ed k values as per SSA because the iron particles were continuously suspended by a physical force. A dditionally as shown in Figure 3 2, the SSA changes over time when in contact with aqueous solutions, therefore
67 expressing the k values per mass will give a better and more accurate insight to the Hg and iron interaction. Su et al (2001) used powder and granular ZVI to remediate arsenic and found that there was no correlation between arsenic remediation and surface area. Therefore all the data shown is normalized based on mass. As a matter of fact, other researchers have encountered similar issues when using nano particles to remediate chromium (VI) (Li et al. 2008) arsenic (Berger et al. 2006; Sharma et al. 2010) carbon tetrachloride (Nurmi et al. 2005) 1,2,3 trichloropropane (Sarathy et al. 2010) fluoride (Mohapatra et al. 2010) and other contaminants expressed loading capacity per mass basis. The kinetic rates for nZVI in Hg spiked WW and DI water both ar e orders of magnitude higher than for ZVI when expressed as per mass basis. The rate constant for n ZVI increased from 1.2505 to 2.6065 day1/g iron, when the initial concentration increased. However for ZVI the rate constants (km) at an initial concentration of 0.5 and 5 mg Hg/L for both DI water and WW were not significantly different. Table 3 4. Pseudofirst order adsorption rate constants for Hg on ZVI and n ZVI based on Hg spiked deionized water and wastewater effluent (* k value s are rate constants) Deionized water (DI) Wastewater effluent (WW) Initial Hg Conc mg/L ZVI nZVI ZVI nZVI k DI k DI k WW k WW r2 day1/ g iron day1/ SSA r2 day 1/ g iron day1/ SSA r2 day1/ g iron day -1/ SSA r2 day 1/ g iron day-1/ SSA 0.05 0.94 0.081 3.5 0.95 1.5 0.010 0.95 0.044 1.9 0.78 1.3 0.008 0.5 0.99 0.065 2.8 0.87 0.83 0.005 0.97 0.081 3.5 0.95 2.0 0.013 5.0 0.94 0.060 2.6 0.92 2.5 0.016 0.99 0.075 3.2 0.97 2.6 0.017
68 Time (days) 0 2 4 6 8 10 0.0 0.2 0.4 0.6 0.8 1.0 C/Co 0.0 0.2 0.4 0.6 0.8 1.0 Time (days) 0 2 4 6 8 10 0.0 0.2 0.4 0.6 0.8 1.0 Hg-spiked WW 0.0 0.2 0.4 0.6 0.8 1.0 C/Co 0.0 0.2 0.4 0.6 0.8 1.0 Hg-spiked DI-water 0.0 0.2 0.4 0.6 0.8 1.0 ZVI nZVI (A) 0.05 mg Hg/L (D) 0.05 mg Hg/L (B) 0.5 mg Hg/L (E) 0.5 mg Hg/L (C) 5.0 mg Hg/L (F) 5.0 mg Hg/L Figure 3 11. Effect of initial Hg concentration dissolved in either Deionized water (DI water) or wastewater effluent (WW) on Hg adsorption by ZVI, ( ), and n ZVI, ( ) particles at pH 7. (A), (B), and (C) for Hg spiked DI water. (D), (E), and (F) for Hg spiked WW. Conclusion The interactions between iron and Hg exhibit two main mechanisms, volatilization and adsorption. In the batch studies the volatilization rate for n ZVI was faster than for
69 ZVI in DI water. When WW effluents were used, the amount of Hg volatilized was the same for ZVI and nZVI. The total Hg volatilized was only 1% ; therefore for both ZVI and n ZVI volatilization is not dominant. All the volatilization experiments in this study w ere done in closed batch reactors, and further investigation is needed on Hg volatilization under open conditions. This would show if the flux of Hg volatilization is constant or if it is contingent on the Hg concentration in the system. In addition the effect of different water chemi stry might also be of interest. It is known that Hg can be volatilized by humic substances (Alberts et al. 1974; Ravichandran 2004) The experimental and t heoretical adsorption capacities show that n ZVI has a higher adsorption capacity than ZVI. Further investigation using other nat ural waters with different chemistries might give more insight into how n ZVI may be used to treat other effluents such as indust rial or even drinking water. This would support the use of nZVI as a more efficient adsorbent than its granular counterpart, similar to the remediation of arsenic. The kinetic studies show that when rate constants are express as per mass basis nZVI has a f aster rate than ZVI. However the controversial issue is that these rate constants should be expressed as per surface area. This is because the hypothesis of using nano particles is to benefit from the much larger surface area. Although, when the rate const ants are expressed as per surface area then ZVI exhibits a faster rate. Others in the literature as explained above have found similar trends. Su et al. (2009) showed that there is no correlation between iron surface area and remediation. Further investiga tion into explain this correlation is needed. Maybe the iron surface area does not play a role but the solution chemistry, which can be seen in Figure 3 2. The surface
70 area changes overtime when in contact with different water chemistries. Additionally the use of nZVI under dynamic conditions needs to be examined. If nZVI leaches into the environment via these remediation techniques, then knowledge of its fate and transport is needed.
71 CHAPTER 4 R EMOVAL OF M ERCURY FROM WASTEWATER EFFLUENT USING FLOW THRO UGH COLUMNS PACKED WITH NANO ZERO VALENT IRON PARTICLES Introduction The development and use of engineered nanomaterials (NMs) are believed to offer great benefits to society through their exploitation within numerous industrial activities, and environmental, agricultural and medical research. With regard to the contaminated environmental systems, NMs have the potential to afford environmental benefits by enabling both the detection of pollutants and the remediation of contaminated environmental matrices. Of particular interest is the use of nanozero valent iron particles (referred to herein as nZVI). Theoretical and practical evidence suggests that nZVI can be used to remediate sites contaminated with certain hazardous organic (e.g. PCBs) and inorganic ( e.g. arsenic) chemicals, a process often termed nanoremediation. The current increasing trend in the use of nZVI finds its origin in past intensive and still ongoing use of bulk zerovalent iron ( or ZVI), primarily as an alternative to the pumpand treat technique for remediation of groundwater contaminated with recalcitrant organic pollutants (Gillham and Ohannesin 1994; Matheson and Tratnyek 1994) The use of granular ZVI in permeable reactive barrier (PRBs) for groundwater treatment started in early 1990s (Gavaskar 1999; Gillham and Ohannesin 1994; Gu et al. 1999; Reynolds et al. 1990) and since, over 100 such PRB structures have been constructed in the U.S. alone (Li et al., 2006). Although several studies have de monstrated that ZVI is effective for the treatment of many pollutants commonly found in polluted ground waters (e.g. perchloroethene, trichloroethene, carbon tetrachloride, energetic munitions such as TNT and RDX, legacy organohalogen pesticides such as li ndane and DDT,
72 heavy metals, etc. important challenges that limit the practical application of granular ZVI technology still exist. For example, a large amount (e.g., tons) of iron powder is us u ally needed even for a modest PRB structure. Costs associated with the PRB construction, especially for deep aquifers, remains too high for many potential users of the technology. Another important limitation is the relative lack of flexibility after a PRB is installed. Relocation or major modifications to the PRB i nfrastructure is often impractical. And besides ZVI, a number of sorbents including zeolite, hydroxyapatite have also been tested in PRB structures (Cundy et al. 2008; Moraci and Calabro 2010; Mulligan et al. 2001) The nZVI technology is therefore seen as an extension of the above ZVI technology. It takes advantage of physicochemical characteristics specific to nanosize particles that result in distinctive mechanical, magnetic, electronic, catalytic, and chemical properties. The large surfaceto volume ratio characteristics of nanosize materials can lead to interesting and some time surprising surface and quantum size effects. Overall, as the size of the particles decreases, the proportion of surface and near surface atoms increases; and surface atoms would tend to have more unsati sfied or dangling bonds with concomitantly higher surface energy (Li et al. 2006a) Thus, the surface atoms would have a stronger tendency to interact, adsorb, and react with other atoms or molecules in order to achieve surfac e stabilization (Service 1998) nZVI is anticipated to serve as either an alternative or supplement to the conventional ZVI based PRB technology. For example, injections of nZVI could be used to address the heavily contaminated source area or other hot spots whereas the traditional ZVI PRB would function as barrier to contain the dispersion of contaminants.
73 Because of their small size, nanoparticle slurries could be injected under pressure and/or even by gravity flow to the contaminated area and under certain conditions remain in suspension and flow with water for extended periods of time. Alternatively, nZVI can also be used in the treatment of wastewater effluent as investigated in this research. Column s have been used in several past studies to determine the efficiency of sorbents such as ZVI in dynamic systems (Blowes et al. 1997; Gu et al. 1999; Matheson and Tratnyek 1994; Moraci and Calabro 2010) Example studies of the removal of pollutants from aqueous solutions using columns have use d water contaminated with inorganic species such as hexavalent chromium (Cr+6) (Astrup et al. 2000; Blowes et al. 1997; Puls et al. 1999) uranium (Morrison et al. 2001; Simon et al. 2003) arsenic (Koeber et al. 2005; Lien and Wilkin 2 005b; Su and Puls 2003) nitrate (Till et al. 1998; Westerhoff and James 2003) and mercury (Hg) (Weisener et al. 2005) Column studies have also been used to determine the efficiency of sorbents on complex water matrices such as acid mine drainage (Wilkin and McNeil 2003) In the only field pilot study of removal of Hg from groundwater using columns packed with ZVI, Weisener et al. (2005) were able to lower the concentration of treated waters from 40 g/L to 4.0 g/L (or 4000 ng/L) This drop of initial concentration to low ppb level is rather promising. T he availability of highly sensitive analytical techniques (e.g. cold vapor atomic fluorescence spectroscopy or CV AFS) detecting aqueous Hg in parts per trillion levels on one hand; and the potential for harm to aquatic biota by long term discharges of wastewater effluents containing Hg concentrations in the low ppb range has led the US EP A to target 12 to 15 ng/L as action limit concentrations for
74 discharged wastewater effluents (Hanlon 2007) Unfortunately, there is no remediation technique available, which removes aqueous Hg from wastewater effluents to such low levels. The objec tive of this research is to use nZVI in column experiments to: (i) assess the potential of these NMs to remove Hg from aqueous effluents in flow through columns; (ii) investigate the different parameters that affect the efficiency of Hg removal; and (iii) lay the groundwork for the development of remediation approach that meets the Hg concentration range targeted by the US EPA. Materials and Methods Particle Characterization For comparison purposes, the initial set of column experiments used both ZVI and n ZVI. For these experiments, ZVI particles had a size ranging from 1 to 2 mm and were obt ained from Alfa Aesar (PA, USA). The nZVI particles were obtained from Quantum Sphere, Inc. (CA, USA ), with vendors reported particle size range of 15 to 25 nm. For th ese particles, the N2 BET specific surface areas (SSA) were determined using a Quanto Chrome NOVA 1200 at the Particle Engineering Research Center, University of Florida. The measured SSA values were 0.116 m2/g and 30.5 m2/g for ZVI and nZVI, respectively. Column Preparation and Setup Polyvinyl chloride (PVC) columns were purchased from Alsco Industrial Co (Lithia Springs, GA). The columns were 20 cm long and had an inner diameter of 2.54 cm. Each column was fitted with PVC screw caps at both ends. The sc rew caps were perforated to hold a Teflon tubing which allowed controlled water pumping in and out of the columns. Before being capped, columns were packed with a reactive solid
75 mixture made of sand and iron particles. In fact, columns packed with 10 0% ZVI have been used by other researchers to remove Hg from ground water. Findings from these studies show that ZVI oxidation over time leads to t he formation of thick oxyhydrox ide layers and clogging of the columns (Cundy et al. 2008; Weisener et al. 2005) To avoid this outcome, a reactive mixture of sand and iron particles is used instead, and the mass of iron particles used to spike the sand was based on SSA equivalent. Accordingly, in these column studies, the reactive mixtures were made of sand containing either 0.03 g of nZVI or 8 g of ZVI dispersed in sand with a final total mass of 25 g. This reactive mixture was then sandwi ched between support layers of 5 g of plain sand and glass wool, (see Figure 41). The inlets o f the col umns were connected to the Hg containing influent container by pumps High pressure pumps ( LC 10AT Shimadzu) commonly used in High Performance Liquid Chromatography (HPLC) were used here to obtain an accurate control of flow rates. In addition, the use of such pumps allowed for a complete wetting of the columns and helped avoid the formation of voids filled with trapped air (Bartzas et al. 2006) The Teflon tubing on the outlet end was used to collect sample effluents over time. Water Used in Column Studies and Experimental Protocol The water used as influent in these column studies was a filtered (0. 45 m) wastewater effluent (WW) collected from the wastewater treatment facility located on the campus of the University of Florida. The background solution chemistry of used WW is presented in Table 4 1. Note a background total Hg concentration of 55 g/L and a dissolved organic carbon (DOC) content of ~8mg C/L was present in the WW.
76 Figure 41. Picture of the experimental setup used in described column studies. The different layers are labeled, where the reactive mixture corresponds to the p or tion of the column containing a mixture of sand and iron particles in fixed ratios. 1:2.12 g/g for ZVI and 0.01:6.66 g/g for nZVI The above background Hg level was increased intentionally prior to use in column experiments by spiking with aliquots of a Hg(NO3)2 solut ion (1000ppm standard solution obtained from Fisher Scientific, USA). Table 41. Chemical composition of wastewater effluent used in column studies Parameter Concentration (mg/L) Chlorine, Cl 125.5 Fluoride, F 0.583 Phosphate, PO 4 7.079 Nitrate, NO 3 4.589 Potassium, K + 10.930 Magnesium, Mg 2+ 33.620 Calcium, Ca 2+ 55.940 Sodium, Na + 60.230 Sulfate, SO 4 2 201.5 Dissolved Organic Carbon 8.000 Carbonaceous Biological Oxygen Demand (CBOD) 3.070 Mercury Concentration 0.055
77 Prior to initiating the Hg removal experiments, columns packed with iron particles were first rinsed by pumping through deionized (DI) water for 24 hours at a flow of 50 L/min. This first step was then followed by the introduction of a 0.1M sodium hydroxide (NaOH) solution. Columns were filled up with NaOH and left without flow for 120 hours contact, to help expedite the oxidation of metallic iron surfaces (Feng et al. 2007) This step was intended to favor the formation of oxyhydroxide layers that s upport the removal of Hg by adsorption and minimizes Hg losses from solution by volatilization. After rinsing with DI water, Hg containing influents were pumped up gradient through the packed columns to initiate the Hg removal from WW effluent experiments. On temporal basis, effluent aliquots were collected at the outlet end of the column and analyzed for total Hg concentration. Hg was analyzed by the SnCl2reduction technique after sample oxidation by bromine monochloride (BrCl), prereduction by hydroxylamine and then detection by atomic fluorescence spectrometry (Tekran Model 2600). Details on used analytical procedures and QA/QC criteria (blanks and Hg standard solutions) have been described in several earlier publications by our research group (Hovsepyan and Bonzongo 2009; Warner et al. 2005; Warner et al. 2004) Optimization of Mercury Removal in Columns Containing Nano Zero Valent Iron : Effects of Particle M ass, Flow Rate, and W ater Chemical Composition Additional experiments were conducted to assess the effects of selected key parameters on the ability of nZVI to remove Hg in flow through columns. First, the effect of nZVI mass used to prepare the reactive media was investigated by running columns with nZVI content of 0.1% and 1.0% on a weight basis. Hg containing effluents were pumped up gr adient through the columns at a flow rate of 50 L/min.
78 In a second set of experiments, by maintaining the nZVI content at 0.1%, the flow rate of 25 and 75 L/min was used in addition to the 50 L/min initial value and the effect of Hg removal assessed. A nother set of experiments used hydrogen peroxide (H2O2) treated WW effluent at a final concentration of ~1% to help oxidize the dissolved organic compounds, and therefore, release Hg complexed by such organic ligands prior to introduction to the 0. 1% nZVI columns. Use of columns in series and study of the effects of competitive cations In these experiments, two columns packed with 0.1% nZVI, and connected in series were used to (i) investigate the potential for increased Hg removal to ng/L level, and (ii) assess the impact of cation competition. The WW influent was pumped up gradient of column 1 and the effluent of column 1 was pumped up gradient of column 2. The effluent of column 2 was then sampled over time, and analyzed for total Hg as described earlier A similar setup was used. However, to investigate the effects of Zn and Cd (Zn and Cd were added to used influents to produce final concentrations 100 g/L or ppb for each) as competitive ions on Hg removal by nZVI. Statistical Analysis For all of the above experiments, duplicate columns were run, and whenever necessary, experiments were repeated. To evaluate the significance of differences observed betw een treatments, a simple t test for equal or unequal variance was used at a confidence level of 95%. The analysis was performed using the Microsoft Office Excel 2003 statistical data analysis tools.
79 Results and Discussion Effects of Iron Particle Size on Mercury Removal from Wastewater Effluent Figure 42 gives the trends of Hg removal from the WW effluent in columns packed with nZVI and ZVI. The Hg effluent concentration (C) was normalized by the Hg influent concentration (Co) The column breakthrough oc curs when the effluent concentration matched the influent concentration and C/Co equals 1. Once the column adsorption sites are saturated, the bound Hg has a tendency to be leached off. In Figure 4 2, C/Co values greater tha n 1.0 occur when previously sorbed Hg leaches off the column. However over time this level decreases back to the influent concentration. Figure 42. Comparison of adsorption profiles of Hg onto ZVI and nZVI. The wastewater influent was spiked to produce a final total Hg concentration of 150 g Hg/L. The weight percent of iron particles within the sand matrix was 0.1% and 30% for nZVI and ZVI, respectively and corresponding to a specific surface area of 0.92m2. Column effluent fractions were taken over time, analyzed for total Hg concen tration and the result normalized to the influent Hg concentration. C/C0>1 indicates desorption of previously adsorbed Hg
80 The masses of iron particles used were selected to produce a similar exposed SSA of 0.92 m2 by either ZVI or nZVI. The influent was pumped through these columns at a flow rate of 50 L/min, and this value was determined based on preliminary laboratory tests (data not shown). The adsorption profile of nZVI can be divided into three different sections (Figure 42). The first section, day 1 through day 4, had a constant Hg concentration at around 40 g/L. From day 4 to day 6, the removal efficiency decreased slightly with a second plateau formed at ~65 g Hg/L. The last section is a linear increase in Hg concentration in the effluent leading to the break through at day 9. These trends s uggest that the higher rate of nZVI oxidation results in an early Hg removal of Hg and an apparent progressive saturation that leads to a breakthrough passed 8 days after the start of the experiment. On the other hand, column packed with ZVI show a rather poor removal efficiency in the first 4 days, prior to exhibiting a removal efficiency similar to that obtained with nZVI in the first 4 days T hen a quick breakthrough at about 9 days is observed, as well. Assuming that adsorption onto Feoxyhydroxide is t he main removal mechanism for both particle types; it appears that the rate at which these particles become coated with oxyhydroxide layers dictates their temporal efficiencies. Blowes et al (1997) showed that when remediating Cr(VI), the particle size m atters as they compared iron filings (0.5 1.0 mm) to iron chips (1.0 5.0 mm). The fillings were more efficient. In this study, nZVI had a higher Hg loading when expressed per mass of sorbent basis (1791 and 7.821 g Hg/g for nZVI and ZVI, respectively). However, when Hg loading was expressed per SSA, an opposite trend was obtained, with 58.71 and 67.24 g Hg/m2 for nZVI and ZVI, respectively. From these column
81 studies, nZVI packedcolumns decreased the levels of Hg present in the effluent from 150 g Hg/L to ~30 g Hg/L while ZVI lowered Hg levels down to only 60 g Hg/L even though its Hg loading per SSA is higher than that of nZVI. Effect of Nano Zero Valent Iron M ass on Mercury R emoval Figure 42 showed t hat nZVI has the potential of removing large amounts of Hg based on mass of adsorbent used. Therefore, increasing the mass of nZVI in the reactive mixture should theoretically increase the amount of Hg adsorbed. Figure 43 shows Hg removal profiles in columns packed with different masses of nZVI (0.1% or 0.025 g and 1.0% or 0.250 g). In Figure 43 C/Co values greater than 1.0 occur when previously sorbed Hg leaches off the column as explained previously. Time (days) 0 5 10 15 20 C/Co 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 nZVI 0.1% nZVI 1.0% Breakthrough and Desorption Figure 43. Trends of Hg removal from wastewater influent containing150 g Hg/L in columns packed with nZVI at 0.1% and 1% per mass basis The influent was pumped up gradient at 50 L/min. Column effluent fractions were taken over time, analyzed for total Hg concentration and the result normalized to the influent Hg concentration. C/C0>1 indicates desorption of previously adsorbed Hg
82 Increasing the mass of nZVI does increase the total amount of Hg adsorbed, and the breakthrough point is shifted from day 9 to day 14 (Figure 43). The total mass of Hg adsorbed was 44.77 and 84.03 g Hg for columns with for 0.1% and 1.0% nZVI, respectively. When expressed per SSA basis, the average amount of Hg removed by columns containing 1.0% nZVI was lower (11.02 g Hg /m2) than the Hg mass per SSA in the 0.1% nZVI column. Alowitz et al. (2002) obtained a similar trend with Cr(VI) adsorption studies, and where they showed a lack of correlation between metal removal and SSA (Alowitz and Scherer 2002) This observation suggests there are Hg fractions in solution not available for adsorption. One possibility is that Hg complexed to organic ligands could become unavailable for adsorption on Feoxyh yd roxide layers, therefore, limiting the removal capacity of the iron particles. If true, the oxidation of dissolved organic matter present in the sample could help improve the removal efficiency which is investigated in a latter section. Effect of F low R ate on Mercury R emoval Mercury adsorption was investigated under flow rate conditions of 25, 50 and 75 L/min. The results are presented in Figure 44 ; C/Co values greater tha n 1.0 occur when previously sorbed Hg leaches off the column. As the flow rate is increased from 25 to 75 L/min, the Hg loading increases and the lifetime of the column increases as w ell. However when a flow rate of 500 L/min was used (data not shown), no Hg removal occurred. This trend is opposite to some of the findings reported in the literature, in that decreasing flow rates increased removal efficiencies of ZVI for uranium (Morrison et al. 2001) and Hg (Weisener et al. 2005)
83 This discrepancy probably stems from a number of factors. First, the particle size used, and second, the amount of particles and column packing methods (100% ZVI versus reactive medium mixing sand and a small fraction of nZVI). Time (days) 0 2 4 6 8 10 12 14 16 C/Co 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 25 ul/min 50 ul / min 75 ul / min Breakthrough and Desorption Figure 44. Effect of flow rates on adsorption profile of nZVI. Influent s are Hg spiked wastewater effluent (150 g Hg/L). The weight percent of reactive media used was 0.1% nZVI. The influent was pumped up gradient at different flow rates (25, 50, 75 l/min). Column effluent fractions were taken over time, analyzed for total Hg concentration and the result normalized to the influent Hg concentration. C/C0>1 indicates desorption of previously adsorbed Hg In fact, for the mixture of sand and nZVI used in this study, the excess sand might cause the so called null pathways to fo rm in the column due to differences in particle size between nZVI and sand (Wilkin et al. 2005a; Wilkin et al. 2005b) Fig ure 45, shows a n idealized vertical cross section of a mixed reactive media. In this case, t he white bigger circles represent sand particles and the smaller black circles represent nZVI particles Two differ ent pathways are shown in red. Solution in pathw ay 1 comes into
84 contact with more nZVI particles than in pathway 2, which can be qualified as null pathway For adsorption to take place, the Hg needs to come into direct contact with the nZVI particles I t is possible that at very low flow rates throug h the column, null pathways are formed and consequently Hg is not adsorbed. However as the flow rate is increased Hg removal increases as well, which suggests a more uniform flow characteristics through the column and very little to no null pathway. 2 1 Figure 45. Cross sectional representation of the columns reactive media Hydrogen Peroxide T reatment of the I nfluent and O xidation of D issolved O rganic C ompounds to Improve Mercury R emoval from S olution The influent used in this study contained 8 mg /L of dissolved organic carbon (see Table 41). Accordingly, the high affinity of Hg for organic ligands may hinder the removal of Hg and limit the efficiency of these columns to levels much higher than the ng/L concentration range targeted by regulation agencies This is likely one of the
85 reasons explaining the fact that Hg concentrations in the treated effluents never dropped down to low ng/L levels. Therefore, treating the influent with an oxidizing agent such as H2O2 (used at a final concentration of 1% ( v/v) in this study) could help release Hg previously bound to dissolved organic matter (DOM). Time (days) 0 2 4 6 8 10 12 14 16 C/Co 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 Untreated Influent Influent treated with 1% Hydrogen Peroxide Breakthrough and Desorption Figure 46. Effect of influent pretreatment on adsorption profile of n ZVI. Influent for column was Hg spiked wastewater eff luent (150 g Hg/L). The influent was treated with 1% hydrogen peroxide to degrade any Hg bound to organic matter. The weight percent of reactive media used was 0.1% nZVI. The influent was pumped up gradient at 50 l/min. Column effluent fractions were tak en over time, analyzed for total Hg concentration and the result normalized to the influent Hg concentration. C/C0>1 indicates desorption of previously adsorbed Hg Using such an approach and running the influent through a nZVI packed column, the removal pr ofile shown in Figure 46 was obtained. The H2O2treated influent resulted in a steady concentration of 35 g Hg/L in the effluent, and a significant delay of the breakthrough (>30 days) as compared to nontreated influent However, this
86 treatment did not improve the column efficiency in terms of lowering H g levels in the effluent to the1 2 to 15 ng/L levels Th is could come from the inefficiency of the oxidizing agent at used concentration. In fact, t he analysis of DOC in both untreated and treated influent s showed only a slight drop from the initial 8 mg C/L to about 6 mg C/L in the untreated and treated influe nts respectively (Figure 47). Influent Untreated Influent Treated Effluent DOC (mg/L) 0 2 4 6 8 10 Figure 47. Effect of hydrogen peroxide (H2O2) treatment on dissolved organic matter present in wastewater used as influent in column studies. This shows that the 1% H2O2 treatment used was not adequate to obtain a complete oxidation of DOM present in the effluent. A higher concentration of H2O2 or a much stronger oxidizing agent w ould have been more appropriate, and unfortunately, this was not accomplished in this study. One possibility for the poor DOM oxidation result could be the complex structure of organic matter present in used influent. A literature review published by Shon et al. (2006), and focusing on the type of organic matter present in biologically treated wastewater efflue nts identified a wide variety of
87 organic compounds (Figure 4 8). As shown in this figure, DOC present in such WW can contain recalcitrant compounds such as humic acid s, which may strongly bind with metals while remaining hard to oxidize. The influent and effluent solutions used in this study were further characterized by fluorescence excitationemission spectrometry. This technique scans the sample to produce an excitationemission matrix (EEM). Figure 48. Organic constituents present in wastewater effl uents (Shon et al. 2006) Figure 49. Excitation emission peaks of dominant soluble organic matter compounds based on a literature review published by Chen et al. (2003)
88 The matrix consisted of emission wavelengths from 250 nm to 550 nm at 2 nm intervals and excitation wavelengths from 220 nm to 400 nm at 5 nm intervals with a scan speed of 15 00 nm/min. In general, data obtained from EEM studies are graphed in three dimension plot s and wh ere the y axis represents the ex citation wavelength, the x axis represents emission wavelength, and the z axis represents the intensity of the peak at that specific excitation and emission point. Using this technique, Chen et al. (2003) mapped different types of organic matter into diff eren t spectral regions (Figure 49) This graph is subdivided into five regions based on the most dominant organic fractions as follows: Region s I and II: a romatic proteins, boundaries are at shorter excitation wavelength <250 nm and shorter emission wavelength <350 nm Region III: fulvic acid like organics, boundaries are at shorter excitation wavelength <250 nm and longer emission wavelength >350 nm Region IV: s oluble microbial by product, boundaries are at intermediate excitation wavelength 250280 nm an d shorter emission wavelength <380 nm Region V: h umic acid like organics, boundaries are at longer excitation wavelength >280 nm and longer emission wavelength >380 nm This map can be used as a guide to distinguish the type of organic carbon present in the sample by matching peaks in the different regions shown above. Figure 410 gives the EEM of the influent without treatment (A), H2O2treated influent (B), and that of the column effluent from treated influent (C) From these figures two m ain observations can be made. First, peaks corresponding to 2 to 3 major organic compound types are present with the same intensity in both the treated and untreated water influent s. Second, compounds giving the dominant peaks can be roughly assigned to r egions I, III, an d IV (e.g. aromatic proteins, fulvic acids, and soluble microbial by p roducts) IV.
89 Figure 410. Excitation emission spectra to determine dominant organic species present. ( A ) column influent untreated, (B) column influent treat ed w ith 1% hydrogen peroxide, and (C) column effluent. The x axis is the emission wavelength (nm), the y axis is the excitation wavelength (nm) and the z axis is the peak intensity. ( A ) (B) (C)
90 The used H2O2 treatment was n therefore n ot strong enough t o break down these organic compounds. As observed in Figure 410, the influent did have some minor peaks in region II, III, IV, and V which were reduced after the H2O2 treatment However, the extent of the dominant peaks remained the same. This is a likely explanation as to why Hg removal efficiency did not increase after H2O2 treatment. However, the observed delayed breakthrough (Figure 46) on the other hand suggests that H2O2 treatment some beneficial effects including a contribution to the oxidation of iron particle surfaces, and therefore the increase in Hg adsorption capacity Unfortunately, Hg bound to complex DOM compounds remained in solution and were able to pass through the column unr etained. The potential of H2O2 to decrease total Hg levels of the influent by volatilization prior to pumping through the column was also investigated. This is because H2O2 has been linked to the volatilization of Hg in natural waters (Amyot et al. 1997; Amyot et al. 1994; Schroeder et al. 1992) E xperiments to verif y this possibility were conducted over a 20day time period. 25 mL of WW was treated with 1% H2O2 was added to 50 mL headspace vials The vials were capped and allowed to incubate for 20 days, after which 100 L headspace gas samples were analyzed for Hg0. Levels of Hg0 in the corresponding aqueous phase ( 0 aqHg ) were calculated using Henrys Law. This approach allowed for the determination of the total amount of Hg0 ( 0 THg ) produced in the reactor (i.e. 000 TaqgasHg=Hg+Hg) Figure 411, shows the e xtent of Hg volatilization from WW samples untreated and treated with H2O2.
91 ng Hg Volatilized 0 20 40 60 80 100 120 WW WW + 1%H 2 O 2 Figure 411. Volatilization of Hg in wastewater effluent (W W) and WW treated with 1% hydrogen peroxide (H2O2) with an incubation time of 20 days in closed headspace vials. Levels of Hg0 in the corresponding aqueous phase ( 0 aqHg ) were calculated using Henrys Law. This approach allowed for the det ermination of the total amount of Hg0 ( 0 THg ) produced in the reactor at any sampling time (i.e. 000 TaqgasHg=Hg+Hg ) Y axis represents the mass of total Hg0 produced in ng. WW samples treated with 1% H2O2 volatilized more Hg, 91 ng H g0, tha n untreated WW samples, 39 ng Hg0, over the 20 day incubation period. This was about 2.42% and 1%, of the total Hg in the system, which was 3750 ng Hg, for treated and untreated samples, respectively. These results suggest that H2O2 induced Hg loss by volatilization could be considered negligible within the time frame used for column studies described herein. Mercury R emoval U sing C olumn in S eries and E ffect of Z inc and C admium as C ompe ti tive C ations The above tested c hanges in nZVI mass, influent flow rate s, and influent oxidation with H2O2 have decreased the effluent concentration of Hg but not down to the acceptable low ng/L levels mentioned earlier To verify the fact that this Hg removal limitation was likely due to Hg speciation, two columns joined in series were used to
92 expose Hg exiting the first column to fresh nZVI surfaces and potentially decrease the concentration of Hg at the outlet of the second column. Figure 4 12 shows data comparing Hg trends in effluents from a single column vers us two columns used in series with all operating parameters being identical In Figure 412 C/Co values > 1.0 occur when previously sorbed Hg leaches off the column. Figure 412. Mercury removal from wastewater (WW) effluent in a single column and two col umns used in series. Used influent solution was spiked with Hg to a final concentration of 150 g Hg/L. The weight percent of reactive media used was 0.1% n ZVI. The influent was pumped up gradient at 50 L/min. Column effluent fractions were taken over tim e, analyzed for total Hg concentration and the results normalized to the influent Hg concentration. C/C0>1 indicates desorption of previously adsorbed Hg The first 4 days of the column in series brought Hg concentrations down to 10 g Hg/L. However this range increases over time and plateau around 45 g/L from day 4 to
93 day 9. The residual Hg not being adsorbed may be due to the unavailability of Hg and lack of interaction with oxidized iron particle surfaces. Additionally based on EEMs data presented in F igure 4 10, Hg is likely bound to aromatic proteins, fulvic acids, and soluble microbial byproducts. Since columns used in series seemed to slightly improve Hg removal from solution, this setup was used to test if the presence of competitive cations in in fluent water would compete with Hg for adsorption sites For that purpose, cadmium (Cd) and zinc (Zn) were chosen. Cd, Hg, and Zn belong to Group IIB in the periodic table of the elements and have a number of similar geochemical behavior, such as their str ong affinity for sulfur containing compounds and their classification on the Goldschmidts scale (Krauskopf and Bird 1995) In addition, they exhibit similar ionic radius and electronegativities (Table 42) (Jing et al. 2007; Krauskopf and Bird 1995) Table 42. Ionic radii and elec tronegativities (Krauskopf and Bird 1995) Element Ionic radius( ) Electronegativity Hg 1.02 1.9 Cd 0.95 1.7 Zn 0.74 1.7 Figure 413 shows how the addition of Cd and Zn affects the adsorption of Hg and C/Co values greater than 1.0 occur when previously sorbed Hg leaches off the column. These cati ons at the concentrations tested did not have a significant effect on the Hg adsorption onto nZVI. The slightly higher electronegativity of Hg as compared to Cd and Zn leads to a greater affinity of ionic Hg for negatively charged oxygen from iron oxyhyd roxide layers.
94 Figure 413. Effect of cadmium (Cd) and zinc (Zn), on the removal of Hg from a wastewater used as influent in a study using two 0.1% nZVI packed columns in series. Used influent solution was spiked with Hg to a final concentration of 150 g Hg/L. The weight percent of reactive media used was 0.1% n ZVI. The influent was pumped up gradient at 50 L/min. Column effluent fractions were taken over time, analyzed for total Hg concentration and the result normalized to the influent Hg concentrati on. C/C0>1 indicates desorption of previously adsorbed Hg S canning E lectron M icroscopy Characterization of the Reactive Media Scanning electron microscopy (SEM) was used to analyze the nZVI particles before and after Hg removal experiments. The purpose of this effort was to indentify the presence of Hg on the surface of nZVI particles used in the make up of the reactive media. SEM images of raw nZVI (Figure 414) show that the raw nZVI particles are dominantly aggregated The coupling of SEM to energy dispersive spectroscopy (EDS) analysis of the sample showed oxygen peak s (see red box), suggesting the presenc e of oxides on nZVI surfaces before they were treated with NaOH and used in Hg removal laboratory experiments. This could have contributed to the extent of Hg retention by
95 nZVI containing columns in the first few days of the Hg removal experiments (see Figure 42). Figure 414. (A) SEM image of pu re nZVI at 600x. (B) EDS spectrum of pure nZVI The presence of a large oxygen peak is highlighted by red b ox At the end of these Hg removal experiments, a fraction of the reactive media ( i.e. mixture of sand and nZVI) was sampled, freezedried and then analyzed by SEM EDS. Figure 415 shows the corresponding SEM images of materials obtained from the first col umn (column 1). Figure 415 A shows a sand particle with iron oxides on the surface.
96 Similarly, analysis of column 2 shows iron oxides on the surface (Figure 4 15B). Figure 4 16C shows a higher resolution of the framed red box where the iron oxides are form ed on the surface of the sand particles. Overall, since the dominant sand matrix was made of much larger particles than those of nZVI, even when aggregated, Fe is therefore detected onto those sand particles, which are acting as support materials. Figure 415. SEM image of the columns reactive media in series. (A) column 1 resolution of 400x, (B) column 2 resolution of 2500x ( A ) (B)
97 Figure 416. SEM image of the column s reactive media in series (C) red box zoomed in of Figure A at a resolution of 4000x EDS was done on the same particle s and the results are shown in Figure 417. The spectrum shows high peaks of silica and oxygen, and peaks for iron and aluminum. The silica as expected comes from the sand particles. The aluminum and chlorine may be from the WW as these substances are present in WW treatment. The iron intens ity is very small and comes from the nZVI. This technique was used to show that Hg is adsorbed on the surface. Due to sensitivity in the technique, Hg was not detected with 100% confidence. It is speculated that the peak highlighted in t he red box in Figure 417A, comes from Hg. The high intensity of silica and oxygen can produce peaks for a silica oxygen compound and when the energies are summed the peak should be seen at 2.26 keV. Sim ilarly the Hg peak should be seen at 2.20 keV. Therefore based on just this EDS spectrum it is unclear if the peak is from Hg or from silica oxygen. However, the SEM/EDS sample of colum n 2 in the series, Figure 417 B, did not show a peak in this region. Column 2 should have less Hg than column 1.Thus, it can be said that the peak from column 1 c omes from the Hg on the surface. To summarize column 1 ( C )
98 adsorbed av ailable Hg (peak in Figure 417A), while column 2 did not adsorb any Hg (no peak). This confirms that Hg bound organic compounds are unavail able for adsorption; consequently column 2 did remove any Hg. Figure 417. Energy dispersive spectra analysis of column 1 (A) and column 2 (B) in the column series Conclusion This study investigated the poss ibility of using the different interactions between zero valent iron particles and dissolved Hg towards the development of a technology for Hg removal from contaminated wastewater effluents. Results obtained from column
99 studies show that nZVI has the potential to adsorb Hg from aqueous solutions, but several factors may play a significant role in reaching an ideal efficiency in terms of Hg removal. Such factors include influent characteristics, the speciation of Hg in the solution to be treated, the flow r ate used to pump the Hg contaminated water through the column, the mass and probably the degree of aggregation of nanosized particles the pH of solution, and specific characteristics of the columns (length, width, single column, more than 1 column used i n series) The effect of particle size seems to be in favor of nZVI as compared to ZVI, mostly in terms of rate s of particle oxidation and early Hg removal by adsorption. Based on experimental conditions used in this study, over time nZVI not only becomes saturated, but also releases previously adsorbed Hg. With regard to the pumping flow rate, an increase from 25 to 75 L /min improved the removal efficiency and delayed the breakthrough time. However, when the flow rate was increased to values as high as 5 00 L /min no Hg adsorption occurred. This can be a serious limitation with regard to large scale application. Ideally, further studies should focus on the identification of ideal conditions for higher Hg removal efficiencies under conditions that can be easily scaled up for industrial application. Overall, the use of nZVI to remove Hg from aqueous effluents could be considered promising, but more studies are to obtain operating conditions that will decrease effluent Hg concentrations to the ideal ng/L range. Only then, effort should be invested in the evaluation of the potential for the regeneration of nZVI and its re use.
100 CHAPTER 5 IRON MERCURY INTERACTIONS : EFFECTS ON MERCURY BIOAVAILABILITY AND METHYLMERCURY PRODUCTION IN FRESHWATER S EDIMENTS Introduction The speciation and concentration of mercury (Hg) available to methylating bacteria are key variables for the rate of methylmercury ( MeHg ) production in sediments. The chemical speciation of Hg in aquatic systems is strongly influenced by several para meters including the redox potential and pH conditions as well as the concentrations of inorganic and organic complexing agents. Both Hg2+ and MeHg+ cations have a high tendency to form thermodynamically stable complexes, in particular with soft ligands such as reduced sulfur species (Ullrich et al. 2001). In the absence of reduced sulfur ligands the speciation of inorganic Hg in freshwaters seems to be dominated by three uncharged complexes : Hg(OH)2, HgOHCl, and HgCl2. With regard to organomercury speci es, thermodynamic calculations predict that CH3HgOH would be the most stable species in most freshwater environments, whereas CH3HgCl would dominate in seawater (Craig 1986; Stumm and Morgan 1996). T he mobility and bioavailabili ty of Hg species in aquati c and soil environments are affected by physicochemical processes such as complexation, precipitation, dissolution, and adsorption. In Hg impacted watersheds, organic matter is usually a very good predictor of Hg2+ in surface waters as dissolved organic m atter (DOM) increase s the solubility of Hg, mostly through coordination to thiol groups associated with complex organic ligands such as humic and fulvic substances. A n extensive review of factors controlling Hg methylation including redox conditions was published by Ulrich et al., (2001). Understanding the distribution of redox processes is essential for predicting the fate of certain environmental contaminants.
101 With regard to Hg, the dissolution of certain oxide minerals, primarily Mn and Fe oxyhydroxides in reducing environments releases Hg and other metals adsorbed onto them, thereby increasing Hg bioavailability to methylating microorganisms. This is due to their large surface area and high capacity to adsorb and coprecipitate Hg, and to rerelease it as redox conditions shift from oxic to anoxic (Fagerstr.T and Jernelov 1972). Many workers have found the distribution and concentration of dissolved and particulate Hg species to be influenced, among other factors, by the redox cycling of Fe, and less fre quently Mn (e.g. (Bloom et al. 1999; Bonzongo et al. 1996; Gagnon et al. 1997; Gobeil and Cossa 1993; Hurley et al. 1991; Mason and Sullivan 1999; Quemerais et al. 1998; Regnell et al. 1997). Bloom et al. (1999) reported, for example, that the mobility of MeHg in estuarine surface sediments was linked to the Fe redox cycle, while the mobility of Hg2+ was controlled by the formation of soluble polysulfide or organic complexes. The formation and dissolution of Fe and Mn oxides is strongly controlled by the re dox state and oxygen content of waters and sediments. In anoxic conditions, oxyhydroxides dissolve and release any associated Hg (Cossa and Gobeil 2000; Gagnon et al. 1997; Gobeil and Cossa 1993), which is thought to be one reason for the frequently observ ed Hg and M e Hg enrichment in seasonally anoxic waters (Cossa et al. 1994; Hurley et al. 1991; Watras et al. 1995). Seasonal and diurnal trends in MeHg concentrations in sediment pore waters may also be linked with redox effects (Covelli et al. 1999; Gill et al. 1999). Oxyhydroxides can also form labile complexes with organic matter and clay minerals and this has the potential to increase their metal scavenging capacity (Meili 1997).
102 T he production of MeHg in sediment s is controlled primarily by the availability of the different mineral and organic Hg fractions. The use of KOH as an extractant in the chemical Hg fractionation process is based on its ability to solubilized organic matter boundmetals. Similarly aqua regia is used to solubilize the Hg that is tightly bound to sulfide minerals. T he relationship between potential rates of MeHg production and certain sediment Hg fractions ( KOH and aqua r egia extracted) can be used as proxy for Hg availability to methylating agents For instance, as organic matter undergoes degradation in sediments, it releases previously boundHg, making it available to microorganisms Another important aspect of Hg bioavailability is its affinity with soft ligands, mostly reduced sulfur species R eduction of sulfate and precipitatio n of Hg sulfides can limit bioavailability. However, under specific laboratory conditions, it has been hypothesized that neutrally charged aqueous Hg sulfide species could become methylated by SRB (Benoit et al. 2001; Benoit et al. 1999) This latter observation is based on simply on geochemical equilibrium predictions and not the chemical analysis of different Hg species. In aquatic environments, zerovalent i ron particles can react with water under both aerobic and anaerobic conditions to produce the oxidized Fe2+ and Fe3+ species (Bierna t and Robins 1972; Gu et al. 1999; Kenneke and McCutcheon 2003; Majewski 2006; Sayles et al. 1997) These conditions would favor the formation of relatively stable i ron oxyhydroxide amorphous minerals such as FeOOH, FeOOH, FeOOH, Fe3O4, and FeCO3, (Farrell et al. 2000; Gu et al. 1999; Huang and Zhang 2005; Huang et al. 2003; Phillips et al. 2000; Rangsivek and Jekel 2005; Ritter et al. 2003) While the formation of such oxyhydroxides Fe minerals would likely increase the adsorption of
103 dissolved Hg2+, the release of dissolved Fe2+ ions under predominant anoxic conditions could increase the availability of Hg through the removal of diss olved sulfide through formation of pyrite (FeS) minerals and reduction of Hg precipitation as HgS minerals. The balance between these different processes would be dictated by key environmental parameters such as pH, type and quantity of dissolved organic m atter, and redox conditions. In this study, the Hg methylation potential of anoxic sediments in the presence of zero valent iron particles w as investigated with the goal to determine if Hg naturally occurring or added to sediments could be made available t o methylating microorganisms. Sulfate has also been identified as the primary parameter controlling the methylation of Hg in aquatic systems (Ulrich et al., 2001). Therefore, experiments were designed to take into account the potential effects of sulfate c oncentrations. M aterials and M ethods Sediment and Water Samples Used in Hg Methylation Studies Sediment and water samples were collected from the East Fork Poplar Creek watershed (latitude 35.990683 and longitude 84.317183) near Oak Ridge Tennessee, USA. Figure 51 shows the geographical location of the sampling site, labeled by a star. This site has been on US EPA Superfund list since 1989 and is known to be heavily contaminated with Hg (Dong et al. 2009; Miller et al. 2009) Water samples were collected in large acid precleaned high density polyethylene (HDPE) containers using ultraclean sampling techniques for trace metal sample collection. Sediment samples were collected after removing the surface layer of decomposing litter and scooping of exposed sediment s.
104 Figure 51. Map of East Fork Poplar Creek in Big Turtle Park Greenway in Oakridge,TN. Sediment and water samples were collected from a noted by a star on the map (latitude 35. 990683 and longitude 84.317183 in decimal degrees). Sediment samples were placed into clean plastic bags. Both water and sediment samples were kept in large coolers packed with ice, and the load transported back to our laboratory at the University of Florida. Once back to the laboratory, the sediment was sieved through a 2 .0 mm mesh to remove large plant and rock debris while the water was filtered through a 0.45 m filter. Both types of samples were then stored refrigerated at 4oC pending use in different laboratory analyses. Ion chromatography (IC), inductively coupled plasma atomic emission spectroscopy (ICP AES), and loss on ignition techniques were used to det ermine the concentrations of major ions, metals other than Hg, and the organic content of sediment samples Although important the analysis of the data, the physical characteristics of used sediments (size fraction and mineralogy) were not determined due t o limited access to analytical equipment. Determination of Total and Methyl mercury Sediment s samples placed in Teflon tubes were first digested by aqua regia (HCl and NO3 in a 3:1 ratio volume/volume) at sub boiling temperature for about 12 hours (overn ight digestion). Cold acid digestion was used for aqueous samples using bromine monochloride. In both cases, digested samples were analyzed by the SnCl2reduction
105 technique, followed by detection with cold vapor atomic fluorescence spectrometry ( CV AFS, Te kran Model 2600). Details on used analytical procedures and QA/QC criteria (blanks and Hg standard solutions) have been described in several of our earlier publications (Hovsepyan and Bonzongo 2009; Warner et al. 2005; Warner et al. 2004) MeHg present in sediment samples was determined using methods published in the literature (Bloom 1989) Sediment samples (~1g) were first digested with 10 mL of 25% KOH in methanol at 70oC for 12 hours. This step is used for the solubilization of organic compounds present in sediment including MeHg. Next, the obtained extract is diluted to a known volume (25 mL) of which an aliquot was used for MeHg determination. Prior to analysis, the pH of the aliquot sample was buffered to 4. 5 using an acetic acid buffer The sample was then treated 100 L, of sodium tetraethylborate to convert Hg compounds present in the samples to highly volatile species as shown in equations 51 and 52 Through these reactions, ionic Hg present in the sampl e becomes converted to di ethyl mercury (Eq. 5 1) while MeHg cation becomes ethylated to form methyl ethyl Hg (Eq. 5 2). 2+ + 254 2532522NaB(CH)+Hg2B(CH)+Hg(CH)+2Na (5 1) ++ 2543 253325NaB(CH)+CHHgB(CH)+CHHgCH+Na (5 2) This approach allows for increased volatilization of these organomerc uric compounds. Sampling of the headspace after equilibration and species separation by gas chromatography and pyrolysis allowed for the detection of atomic Hg by CV AFS. Speciation of Solid Phase Mercury in Collected Sediment Samples The selective Hg e xtraction method specific for biogeochemically relevant Hg fractions in solid materials developed by Bloom et al., (2003) was used in this study.
106 The method identifies five different Hg fractions based on the type of chemical extractant used during a sequential procedure. The different chemical solutions used as extractants as well as the corresponding Hg fractions and time of extraction are shown in Table 51. Table 51. Selective extraction fractions of mercury. Adapted from Bloom et al., 2003 Targeted G eochemical Fractions Extractant Solutions (V=30 mL) Extraction Time (Hours) F1: Easily exchangeable Hg DI water 2 F2: Humic stomach acid soluble 0.1MCH 3 COOH and 0.01MHCl 4 F3: Organo chelated 1M KOH 6 F4: Elemental Hg 12M HNO 3 8 F5: Mercuric sulfi de Aqua Regia (10:3 ratio of HCl:HNO 3 ) 12 At the end of each shaking step, except for step number 5, sediment samples were centrifuged for 30 min at 3000 rpm and the supernatant removed and placed in acid p re cleaned Teflon containers. Next, the solid pellet was rinsed with digestion solution, centrifuged, and the c ollected supernatants combined. The latter was then filtered through a 0.45 m filter and treated with bromine monochloride, a strong oxidizing solution, to help maintain the extracted Hg into solution pending analysis. Investigation of the Effect of Changing Sulfate Concentrations on the Biotransformation of Mercury in Sediment Slurries Containing Fixed Iron Masses In this series of laboratory experiments, microcosm studies were conducted t o investigate the interaction of iron and Hg already present in sediments (i.e. no Hg spikes) on the bioavailability and biotransformation of sediment Hg. Serum vials (50mL) pre cleaned by soaking in 10% trace metal grade HNO3 for 24 hours and thoroughly rinsed with Nanopure water were used as reactors for batch incubation studies. Sediment slurries were prepared by adding 10 mL of site water to 1 g of sediment
107 following a procedure widely reported in the literature on methyl Hg production studies (Slowey and Brown 2007; Ullrich et al. 2001) The above mixture was first allowed to equilibrate for 24 hours and a preweighed and fixed mass of ZVI (0.2g) or nZVI (0.01) was added to the sediment slurries. Next, sulfate was added in increasing concentrations (200, 500, 1000 M) to produce three different Fe/sulfate ratios. Sulfate is used by sulfate reducing bacteria (SRB) as terminal electron acceptor during anaerobic respiration, and MeHg production in anoxic environments has been found linked primarily to the activity of S RB (Ulrich et al., 2001). Experimental parameters used in this study are summarized in Table 52. Slurries were next deaerated by bubbling with ultrapure nitrogen, the vials hermetically sealed, and the mixture left to incubate for 25 days. Note that no H g was added to these vials as these experiments focused on the biotransformation of Hg already present in sediment as a function of changing Fe/SO4 ratios. At the end of the incubation period, 2 mL of slurries were withdrawn and processed for the analysis of both MeHg and total Hg concentrations. Table 52. Experimental Design for Hg Methylation in Sediment slurries containing Fixed Iron Amounts and Changing Sulfate Concentrations Sediment (g ) Site Water (mL ) Sulfate Conc. Added ( M) Iron Added (g) Contr ol 1 1 .0 10 0.0 No ne Control 2 : ZVI 1 .0 10 0.0 0.20 Control 2 : nZVI 1.0 10 0.0 0.01 Control 3 1 .0 10 500 No ne Treatment 1 : ZVI 1 .0 10 200 0.20 Treatment 1 : nZVI 1.0 10 200 0.01 Treatment 2 : ZVI 1.0 10 500 0.20 Treatment 2 : nZVI 1 .0 10 500 0.01 Treatment 3 : ZVI 1 .0 10 1000 0.20 Treatment 3 : nZVI 1 .0 10 1000 0.01 Treatment 4 : ZVI 1 .0 10 500 2.60
108 Investigation of the Effect of Changing Iron Masses on the Biotransformation of Mercury in Sediment Slurries Containing Sulfate Concentrations Us ing a procedure similar to the one described above, batch experiments were conducted, and this time, the mass of iron added to the sediment slurries varie d while the concentration of sul fate was maintained constant. The experimental conditions are summariz ed in Table 5 3. Table 53. Experimental Design for Studies on Hg Methylation in Sediment Slurries with Fixed Sulfate Concentrations and Changing Iron Masses Sulfate Conc. Added ( M ) Mass of Iron (g) Iron 1: ZVI 1000 0.500 Iron 1: nZVI 1000 0.020 Iron 2: ZVI 1000 0.100 Iron 2: nZVI 1000 0.005 Effect of Iron on the Biotransformation of Newly Added Hg into Sediment s In contrast to the above experiments focusing on in situ Hg, this second set of incubations assessed the effect of ironmercury interactions on the availability of Hg newly added to sediments using microbial Hg methylation as surrogate for bioavailability. The experimental approach was similar to the one described a bove, except that in this specific case, sediment slurries were spiked with Hg (using a 1000mg/Lstock sol ution of Hg(NO3)2 from Fisher Scientific, USA ) to produce a final slurry concentration of 1.0 mg/L. Experimental parameters are presented in Table 54 Table 54. Experimental Design for Hg Methylation Studies in Sediment Slurries Spiked with Hg(NO3)2 in Addition to Iron and Sulfate Sediment (g ) Site Water (mL ) Sulfate Conc. Added ( M) Hg Conc. Added (mg/L) Iron Added (g) Control 1 1 .0 10 0 1.0 no n e Control 2 1 .0 10 200 1.0 none Control 3 1 .0 10 1000 1.0 no ne Treatment 5: ZVI 1 .0 10 200 1.0 0.20
109 Table 54. Continued. Sediment (g ) Site Water (mL ) Sulfate Conc. Added ( M) Hg Conc. Added (mg/L) Iron Added (g) Treatment 5: nZVI 1 .0 10 200 1.0 0.0 1 Treatment 6: ZVI 1 .0 10 1000 1.0 0.20 Treatment 6: nZVI 1 .0 10 1000 1.0 0.01 Statistical Analysis To evaluate the significance of differences observed between treatments, a simple t test for equal or unequal variance was used at a confidence level of 95%. The calculations were performed using Microsoft Office Excel 2003 statistical data analysis tools see Appendix A Results The site water used in these laboratory experiments had a pH of 7.02 and a total Hg concentration of 140 g Hg/L (or ppb), a rather high dissolved Hg concentration indicative of the contamination level of the creek under study. This water contained an average DOC concentration of 7.2 mg C/L and a major ion composition listed in Table 5 6. Used sediments had a pH of 7.46, and a total Hg concentration of 11 mg Hg/kg (or ppm). The sediments organic content was about 8% as assessed by loss on ignition. Metal and major ion data for both water and sediment used are presented in Tables 55 and 56. Based on these data, the concentration of sulfate naturally present in used sediment slurries was approximately 0.18 M, based on the mixture of 1 g of sediment and 10 mL of site water.
110 Table 55. Average metal concentrations found in sediment and site water determined by ICP AES. Metals Sediment ( mg/g ) Site Water ( mg/L ) A luminum (Al) 3.026 0 .117 A rsenic (As) 0.002 0. 003 C alcium (Ca) 0.698 29.341 C obalt (Co) 0.004 0.003 I ron (Fe) 6.562 0.194 P otassium (K) 0.235 2.702 M agnesium (Mg) 0.515 8.059 M anganese (Mn) 0.44 0 0.004 Sodium (Na) <0.033 2.093 Table 56. Average major ions found in sediment (water soluble fraction) and used natural water, analyzed by ion chromatography (ND = not determined). Ions Sediment ( water soluble fraction ) ( mg/g ) Site water ( mg/L ) Chloride Cl 0.11 3.93 Sulfate, 2 4SO 0.07 9.85 Nitrate 3NO 0.05 2.13 Sodium Na + N D 2.94 Potassium K + ND 1.34 Calcium Ca 2+ ND 22.84 Accordingly, sulfate concentrations used to spike the sediment slurries resulted in a significant increase (200, 500, and 1000 M), and this choice was based on reports from the l iterature on sulfate concentrati ons that stimulate Hg methylati on in sediments of different aquatic systems (Chen et al. 1997; Gilmour and Henry 1991; Harmon et al. 2004; Harmon et al. 2007; Mehrotra and Sedlak 2005) Figure 52 shows the results of the effect of varying sulfate concentrations in sediment slurries with fixed amounts of iron particles. No significant differences were observed between values obtained from control samples and sulfatetreated slurries at the 95% confidence level. Additionally, there was no significant difference between the samples containing ZVI and nZV I.
111 The measured specific surface area (SSA) was 0.116 m2/g and 30.5 m2/g for ZVI and nZVI, respectively. Based on the total mass of either ZVI or nZVI used in these experiments, an estimated SSA 0.0232 m2 and 0.305 m2 could be calculated for the two types of iron particles respectively. There was no significant difference between the MeHg produced when equal surface area was used for ZVI and nZVI. C C (Trt 2) C (iron) Trt 1 Trt 2 Trt 3 ug MeHg/g 0.0 0.5 1.0 1.5 2.0 2.5 no iron ZVI nZVI Figure 52 Methyl mercury produced in Hg contaminated sediments with sulfate addition after 25 days of incu bation. Control: sediments only, C(Trt 2): sediments + 500 M sulfate, C(iron): sediments + iron only, Trt 1: sediments + 200 M sulfate added + iron, Trt2: sediments + 500 M sulfate added + iron, Trt 3: sediments + 1000 M sulfate added + iron
112 ZVI nZVI ug MeHg/g 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 Figure 53. Effect of partic le size with a sulfate concentration of 500 M treatment added to sediments. The mass of ZVI used matched the surface area of the mass of nZVI used. Varying the mass of iron particles in sediment slurries with sulfate concentration fixed at 1000 M (or 1mM) gave the results presented in Figure 54. For ZVI, as the iron particle mass decreased the amount of MeHg produced decreased as well. In the presence of nZVI,a bell shaped trend was observed, in that slurries spiked with the intermediate tested mass on nZVI (0.01g) exhibited the highest MeHg production. The highest mass of iron particles, (0.2 g), had a significantly lower MeHg production than both slurries with 0.01 g and 0.005 g.
113 Mass of nZVI (grams) ug MeHg/g 0.0 0.5 1.0 1.5 2.0 0.5 0.2 0.1 0.02 0.01 0.005 Mass of ZVI (grams)* *ZVI nZVI* Figure 54. Effect of different iro n masses on MeHg production in sediment slurries containing 1 mM of sulfate. (*) represents significant differences at 95% confidence level The results obtained from the above experiments show no to very little impact of iron particles on the transformat ion of Hg in sediments. One reason could be the limited bioavailability of Hg naturally occurring in used sediments. The speciation of solid phase Hg in used sediments based on the sequential extraction technique described in the method section showed that Hg in used sediments is present predominantly in the non available form, including up to 68% of total Hg found in the organic bound fraction and another 14 to 15% found in fractions which are not readily bioavailable (Bloom et al., 2003). It was therefor e assumed that spiking the sediments with a known amount of Hg would increase the amount of its bio available fraction and therefore MeHg production Under such conditions, a better assessment of the potential ironmercury interactions on
114 MeHg production i n sediments could be made. Therefore, experiments similar to treatments 1 (200 M sulfate added) and 3 (1000 M sulfate added) shown in Figure 52 were used, except that in this case, sediment slurries were supplemented with Hg salt to a final concentrati on of 1 mg Hg/L. The obtained results are shown in Figure 55 in comparison with control non spiked s lurrie s. In ZVI containing slurries, no significant difference in levels of MeHg produced was observed between the Hg spiked and nonspike H g sediments H owever in nZVI treated slurries, a significant difference in the amount of MeHg produced was observed for slurries containing 1 mM of sulfate. Figure 55. Methyl Hg produced in sediment slurries spiked and nonspiked with Hg and containing a fixed amount (0.02 g for ZVI and 0.01 g for nZVI) of iron particles but different sulfate concentrations (200 and 1000 m M). (*) represents significant difference at 95% confidence.
115 Discussion Based on general knowledge of the biogeochemistry of Hg in aquatic systems and results on Hg iron interactions presented in earlier chapters, at least three key parameters can help explain MeHg trends or lack of, seen in this study. First, the speciation of Hg naturally present in used sediments and its impact on the bioavailabi lity and transformation processes that lead to MeHg production. Second, used experimental conditions (e.g. ratios of used components and redox conditions). Third, the potential impact of cooccurring methylation and demethylation processes and the signific ance of their potential rates. Speciation of Mercury and its Methylation in Aquatic Sediments The literature on the methylation of Hg in sediment is quite abundant. Although Hg methylation by chemical reactions not catalyzed by nonmicrobial processes (E q. 5 3,4,5), its relative contribution to the overall pool of produced MeHg has been estimated to be negligible based on conditions required for such reactions to take place under most common natural conditions (Celo et al. 2006; Craig 1986; Gardfeldt et al. 2003; Gilmour and Henry 1991; Monperrus et al. 2007; Weber 1993) 2+ + + 22 222MeCo(dmg)HO+HgMeHg+MeCo(dmg)(HO) (5 3) 32 MeSn(IV)+Hg(II)MeSn(IV)+MeHg(II) (5 4) 2+ ++ 2Hg+MeIMeHg+HgI (5 5) In contrast, biotic methylation of Hg is known to dominate in aquatic sediments and sediment s, particularly when conditions for microbial reduction of sulfate by SRB are prevalent. However, the exact mechanisms of Hg methylation are still not fully understood (Beijer and Jernelov 1979; Choi et al. 1994; Jay et al. 2002; Jensen and
116 Jernelov 1969; Pak and Bartha 1998) Besides Hg methylation by SRB, recent studies have als o linked iron reducing bacteria (IRB) to MeHg production (Fleming et al. 2006; Han et al. 2008; K erin et al. 2006) Regardless of the pathway, the biotic or abiotic conversion of inorganic Hg species to MeHg depends primarily on Hg speciation, which controls Hg availability to methylating agents. The results of Hg speciation in used sediments are pr esented in Figure 57. Fractions F1 F2 F3 F4 F5 % of Total Hg 0 20 40 60 80 100 Figure 56 Percentage of Hg associated with different sediment fractions based on sequential selective extractions using a method adapted from Bloom et al. (2003). F1: Easily exchangeable Hg; F2: humic stomach acid soluble; F3: organo chelated; F4: elemental Hg; and F5: mercuric sulfide. From the results presented in Figure 56 Hg speciation in used sediments is primarily associated with the organic fraction (68%). The remaining fractions F1, F2, F4, and F5 accounted for 16%, 9%, 6% and <1%, respectively. Based on this Hg distribution among the different sediment organic and inorganic fractions, the extent of MeHg production is likely limited primarily due to the nonavailability of sediment Hg. Complexation by sediment organic matter would therefore constitute a key limitation to
117 Hg bio methylation. This observation would support the result trend shown in Figure 52. In fact, it appears that the Hg binding capacity of used sediments is so hi gh that the use of Hg spiked sediments produced MeHg levels that were similar in both control and Hg spiked sediment slurries containing either ZVI or nZVI, and regardless of sulfate concentrations. Effect of Specific Experimental Parameters Sulfate Sulfa te concentrations can either inhibit or enhance MeHg production (Compeau and Bartha 1985; Gilmour and Henry 1991) Figure 57 shows a theoretical ly determined range for Hg methylation as a function of sulfate concentration with a potential optimal Hg methylation supported by sulfate concentrations ranging from 200 to 500 M. However, the optimum response within the above range is also dependent on a number other parameters such as pH, temperature, sediment porosity, and organic carbon content Above this range it is hypothesized that reduction of sulfate will produce reduc ed sulfur species which on a molar ratio basis would likely precipitate most of the soluble ionic Hg as mercury sulfide. In contrast, s ulfate concentrations <200 M would limit SRB activity and therefore Hg methylation (Gilmour and Henry, 1991) Figure 57 Sulfate concentration range for optimal mercury methylation rates in sediments (Gilmour and Henry 1991)
118 Published experimental data on the other hand do not necessarily agree with the above theoretical predictions, and due to the synergic effect of several environmental factors on the methylation pr ocess, the observed response to sulfate concentrations varies from site to site. For example, high MeHg production rates were found sulfaterich salt marsh sediments (Compeau and Bartha 1984) low sulfate concentrations (about 92.9 M) in a wetland microcosm study produced more MeHg than the highest tested sulfate concentration of 482 M (Harmon et al. 2004) Trends observed in this study follow those reported by Hamon et al (2004). At the lowest tested sulfate concentration (200 M), for both ZVI and nZVI treated slurries, the amount of MeHg produced was the highest observed ( Figure 52). While at the highest tested sulfate concentration (1000 M), both ZVI and nZVI treated slurries produced low MeHg (Figure 5 2). This may be due to the formation of mercuric sulfide based on the hy pothesis presented in Figure 57 as a sulfate concentration of 1000 M is outside the range of values that favor Hg methylation (Gilmour and Henry, 1991). Organic carbon It is believed that low rates of Hg methylation in systems containing reasonable sulfate levels (i.e. levels that stimulate Hg methylation) could be attributable to the nonavailability of Hg to methylating mic r oorganisms, and the strong association of Hg to sediment organic matter explains, at least partly, measured low Hg methylation rates (Harmon et al. 2004) The results of solid phase Hg speciation conducted in this study showed that approximately 68% of total Hg present in used sediments was organochelated, and therefore not directly available to microorganisms (Frac tion 3, Table 51 and Figure 56 ).
119 Effect of iron addition to sediment slurries Tables 55 and 56 show that Fe and sulfate do occur in measurable concentrations in both sediment and water samples used in this study. The use of iron particles would therefore increase Fe(II) levels in the slurries through reduction and dissolution of initially oxidized iron at the surface of the particles. Under sulf ate reducing conditions, the presence of reduced iron could lead to the formation of iron sulfides, which in turn can coprecipitate Hg (Gilmour and Henry 1991) Reactions between water and iron particles (see chapter 2) and the interaction of Hg with such particles (chapter 3) show that Hg could be rem oved from solution through sorption onto iron particles as oxyhydroxides form on the surface of the iron particles and decrease Hg availability (Weisener et al. 2005) However, this formation is rather slower under anoxic conditions than in the presence of oxygen. An incubation time of 25 days used in this study may have given the iron particle enough time to undergo oxidation and then reduction, reactions which may help explain why sediment slurries with iron, sulfate, and no Hg addition had rather similar amount of produced MeHg (Figure 52). The difference in the size of particles used (Figures 52 and 53) had apparently no effect on MeHg production. At the highest tested concentration of nZVI, the smallest amount of MeHg was produced while the highest produced amount of MeHg was measured in sediment slurries spiked with 0.01 g nZVI. The different trends seen between ZVI and nZVI, (Figure 54), can only be explained by the differenc e in SSA of used iron particles. However, the exact mechanisms as to why is not clear. Methylation and Demethylation MeHg produced i n sediments is a net product of cooccurring Hg methylation and MeHg demethylation reactions. Measurements of potential rates of Hg methylation and
120 MeHg demethylation rates were beyond the scope of this study Never theless, these potential rates commonly used to identify the tendency for a given sediment to produce and accumulate MeHg. For instance, when rates of Hg methylation (M) and MeHg demethylation (D) are equal, M/D would be equal to 1 and MeHg would not accumulate to measurable levels. Only when M/D>1 that MeHg builds up in the environment. Unfortunately, the accurate determination of these rates requires the use of stable Hg isotopes and ICP MS, not available for this study due to fund limitation. Conclusion This study assessed the potential role of iron added in the metallic form and in different particle sizes to sediments on the biotransformation of both native and newly added Hg. The major findings can be summarized as follows: Relationship between the different fractions of Hg present in the solid phase and the amount of MeHg produced during the incubation of sedi ment slurries points to Hg speciation as a determinant factor in controlling the production of MeHg. In subsequent set of experiments using the same sediments but spiked with Hg, no significant amount of MeHg was produced neither This suggests that sediments used had a very high binding capacity for Hg whic h in turn controls the Hg availability and potential for methylation. T he addition of metallic iron particles in sediments could impact the production of MeHg through two main mechanisms: (i) adsorption onto oxyhydroxides layers or (ii) release of Fe(II) which co precipitates with Hg in iron sulfide minerals depending on redox conditions T hese processes would reduce the bioavailable fraction of Hg, and therefore, its methylation by sediment s microorganisms Ideally measurements of potential rates of Hg methylation (M) and MeHg demethylation (D) would have been necessary to support the above conclusions. Further studies should consider (i) the concentration and types of organic carbon sources, (ii) M/D ratios and (iii) kinetic aspects of MeHg production
121 CHAPTER 6 CONCLUSION AND RE COMMENDATIONS This research focused on the interaction of iron and mercury (Hg) in aquatic systems. With the ultimate goals of understanding the potential mechanism of such interactions and assessing the possibility of using identified biogeochemical processes in controlling Hg pollution issues. The rational behind this study is shown conceptually in Figure 6.1. Figure 6 1: Hg iron interactions investigated in this study. Three different reaction pathways are anticipated to play a significant role in the fate of Hg in aqueous and sediment phases (Figure 61). First Hg can be eliminated f rom the aqueous phase through volatilization following its interaction with metalli c iron particles. This study showed that this specific process does not play a significant role with regard to the total mass of Hg removed from contaminated water matrixes. Second corrosion of iron particles and the resulting formation of oxyhydroxide layers provide adsorption sites for removal of Hg from aqueous solution. Third metallic iron impact s the bioavailability of inorganic Hg to methylating agents by (i) iron oxidation and Hg FeHg(OH) n Fe(OH) n Adsorption Hg 2+ Hg 0 Fe(OH) n Fe 0 2e CH 3 Hg + Methylation z y n 2 2 4 2 2 0S (Hg) Fe S SO Fe O /H Fe Volatilization
122 adsorption and (ii) iron reduction and formation of Fe(II) which in combination with produced sul fides would coprecipitate dissolved Hg, hence limiting its availability for methylation. The major findings based on lab experiments testing the above hypotheses are as follows: Batch experiments demonstrated that rates of Hg volatilization were impacted primarily by water chemistry (DI water versus wastewater effluents (WW)), while the particle size of metallic iron seemed to play a minor role. Depending on how data are normalized (specific surface area (SSA) versus mass), kinetic studies show that rates of Hg adsorption are faster for nZVI when data are normalized by mass and the opposite when normalized by SSA. The faster rate of nZVI oxidation allows for early removal of aqueous Hg while ZVI would require a much longer time for oxidation and efficient removal of Hg. Experimental column studies have shown that the flow rate play a significant role in the efficiency of Hg removal by metallic iron particles. Unfortunately, good Hg removal is obtained at low rate such as 75 l/min while much higher flow rates such as 500 l/min and up resulted in no Hg removal at all. Further studies are needed to improve Hg removal from columns using conditions that can be scaled up for industrial applications. The mass of nZVI that is necessary for an efficient removal of Hg is a parameter which needs to be considered. First, column entirely packed with nZVI could be expensive but at the same time result in clogging as observed by Weisner et al. (2003) using bulk ZVI. There is a need for a pretreatment of water due to the role of binding ligands such as dissolved organic matter, which limits Hg adsorption on oxyhydroxide layer s. The presence of competitive cations did not significantly impact Hg adsorption and removal at tested concentratio ns. Further studies would be needed to test a much broader range of cation concentrations. Lab experiments adding i ron particles to sediment slurries showed a potential for controlling Hg methylation in sediments. However, preliminary results obtained from this study need to be supplemented by further and more detailed investigations to validate this hypothesis.
123 Recommendations : All the volatilization experiments in this study were done in closed batch reactors, and further investigations are needed using open and column type conditions. However, the low rates of volatilization from batch experiments tend to suggest that this pathway may not be significant. In addition the effect of water chemi stry on Hg volatilization should be investigated further as cer tain dissolved compounds such as humic substances can substantially increase rates of Hg volatilization (Alberts et al. 1974; Ravichandran 2004) Given the wide variety of chemical composition of WW a study based on the types of WW samples would be necessary to validate the efficiency of nZVI technology. To efficiently account for the effect of SSA, methods should be developed to make sure that nZVI is not used as large aggregates, but w ell dispersed suspensions to take advantage of specific physical characteristics of nano size particles. Further r esearch is needed on the release of Hg bound to complex organic constituents by use of different concentrations and types of oxidizing agents Results from Hg methylation studies suggest that further research is needed to validate the hypothesis put forward in this study.
124 APPENDIX A STATISTICAL ANALYSIS DATA The statistical data for the methyl mercury experiments put forth in Chapter 5 are shown below. The t test compares the means of two groups to assess whether there is a significant difference between the groups. Group 1 Group 2 Pooled Standard Deviation sp t statistic Degrees of Freedom df Alpha Level Standard Table of Significan ce t value Significant difference Control Control Trt 2 0.52 0.66 4 0.05 2.78 No Control Control ZVI 0.57 0.31 4 0.05 2.78 No Control Control nZVI 0.52 0.59 4 0.05 2.78 No Control Trt 1 ZVI 0.52 0.27 4 0.05 2.78 No Control Trt 1 nZVI 0.64 0.49 4 0.05 2.78 No Control Trt 2 ZVI 0.52 1.28 4 0.05 2.78 No Control Trt 2 nZVI 0.55 1.01 4 0.05 2.78 No Control Trt 3 ZVI 0.56 1.23 4 0.05 2.78 No Control Trt 3 nZVI 0.65 0.71 4 0.05 2.78 No ZVI 0.5 g ZVI 0.2 g 0.32 0.59 4 0.05 2.78 No ZVI 0.5 g ZVI 0.1 g 0.14 7.98 4 0.05 2.78 Yes nZVI 0.01 g nZVI 0.02 g 0.46 2.81 4 0.05 2.78 Yes nZVI 0.01 g nZVI 0.05 g 0.47 1.66 4 0.05 2.78 No ZVI Trt 5 ZVI Trt 6 0.66 0.90 4 0.05 2.78 No ZVI Trt 1 ZVI Trt 5 0.50 0.01 4 0.05 2.78 No ZVI Trt 3 ZVI Trt 6 0.66 0.37 4 0.05 2.78 No nZVI Trt 5 nZVI Trt 6 0.43 3.32 4 0.05 2.78 Yes nZVI Trt 1 nZVI Trt 5 0.57 0.74 4 0.05 2.78 No nZVI Trt 3 nZVI Trt 6 0.61 2.90 4 0.05 2.78 Yes
125 LIST OF REFERENCES Aberg, B., Ekman, L., Falk, R., Greitz, U., Perss on, G., and Snihs, J. O. (1969). "Metabolism of Methyl Mercury (203hg) Compounds in Man Excretion and Distribution." Archives of Environmental Health, 19(4), 478&. Adams, N., Caroll, D., Madalinski, k., Rock, S., Wilson, T., and Pivetz, B. (2000). "Intr oduction to Phytoremediation." National Risk Management Research Laboratory. Office of Reasearch and Development.US Environmental Protection/600/R 99/107. Alberts, J. J., Schindle.Je, Miller, R. W., and Nutter, D. E. (1974). "Elemental Mercury Evolution M ediated by Humic Acid." Science 184(4139), 895896. Alkorta, I., and Garbisu, C. (2001). "Phytoremediation of organic contaminants in soils." Bioresource Technology 79(3), 273276. Alowitz, M. J., and Scherer, M. M. (2002). "Kinetics of nitrate, nitrite, and Cr(VI) reduction by iron metal." Environmental Science & Technology 36(3), 299306. Alvarez, P. J., and Illman, W. A. (2006). Bioremediation and Natural Attenuation: Process Fundamentals and Mathematical Models John Wiley & Sons, Inc, Hoboken, NJ. A myot, M., Gill, G. A., and Morel, F. M. M. (1997). "Production and loss of dissolved gaseous mercury in coastal seawater." Environmental Science & Technology 31(12), 36063611. Amyot, M., Mierle, G., Lean, D. R. S., and McQueen, D. J. (1994). "SunlightIn duced Formation of Dissolved Gaseous Mercury in Lake Waters." Environmental Science & Technology 28(13), 23662371. Anirudhan, T. S., Senan, P., and Unnithan, M. R. (2007). "Sorptive potential of a cationic exchange resin of carboxyl banana stem for mercury(II) from aqueous solutions." Separation and Purification Technology 52(3), 512519. Astrup, T., Stipp, S. L. S., and Christensen, T. H. (2000). "Immobilization of chromate from coal fly ash leachate using an attenuating barrier containing zerovalent i ron." Environmental Science & Technology 34(19), 41634168. Atia, A. A., Donia, A. M., and Elwakeel, K. Z. (2005). "Selective separation of mercury (II) using a synthetic resin containing amine and mercaptan as chelating groups." Reactive and Functional P olymers 65(3), 267275.
126 Atkins, P. (1994). Physical Chemistry Oxford Univeristy Press. Atwood, D. A., and Zaman, M. K. (2006). "Mercury removal from water." Recent Developments in Mercury Science, 163182. Bartzas, G., Komnitsas, K., and Paspaliaris, I. (2006). "Laboratory evaluation of Fe0 barriers to treat acidic leachates." Minerals Engineering Selected papers from Processing and Disposal of Minerals Industry Wastes '05 19(5), 505 514. Beijer, K., and Jernelov, A. (1979). "Methylation of mercury in aquatic environments." The Biogeochemistry of Mercury in the Environment, J. O. Nriagu, ed., Elsevier/North Holland Biomendical Press, Amsterdam, 203210. Bennicelli, R., Stepniewska, Z., Banach, A., Szajnocha, K., and Ostrowski, J. (2004). "The ability of Azolla caroliniana to remove heavy metals (Hg(II), Cr(III), Cr(VI)) from municipal waste water." Chemosphere, 55(1), 141146. Benoit, J. M., Gilmour, C. C., Heyes, A., Mason, R. P., and Miller, C. L. (2003). "Geochemical and biological controls over methyl mercury production and degradation in aquatic ecosystems." Biogeochemistry of Environmentally Important Trace Elements, 262297. Berger, C. M., Geiger, C. L., Clausen, C. A., Billow, A. M., Quinn, J. W., and Brooks, K. B. (2006). "Evaluating trichloroethyl ene degradation using differing nanoand micro scale iron particles." Remediation of chlorinated and recalcitrant compounds, 2006. Proceedings of the fifth international conference on remediation of chlorinated and recalcitrant compounds, Monterey, California, 2225 May, 2006, Battelle Press, Colombus USA, C 23. Biernat, R. J., and Robins, R. G. (1972). "HighTemperature Potential/Ph Diagrams for Iron Water and IronWater Sulfur Systems." Electrochimica Acta 17(7), 1261&. Bloom, N. (1989). "Determination of Picogram Levels of Methylmercury by Aqueous Phase Ethylation, Followed by Cryogenic Gas Chromatography with Cold Vapor Atomic Fluorescence Detection." Canadian Journal of Fisheries and Aquatic Sciences 46(7), 11311140. Bloom, N. S. (1992). "On the Ch emical Form of Mercury in Edible Fish and Marine Invertebrate Tissue." Canadian Journal of Fisheries and Aquatic Sciences 49(5), 10101017. Blowes, D. W., Ptacek, C. J., Benner, S. G., McRae, C. W. T., Bennett, T. A., and Puls, R. W. (2000). "Treatment of inorganic contaminants using permeable reactive barriers." Journal of Contaminant Hydrology 45(12), 123137.
127 Blowes, D. W., Ptacek, C. J., and Jambor, J. L. (1997). "Insitu remediation of Cr(VI) contaminated groundwater using permeable reactive walls: Laboratory studies." Environmental Science & Technology 31(12), 33483357. Boening, D. W. (2000). "Ecological effects, transport, and fate of mercury: a general review." Chemosphere 40(12), 13351351. Boffetta, P., Garcia Gomez, M., Pompe Kirn, V., Zarid ze, D., Bellander, T., Bulbulyan, M., Caballero, J. D., Ceccarelli, F., Colin, D., Dizdarevic, T., Espanol, S., Kobal, A., Petrova, N., Sallsten, G., and Merler, E. (1998). "Cancer occurrence among European mercury miners." Cancer Causes & Control 9(6), 5 91599. Boffetta, P., Merler, E., and Vainio, H. (1993). "Carcinogenicity of Mercury and Mercury Compounds." Scandinavian Journal of Work Environment & Health, 19(1), 17. Bonzongo, J. C. J., Heim, K. J., Chen, Y. A., Lyons, W. B., Warwick, J. J., Miller, G. C., and Lechler, P. J. (1996). "Mercury pathways in the Carson River Lahontan reservoir system, Nevada, USA." Environmental Toxicology and Chemistry 15(5), 677683. Bostick, W. D., Shoemaker, J. L., Osborne, P. E., and Evansbrown, B. (1990). "Treatment and Disposal Options for a Heavy Metals Waste Containing Soluble Tc 99." Acs Symposium Series, 422, 345367. Boussahel, R., Harik, D., Mammar, M., and LamaraMohamedl, S. (2007). "Degradation of obsolete DDT by Fenton oxidation with zerovalent iron." Des alination 206(13), 369372. Cantrell, K. J., Kaplan, D. I., and Wietsma, T. W. (1995). "Zero Valent Iron for the inSitu Remediation of Selected Metals in Groundwater." Journal of Hazardous Materials 42(2), 201212. Celo, V., Lean, D. R. S., and Scott, S. L. (2006). "Abiotic methylation of mercury in the aquatic environment." Science of The Total Environment Selected papers from the 7th International Conference on Mercury as a Global Pollutant, Ljubljana, Slovenia June 27 July 2, 2004, 368(1), 126137. Chen, W., Westerhoff, P., Leenheer, J. A., and Booksh, K. (2003). "Fluorescence excitation Emission matrix regional integration to quantify spectra for dissolved organic matter." Environmental Science & Technology 37(24), 5701 5710. Chen, Y., Bonzongo, J. C. J., Lyons, W. B., and Miller, G. C. (1997). "Inhibition of mercury methylation in anoxic freshwater sediment by group VI anions." Environmental Toxicology and Chemistry 16(8), 15681574.
128 Chingombe, P., Saha, B., and Wakeman, R. J. (2005). "Surface modification and characterisation of a coal based activated carbon." Carbon, 43(15), 31323143. Choe, S., Chang, Y. Y., Hwang, K. Y., and Khim, J. (2000). Chemosphere, 41(null), 1307. Choi, S. C., Chase, T., and Bartha, R. (1994). "Metabolic Pathways Leading to Mercury Methylation in DesulfovibrioDesulfuricans Ls." Applied and Environmental Microbiology 60(11), 40724077. Clever, H. L., Johnson, S. A., and Derrick, M. E. (1985). "The Solubility of Mercury and Some Sparingly Soluble Mercury Salts in Water and Aqueous Electrolyte Solutions." Journal of Physical and Chemical Reference Data, 14(3), 631681. Compeau, G., and Bartha, R. (1984). "Methylation and Demethylation of Mercury under Controlled Redox, Ph, and Salinity Conditions." Applied and Environmental Microbiology 48(6), 1203 1207. Compeau, G. C., and Bartha, R. (1985). "SulfateReducing Bacteria Principal Methylators of Mercury in Anoxic Estuarine Sediment." Applied and Environmental Microbiology 50(2), 498502. Craig, P. (1986). "Organometalli c Compounds in the Environment: Principles and Reactions." Organomercury Compounds in the Environment, P. Craig and H. Longman, eds., Chap 2, 65110. Cundy, A. B., Hopkinson, L., and Whitby, R. L. D. (2008). "Use of ironbased technologies in contaminated land and groundwater remediation: A review." Science of The Total Environment In Press, Corrected Proof. Deng, L., Fu, D. F., and Deng, N. S. (2009). "Photoinduced transformations of mercury(II) species in the presence of algae, Chlorella vulgaris." Jour nal of Hazardous Materials 164(23), 798805. Di Natale, F., Lancia, A., Molino, A., Di Natale, M., Karatza, D., and Musmarra, D. (2006). "Capture of mercury ions by natural and industrial materials." Journal of Hazardous Materials 132(23), 220225. Don g, W. M., Liang, L. Y., Brooks, S., Southworth, G., and Gu, B. H. (2009). "Roles of dissolved organic matter in the speciation of mercury and methylmercury in a contaminated ecosystem in Oak Ridge, Tennessee." Environmental Chemistry 7(1), 94102.
129 Dujardi n, M. C., Caze, C., and Vroman, I. (2000). "Ion exchange resins bearing thiol groups to remove mercury.: Part 1: synthesis and use of polymers prepared from thioester supported resin." Reactive and Functional Polymers 43(12), 123132. Farrell, J., Kason, M., Melitas, N., and Li, T. (2000). "Investigation of the long term performance of zerovalent iron for reductive dechlorination of trichloroethylene." Environmental Science & Technology 34(3), 514521. Farrell, J., Wang, J. P., O'Day, P., and Conklin, M (2001). "Electrochemical and spectroscopic study of arsenate removal from water using zerovalent iran media." Environmental Science & Technology 35(10), 20262032. Fendorf, S., Eick, M. J., Grossl, P., and Sparks, D. L. (1997). "Arsenate and chromate r etention mechanisms on goethite .1. Surface structure." Environmental Science & Technology 31(2), 315320. Feng, N., Bitton, G., Yeager, P., Bonzongo, J. C., and Boularbah, A. (2007). "Heavy metal removal from soils using magnetic separation: 1. Laborator y experiments." Clean Soil Air Water 35, 362 369. Fleming, E. J., Mack, E. E., Green, P. G., and Nelson, D. C. (2006). "Mercury methylation from unexpected sources: Molybdateinhibited freshwater sediments and an ironreducing bacterium." Applied and Envi ronmental Microbiology 72(1), 457464. Gabriel, M. C., and Williamson, D. G. (2004). "Principal biogeochemical factors affecting the speciation and transport of mercury through the terrestrial environment." Environmental Geochemistry and Health, 26(4), 42 1 434. Gao, J., Youn, S., Hovsepyan, A., Llaneza, V. L., Wang, Y., Bitton, G., and Bonzongo, J. C. J. (2009). "Dispersion and Toxicity of Selected Manufactured Nanomaterials in Natural River Water Samples: Effects of Water Chemical Composition." Environmental Science & Technology 43(9), 33223328. Gardfeldt, K., Munthe, J., Stromberg, D., and Lindqvist, O. (2003). "A kinetic study on the abiotic methylation of divalent mercury in the aqueous phase." The Science of The Total Environment Pathways and process es of mercury in the e nvironment. Selected papers presented at the sixth International Conference on Mercury as Global Pollutant, Minamata, Japan, Oct. 1519, 2001, 304(13), 127136. Gavaskar, A. R. (1999). "Design and construction techniques for permea ble reactive barriers." Journal of Hazardous Materials 68(1 2), 4171.
130 Gillham, R. W., and Ohannesin, S. F. (1994). "Enhanced Degradation of Halogenated Aliphatics by ZeroValent Iron." Ground Water 32(6), 958967. Gilmour, C. C., and Henry, E. A. (1991) "Mercury Methylation in Aquatic Systems Affected by Acid Deposition." Environmental Pollution 71(2 4), 131169. Gilmour, C. C., and Henry, E. A. (1992). "Mercury Methylation by SulfateReducing Bacteria Biogeochemical and Pure Culture Studies." Abstra cts of Papers of the American Chemical Society 203, 140GEOC. Givelet, N., Roos Barraclough, F., Goodsite, M. E., Cheburkin, A. K., and Shotyk, W. (2004). "Atmospheric mercuty accumulation rates between 5900 and 800 calibrated years BP in the high Artic o f Canada recorded by peat hummocks." Environmental Science & Technology 38(19), 49644972. Gomez Serrano, V., Macias Garcia, A., ESPINOSA MANSILLA, A., and VALENZUELACALAHORRO, C. (1998). Adsorption of mercury, cadmium and lead from aqueous solution on heat treated and sulphurized activated carbon." Water Research, 32(1), 14. Gu, B., Liang, L., Dickey, M. J., Yin, X., and Dai, S. (1998). "Reductive precipitation of uranium(VI) by zerovalent iron." Environmental Science & Technology 32(21), 33663373. Gu, B., Phelps, T. J., Liang, L., Dickey, M. J., Roh, Y., Kinsall, B. L., Palumbo, A. V., and Jacobs, G. K. (1999). "Biogeochemical dynamics in zerovalent iron columns: Implications for permeable reactive barriers." Environmental Science & Technology 33( 13), 21702177. Gustin, M. S., Chavan, P. V., Dennett, K. E., Donaldson, S., Marchand, E., and Fernanadez, G. (2006). "Use of constructed wetlands with four different experimental designs to assess the potential for methyl and total Hg outputs." Applied Ge ochemistry Mercury: Distribution, Transport, and Geochemical and Microbial Transformations from Natural and Anthropogenic Sources 21(11), 20232035. Hagare, P., Thiruvenkatachari, R., and Ngo, H. H. (2001). "A feasibility study of using hematite to remov e dissolved organic carbon in water treatment ." Separation Science and Technology 36(11), 2547 2559. Han, S., Obraztsova, A., Pretto, P., Deheyn, D. D., Gieskes, J., and Tebo, B. M. (2008). "Sulfide and iron control on mercury speciation in anoxic estuarine sediment slurries." Marine Chemistry 111(34), 214220.
131 Hanlon, J. (2007). "Analytical Methods for Mercury in National Pollutant Discharge Elimination Systems (NPDES) Permits." O. o. W. Management, ed., US EPA. Harmon, S. M., King, J. K., Gladden, J. B., Chandler, G. T., and Newman, L. A. (2004). "Methylmercury formation in a wetland mesocosm amended with sulfate." Environmental Science & Technology 38(2), 650656. Harmon, S. M., King, J. K., Gladden, J. B., and Newman, L. A. (2007). "Using sulfatea mended sediment slurry batch reactors to evaluate mercury methylation." Archives of Environmental Contamination and Toxicology 52(3), 326 331. Hawkins, W. B., Rodgers, J. H., Gillespie, W. B., Dunn, A. W., Dorn, P. B., and Cano, M. L. (1997). "Design and Construction of Wetlands for Aqueous Transfers and Transformations of Selected Metals." Ecotoxicology and Environmental Safety 36(3), 238248. Hepler, L. G., and Olofsson, G. (1975). "Mercury Thermodynamic Properties, Chemical Equilibria, and Standard P otentials." Chemical Reviews 75(5), 585602. Hinton, J. J., Veiga, M. M., and Beinhoff, C. (2003). "Women, mercury and artisanal gold mining: Risk communication and mitigation." Journal De Physique Iv 107, 617620. Hoch, L. B., Mack, E. J., Hydutsky, B. W., Hershman, J. M., Skluzacek, I. M., and Mallouk, T. E. (2008). "Carbothermal synthesis of carbonsupported nanoscale zero valent iron particles for the remediation of hexavalent chromium." Environmental Science & Technology 42(7), 26002605. Hovsepyan, A., and Bonzongo, J. C. J. (2009). "Aluminum drinking water treatment residuals (Al WTRs) as sorbent for mercury: Implications for soil remediation." Journal of Hazardous Materials 164(1), 7380. Hsu Kim, H., and Sedlak, D. L. (2005). "Similarities between inorganic sulfide and the strong Hg(II) Complexing ligands in municipal wastewater effluent." Environmental Science & Technology 39(11), 40354041. Huang, Y. H., and Zhang, T. C. (2005). "Effects of dissolved oxygen on formation of corrosion products and concomitant oxygen and nitrate reduction in zerovalent iron systems with or without aqueous Fe2+." Water Research 39(9), 17511760.
132 Huang, Y. H., Zhang, T. C., Shea, P. J., and Comfort, S. D. (2003). "Effects of oxide coating and selected cations on nitrate reduction by iron metal." Journal of Environmental Quality 32(4), 13061315. Hylander, L. D., Grohn, J., Tropp, M., Vikstrom, A., Wolpher, H., de Castro e Silva, E., Meili, M., and Oliveira, L. J. (2006). "Fish mercury increase in Lago Manso, a n ew hydroelectric reservoir in tropical Brazil." Journal of Environmental Management Mercury cycling in contaminated tropical nonmarine ecosystems 81(2), 155166. Jay, J. A., Murray, K. J., Gilmour, C. C., Mason, R. P., Morel, F. M. M., Roberts, A. L., and Hemond, H. F. (2002). "Mercury methylation by Desulfovibrio desulfuricans ND132 in the presence of polysulfides." Applied and Environmental Microbiology 68(11), 57415745. Jensen, S., and Jernelov, A. (1969). "Biological Methylation of Mercury in Aquatic Organisms." Nature, 223(5207), 753&. Jing, Y. D., He, Z. L., and Yang, X. E. (2007). "Effects of pH, organic acids, and competitive cations on mercury desorption in soils." Chemosphere, 69(10), 16621669. Junyapoon, S. (2005). "Use of ZeroValent Iron for Wastewater Treatment." KMITL Sci. Tech. J. 5(3), 587595. Kamal, M., Ghaly, A. E., Mahmoud, N., and Cote, R. (2004). "Phytoaccumulation of heavy metals by aquat ic plants." Environment International 29(8), 10291039. Kanel, S. R., Greneche, J. M., and Choi, H. (2006). "Arsenic(V) removal kom groundwater using nano scale zerovalent iron as a colloidal reactive barrier material." Environmental Science & Technology 40(6), 20452050. Kanel, S. R., Manning, B., Charlet, L., and Choi, H. (2005). "Removal of arsenic(III) from groundwater by nanoscale zerovalent iron." Environmental Science & Technology 39(5), 12911298. Kenneke, J. F., and McCutcheon, S. C. (2003). Use of pretreatment zones and zerovalent iron for the remediation of chloroalkenes in an oxic aquifer." Environmental Science & Technology 37(12), 28292835. Kerin, E. J., Gilmour, C. C., Roden, E., Suzuki, M. T., Coates, J. D., and Mason, R. P. (2006). "Mercury methylation by dissimilatory ironreducing bacteria." Applied and Environmental Microbiology 72(12), 79197921.
133 Khudenko, B. M., and Garciapastrana, A. (1987). "Temperature Influence on Absorption and Stripping Processes." Water Science and Technology 19(5 6), 877888. Kim, J. P. (1995). "Methylmercury in Rainbow Trout (Oncorhynchus Mykiss) from Lakes Okareka, Okaro, Rotomahana, Rotorua and Tarawera, NorthIsland, New Zealand." Science of the Total Environment 164(3), 209219. King, J. K., Harmo n, S. M., Fu, T. T., and Gladden, J. B. (2002). "Mercury removal, methylmercury formation, and sulfatereducing bacteria profiles in wetland mesocosms." Chemosphere, 46(6), 859870. Koeber, R., Welter, E., Ebert, M., and Dahmke, A. (2005). "Removal of arse nic from groundwater by zerovalent iron and the role of sulfide." Environmental Science & Technology 39(20), 80388044. Krauskopf, K. B., and Bird, D. K. (1995). Introduction to Geochemistry, 3rd, McGraw Hill. Lalonde, J. D., Amyot, M., Orvoine, J., Morel F. M. M., Auclair, J. C., and Ariya, P. A. (2004). "Photoinduced oxidation of Hg 0 (aq) in the waters from the St. Lawrence estuary." Environmental Science & Technology 38(2), 508514. Leonhauser, J., Rohricht, M., Wagner Dobler, I., and Deckwer, W. D. (2006). "Reaction engineering aspects of microbial mercury removal." Engineering in Life Sciences 6(2), 139148. Li, L., Fan, M. H., Brown, R. C., Van Leeuwen, J. H., Wang, J. J., Wang, W. H., Song, Y. H., and Zhang, P. Y. (2006a). "Synthesis, properties, and environmental applications of nanoscale ironbased materials: A review." Critical Reviews in Environmental Science and Technology 36(5), 405431. Li, S. J., Li, T. L., Xiu, Z. M., and Jin, Z. H. (2008). "Reduction and immobilization of chromium(VI) b y nanoscale Fe 0 particles supported on reproducible PAA/PVDF membrane." Journal of Environmental Monitoring, 12(5), 11531158. Li, X. Q., Elliott, D. W., and Zhang, W. X. (2006b). "Zerovalent iron nanoparticles for abatement of environmental pollutants: Materials and engineering aspects." Critical Reviews in Solid State and Materials Sciences 31(4), 111122. Lien, H. L., and Wilkin, R. T. (2005a). "Highlevel arsenite removal from groundwater by zero valent iron." Chemosphere, 59(3), 377386.
134 Lien, H. L., and Wilkin, R. T. (2005b). "Highlevel arsenite removal from groundwater by zero valent iron." Chemosphere, 59(3), 377386. Lin, C. J., and Pehkonen, S. O. (1998). "Oxidation of elemental mercury by aqueous chlorine (HOCl/OCl): Implications for troposp heric mercury chemistry." Journal of Geophysical ResearchAtmospheres 103(D21), 2809328102. Lin, C. J., and Pehkonen, S. O. (1999). "The chemistry of atmospheric mercury: a review." Atmospheric Environment 33(13), 20672079. Loux, N. T. (2004). "A criti cal assessment of elemental mercury air/water exchange parameters." Chemical Speciation and Bioavailability 16(4), 127138. Majewski, P. (2006). "Nanomaterials for Watertreatment." Nanomaterials Toxicity, Health, and Environmetal Issues, C. S. S. R. Kumar ed., Wiley VCH Verlag GmbH & Co, Weinheim, 211233. Manning, B. A., Kiser, J. R., Kwon, H., and Kanel, S. R. (2007). "Spectroscopic investigation of Cr(III) and Cr(VI) treated nanoscale zerovalent iron." Environmental Science & Technology 41(2), 58659 2. Matheson, L. J., and Tratnyek, P. G. (1994). "Reductive Dehalogenation of Chlorinated Methanes by Iron Metal." Environmental Science & Technology 28(12), 20452053. McRae, C. W. T., Blowes, D. W., and Ptacek, C. J. (1997). "Laboratory scale investigati on of As and Se using iron oxides." Proc. Sixth Symposium and Exhibition on Groundwater and Soil Remediation, 167 168. Mehrotra, A., and Sedlak, D. (2005). "Decrease in Net Mercury Methylation Rates Following Iron Amendment to Anoxic Wetland Sediment Slurr ies." Environmental Science & Technology 39(8), 25642570. Miller, C. L., Mason, R. P., Gilmour, C. C., and Heyes, A. (2007). "Influence of dissolved organic matter on the complexation of mercury under sulfidic conditions." Environmental Toxicology and Ch emistry 26(4), 624633. Miller, C. L., Southworth, G., Brooks, S., Liang, L. Y., and Gu, B. H. (2009). "Kinetic controls on the complexation between mercury and dissolved organic matter in a contaminated environment." Environmental Science & Technology 4 3(22), 85488553.
135 Mohapatra, M., Rout, K., Gupta, S. K., Singh, P., Anand, S., and Mishra, B. K. (2010). "Facile synthesis of additiveassisted nano goethite powder and its application for fluoride remediation." Journal of Nanoparticle Research, 12(2), 681686. Mondal, K., Jegadeesan, G., and Lalvani, S. B. (2004). "Removal of selenate by Fe and NiFe nanosized particles." Industrial & Engineering Chemistry Research, 43(16), 49224934. Monperrus, M., Tessier, E., Point, D., Vidimova, K., Amouroux, D., Guyoneaud, R., Leynaert, A., Grall, J., Chauvaud, L., Thouzeau, G., and Donard, O. F. X. (2007). "The biogeochemistry of mercury at the sediment water interface in the Thau Lagoon. 2. Evaluation of mercury methylation potential in both surface sediment and the w ater column." Estuarine, Coastal and Shelf Science Biogeochemical and contaminant cycling in sediments from a humanimpacted coastal lagoon, 72(3), 485496. Moraci, N., and Calabro, P. S. (2010). "Heavy metals removal and hydraulic performance in zeroval ent iron/pumice permeable reactive barriers." Journal of Environmental Management 91(11), 23362341. Morrison, S. J., Metzler, D. R., and Carpenter, C. E. (2001). "Uranium precipitation in a permeable reactive barrier by progressive irreversible dissoluti on of zerovalent iron." Environmental Science & Technology 35(2), 385390. Mulligan, C. N., Yong, R. N., and Gibbs, B. F. (2001). "An evaluation of technologies for the heavy metal remediation of dredged sediments." Journal of Hazardous Materials 85(12) 145163. Nabais, J. V., Carrott, P. J. M., Carrott, M. M. L. R., Belchior, M., Boavida, D., Diall, T., and Gulyurtlu, I. (2006). "Mercury removal from aqueous solution and flue gas by adsorption on activated carbon fibres." Applied Surface Science, 252(17), 60466052. Narr, J., Viraraghavan, T., and Jin, Y. C. (2007). "Applications of nanotechnology in water/wastewater treatment: A review." Fresenius Environmental Bulletin, 16(4), 320329. Niu, S. F., Liu, Y., Xu, X.H., and Lou, Z. H. (2005). "Removal of hexavalent chromium from aqueous solution by iron nanoparticles." J Zhejiang Univ Sci B 6(10), 10227. Noubactep, C. (2008). "A critical review on the process of contaminant removal in Fe0 H2O systems." Environmental Technology 29(8), 909920.
136 Nriagu, J. O. (1988). "A Silent Epidemic of Environmental Metal Poisoning." Environmental Pollution, 50(1 2), 139161. Nurmi, J. T., Tratnyek, P. G., Sarathy, V., Baer, D. R., Amonette, J. E., Pecher, K., Wang, C. M., Linehan, J. C., Matson, D. W., Penn, R. L., an d Driessen, M. D. (2005). "Characterization and properties of metallic iron nanoparticles: Spectroscopy, electrochemistry, and kinetics." Environmental Science & Technology 39(5), 12211230. Oh, Y. J., Song, H., Shin, W. S., Choi, S. J., and Kim, Y. H. (2 007). "Effect of amorphous silica and silica sand on removal of chromium(VI) by zerovalent iron." Chemosphere, 66(5), 858865. Pak, K. R., and Bartha, R. (1998). "Mercury methylation by interspecies hydrogen and acetate transfer between sulfidogens and me thanogens." Applied and Environmental Microbiology 64(6), 19871990. Paquette, K. E., and Helz, G. R. (1997). "Inorganic speciation of mercury in sulfidic waters: The importance of zerovalent sulfur." Environmental Science & Technology 31(7), 21482153. Pearson, R. G. (1963). "Hard and Soft Acids and Bases." Journal of the American Chemical Society 85(22), 3533&. Penichecovas, C., Alvarez, L. W., and Arguellesmonal, W. (1992). "The Adsorption of Mercuric Ions by Chitosan." Journal of Applied Polymer Sc ience, 46(7), 11471150. Phillips, D. H., Gu, B., Watson, D. B., Roh, Y., Liang, L., and Lee, S. Y. (2000). "Performance evaluation of a zerovalent iron reactive barrier: Mineralogical characteristics." Environmental Science & Technology 34(19), 4169 4176. Ponder, S. M., Darab, J. G., Bucher, J., Caulder, D., Craig, I., Davis, L., Edelstein, N., Lukens, W., Nitsche, H., Rao, L. F., Shuh, D. K., and Mallouk, T. E. (2001). "Surface chemistry and electrochemistry of supported zerovalent iron nanoparticles in the remediation of aqueous metal contaminants." Chemistry of Materials 13(2), 479486. Ponder, S. M., Darab, J. G., and Mallouk, T. E. (2000). "Remediation of Cr(VI) and Pb(II) aqueous solutions using supported, nanoscale zerovalent iron." Environmental Science & Technology 34(12), 25642569.
137 Powell, R. M., Puls, R. W., Hightower, S. K., and Sabatini, D. A. (1995). "Coupled Iron Corrosion and Chromate Reduction Mechanisms for Subsurface Remediation." Environmental Science & Technology 29(8), 19131922. Puls, R. W., Paul, C. J., and Powell, R. M. (1999). "The application of in situ permeable reactive (zerovalent iron) barrier technology for the remediation of chromatecontaminated groundwater: a field test." Applied Geochemistry 14(8), 9891000. Rangs ivek, R., and Jekel, M. R. (2005). "Removal of dissolved metals by zerovalent iron (ZVI): Kinetics, equilibria, processes and implications for stormwater runoff treatment." Water Research, 39(17), 41534163. Raposo, R. R., Melendez Hevia, E., and Spiro, M (2000). "Autocatalytic formation of colloidal mercury in the redox reaction between Hg2+ and Fe2+ and between Hg22+ and Fe2+." Journal of Molecular Catalysis A: Chemical 164(12), 4959. Ravichandran, M. (2004). "Interactions between mercury and dissolv ed organic matter a review." Chemosphere, 55(3), 319331. Reynolds, R. L., Fishman, N. S., Wanty, R. B., and Goldhaber, M. B. (1990). "Iron Sulfide Minerals at Cement Oil Field, Oklahoma Implications for Magnetic Detection of Oil Fields." Geological Society of America Bulletin, 102(3), 368380. Ritter, K., Odziemkowski, M. S., Simpgraga, R., Gillham, R. W., and Irish, D. E. (2003). "An in situ study of the effect of nitrate on the reduction of trichloroethylene by granular iron." Journal of Contaminant Hydrology 65(12), 121136. Rumbold, D. G., and Fink, L. E. (2006). "Extreme spatial variability and unprecedented methylmercury concentrations within a constructed wetland." Environmental Monitoring and Assessment 112(13), 115135. Sanchez, I., Stuber, F., Font, J., Fortuny, A., Fabregat, A., and Bengoa, C. (2007). "Elimination of phenol and aromatic compounds by zero valent iron and EDTA at low temperature and atmospheric pressure." Chemosphere, 68(2), 338344. Sanemasa, I. (1975). "Solubility of Eleme ntal Mercury Vapor in Water." Bulletin of the Chemical Society of Japan, 48(6), 17951798. Sarathy, V., Salter, A. J., Nurmi, J. T., Johnson, G. O., Johnson, R. L., and Tratnyek, P. G. (2010). "Degradation of 1,2,3Trichloropropane (TCP): Hydrolysis, Elimi nation, and Reduction by Iron and Zinc." Environmental Science & Technology 44(2), 787793.
138 Sarkar, D., Essington, M. E., and Misra, K. C. (2000). "Adsorption of mercury(II) by kaolinite." Soil Science Society of America Journal 64(6), 1968 1975. Sayles, G. D., You, G. R., Wang, M. X., and Kupferle, M. J. (1997). "DDT, DDD, and DDE dechlorination by zerovalent iron." Environmental Science & Technology 31(12), 34483454. Schluter, K. (2000). "Review: evaporation of mercury from soils. An integration and synthesis of current knowledge." Environmental Geology 39(34), 249271. Schnoor, J. (1997). "Phytoremediation, Technology Evaluation Report." Groundwater Remediation Technologies Analysis Center TE 98 01, 130. Schroeder, W., Lindqvist, O., Munthe, J., and Xiao, Z. F. (1992). "Volatilization of Mercury from Lake Surfaces." Science of the Total Environment 125, 47 66. Service, R. F. (1998). "Superstrong nanotubes show they are smart, too." Science 281(5379), 940942. Sharma, A., Verma, N., Sharma, A., D eva, D., and Sankararamakrishnan, N. (2010). "Iron doped phenolic resin based activated carbon micro and nanoparticles by milling: Synthesis, characterization and application in arsenic removal." Chemical Engineering Science, 65(11), 35913601. Shon, H. K., Vigneswaran, S., and Snyder, S. A. (2006). "Effluent organic matter (EfOM) in wastewater: Constituents, effects, and treatment." Critical Reviews in Environmental Science and Technology 36(4), 327374. Simon, F. G., Segebade, C., and Hedrich, M. (2003). "Behaviour of uranium in ironbearing permeable reactive barriers: investigation with U 237 as a radioindicator." Science of the Total Environment 307(13), 231238. Skinner, K., Wright, N., and Porter Goff, E. (2007). "Mercury uptake and accumulation by four species of aquatic plants." Environmental Pollution 145(1), 234237. Skogerboe, R. K., and Wilson, S. A. (1981). "Reduction of ionic species by fulvic acid." Analytical Chemistry 53(2), 228232. Slowey, A. J., and Brown, J., Gordon E. (2007). "Transformations of mercury, iron, and sulfur during the reductive dissolution of iron oxyhydroxide by sulfide." Geochimica et Cosmochimica Acta, 71(4), 877894.
139 Steffen, A., Douglas, T., Amyot, M., Ariya, P., Aspmo, K., Berg, T., Bottenheim, J., Brooks, S., Co bbett, F., Dastoor, A., Dommergue, A., Ebinghaus, R., Ferrari, C., Gardfeldt, K., Goodsite, M. E., Lean, D., Poulain, A. J., Scherz, C., Skov, H., Sommar, J., and Temme, C. (2008). "A synthesis of atmospheric mercury depletion event chemistry in the atmosphere and snow." Atmospheric Chemistry and Physics 8(6), 14451482. Stumm, W., and Morgan, J. J. (1996). "Aquatic Chemistry Chemical Equilibria and Rates in Natural Waters." Wiley Interscience, New York, Ch 10. Su, C. M., and Puls, R. W. (2001). "Arsenate and arsenite removal by zerovalent iron: Kinetics, redox transformation, and implications for in situ groundwater remediation." Environmental Science & Technology 35(7), 1487 1492. Su, C. M., and Puls, R. W. (2003). "In situ remediation of arsenic in simulated groundwater using zerovalent iron: Laboratory column tests on combined effects of phosphate and silicate." Environmental Science & Technology 37(11), 25822587. Till, B. A., Weathers, L. J., and Alvarez, P. J. J. (1998). "Fe(0) supported autotrophic denitrification." Environmental Science & Technology 32(5), 634639. Uchimiya, M., and Stone, A. T. (2009). "Reversible redox chemistry of quinones: Impact on biogeochemical cycles." Chemosphere, 77(4), 451458. Ullrich, S. M., Tanton, T. W., and Abdrashitova, S. A. (2001). "Mercury in the aquatic environment: A review of factors affecting methylation." Critical Reviews in Environmental Science and Technology 31(3), 241293. US.EPA. (1997). "Mercury Study Report to Congress." US EPA, Washington DC. US.EPA. (2005). "Clean Air Mercury Rule." US EPA. Valenzuela, A., and Ftyas, K. (2002). "Mercury Management in Small Scale Mining." Minig Environmental Management 6(10). W.H.O. (1990). "Environmental Criteria 101: Methylmercury." World Health Organization, Gen eva, Switzerland. W.H.O. (1991). Inorganic Mercury. Environmental Health Criteria. ." World Health Organization Geneva, Switzerland.
140 Wang, Q., Kim, D., Dionysiou, D., Sorial, G., and Timberlake, D. (2004). "Sources and Remediation for Mercury Contaminati on in Aquatic Systems A Literature Review." Environmental Pollution, 131, 323336. Warner, K. A., Bonzongo, J. C. J., Roden, E. E., Ward, G. M., Green, A. C., Chaubey, I., Lyons, W. B., and Arrington, D. A. (2005). "Effect of watershed parameters on merc ury distribution in different environmental compartments in the Mobile Alabama River Basin, USA." Science of The Total Environment 347(1 3), 187207. Warner, K. A., Roden, E. E., and Bonzongo, J. C. (2004). "Microbial mercury transformation in anoxic fres hwater sediments under ironreducing and other electronaccepting conditions (vol 37, pg 2153, 2003)." Environmental Science & Technology 38(1), 352352. Watras, C. J., Bloom, N. S., Claas, S. A., Morrison, K. A., Gilmour, C. C., and Craig, S. R. (1995). "Methylmercury Production in the Anoxic Hypolimnion of a Dimictic Seepage Lake." Water Air and Soil Pollution 80(14), 735745. Weber, J. H. (1993). "Review of possible paths for abiotic methylation of mercury(II) in the aquatic environment." Chemosphere 26(11), 20632077. Weisener, C. G., Sale, K. S., Smyth, D. J. A., and Blowes, D. W. (2005). "Field column study using zerovalent iron for mercury removal from contaminated groundwater." Environmental Science & Technology 39(16), 63066312. Westerhoff, P. and James, J. (2003). "Nitrate removal in zerovalent iron packed columns." Water Research, 37(8), 18181830. Wilkin, R. T., and McNeil, M. S. (2003). "Laboratory evaluation of zerovalent iron to treat water impacted by acid mine drainage." Chemosphere, 53(7), 715725. Wilkin, R. T., Su, C. M., Ford, R. G., and Paul, C. J. (2005a). "Chromium removal processes during groundwater remediation by a zerovalent iron permeable reactive barrier." Environmental Science & Technology 39(12), 45994605. Wilkin, R. T., Su, C. M., Ford, R. G., and Paul, C. J. (2005b). "Long term geochemical behavior of a zerovalent iron permeable reactive barrier for the treatment of hexavalent chromium in groundwater." Geochimica Et Cosmochimica Acta 69(10), A264A264. Xu, Y. H., an d Zhao, D. Y. (2007). "Reductive immobilization of chromate in water and soil using stabilized iron nanoparticles." Water Research, 41(10), 21012108.
141 Yang, G. C. C., and Lee, H. L. (2005). "Chemical reduction of nitrate by nanosized iron: Kinetics and pat hways." Water Research, 39(5), 884894. Yin, C. Y., Aroua, M. K., and Daud, W. M. A. W. (2007). "Review of modifications of activated carbon for enhancing contaminant uptakes from aqueous solutions." Separation and Purification Technology 52(3), 403415. Yuan, C., and Lien, H. L. (2006). "Removal of arsenate from aqueous solution using nanoscale iron particles." Water Quality Research Journal of Canada, 41(2), 210215. Zavoda, J., Cutright, T., Szpak, J., and Fallon, E. (2001). "Uptake, selectivity, and in hibition of hydroponic treatment of contaminants." Journal of Environmental EngineeringAsce 127(6), 502508. Zhang, H., Jin, Z. H., Han, L., and Qin, C. H. (2006). "Synthesis of nanoscale zerovalent iron supported on exfoliated graphite for removal of nitrate." Transactions of Nonferrous Metals Society of China 16, S345 S349. Zhang, H., and Lindberg, S. E. (2001). "Sunlight and iron(III) induced photochemical production of dissolved gaseous mercury in freshwater." Environmental Science & Technology 35( 5), 928935. Zhang, W. X. (2003). "Nanoscale iron particles for environmental remediation: An overview." J. Nanopart. Res. 5(null), 323.
142 BIOGRAPHICAL SKETCH Julianne D Vernon was born in the Bronx, NY USA in 1981. She received her bachelors degree at T he City College of New York in chemical e ngineering. After which she worked for the New York City Department of Environmental Protection as a Process Engineer in the Bureau of Wastewater. After which, s he received her doctorate of philosophy in Environmental Engineering at the University of Florida under the guidance of Dr. Jean Claude Bonzongo in Spring 2010.