xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EKEY93UBA_ZR1E5J INGEST_TIME 2014-10-03T21:42:38Z PACKAGE UFE0046156_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
1 INSIGHTS INTO THE MECHANISMS OF IRON REDUCTIVE DISSOLUTION IN VADOSE ZONE SOILS AND IMPLICATIONS FOR LANDFILL ACTIVITIES: PREDICTING THE POTENTIAL FOR GROUNDWATER POLLUTION By AKUA BONSU OPPONG ANANE A DISSERTATION PRESENTE D TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2014
2 2014 Akua Bonsu Oppong Anane
3 T o Mum, D ad, Abena and Grandma
4 ACKNOWLEDGMENTS I thank Dr Jean Claude Bonzongo for the discussions and wealth of knowledge I have gained from him as well as all the members of his present and past research group especially Dr Augustine Donkor and his family I would like to thank my committee members, Drs Willie Harris, Timothy Townsend and Joseph Delfino for their helpful suggestions and the use of their laboratories during my research. I thank the staff at the Environmental Engineering Sciences department es pecially Melissa Centurion and Randy Switt for all of their assistance. I thank my parents, Dr Kwame Oppong Anane and Mrs Felicia Oppong Anane for their encouragement and support. I thank my sister, Dr Abena Oppong Anane and her husband Mr Evans Abbey for always being there for me. I thank my grandmother, Comfort Kyeremah for being my rock. I thank the Corry Village staff of the Graduate and Family Housing Department, especially Thomas Germain, Evelyn Jackson, Patricia Jordan, Lesa Boykin, Jason Fraser Nash Sandra Orozco, Lindsay Laytner Shauna Davis and all of the community and student assistants for their support. I thank all the friends at the University of Florida who have supported and shared this journey with me, especially Drs Mariela Rodriguez, Kwa me Sefah, Edmund Azah, Kwabena Amponsah Manager, Lucy Ngatia, Abraham Boateng, Pam Monterolla and Joy Guingab Darina Palacio and Miguel Lugo I thank Jeremy Long, the Ecclestons, Ingrims, Adesogan Beth, Dr Ethel April Owusu, Pastor Mac Wilkins and the co ngregation at the Family Church. I thank members of the Ghanaian community in Gainesville and the Osei family who have helped me to feel at home here a nd my friends in Ghana especially my Ridge Church f amily, the staff at Water Research Institute, Linda Du nn and the Quist Therson family, my extended family in the UK and my former classmates in Australia.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURES ................................ ................................ ................................ .......... 9 LIST OF ABBREVIATIONS ................................ ................................ ........................... 12 ABSTRACT ................................ ................................ ................................ ................... 14 CHAPTER 1 IRON REDUCTIVE DISSOLUTI ON IN SOILS/SEDIMENTS AND GROUNDWATER CONTAMINATION: RESEARCH PROBLEM STATEMENT ..... 16 Introduction ................................ ................................ ................................ ............. 16 Problem Statement ................................ ................................ ................................ 17 Organization of Dissertation ................................ ................................ .................... 24 2 BIOGEOCHEMISTRY OF I RON OXIDES IN SOILS: A REVIEW ........................... 29 Iron as an Important Nutrient ................................ ................................ .................. 29 Occurrence of Iron in the Environment ................................ ................................ ... 29 Iron Oxides ................................ ................................ ................................ ....... 30 Properties of Iron Oxides ................................ ................................ .................. 31 Iron Oxides in Soils ................................ ................................ ................................ 33 Background Sources of Iron Oxides ................................ ................................ 33 Anthropogenic Sources of Iron Oxides in Soils ................................ ................ 35 Mechanisms of Dissolution of Iron Oxides ................................ .............................. 36 Dissolution by Protonation ................................ ................................ ................ 37 Dissolution by Complexation ................................ ................................ ............ 38 Reductive Dissolution of Iron Oxides ................................ ................................ 39 Biotic and Abiotic Reductive Dissolution of Iron Oxides ................................ .......... 41 Biotic Reductive Dissolution of Iron Oxides ................................ ...................... 42 Abiotic Reductive Dissolution ................................ ................................ ........... 44 Potential for Soil Fe (III) Oxides to Act as an Fe Source of Groundwater Contamination ................................ ................................ ............................... 44 3 IRON REDUCTIVE DISSO LUTION IN SOILS AS A FUNCTION OF ORGANIC CARBON TYPES AND DEG REE OF CRYSTALLIZATI ON OF IRON OXIDE MINERALS : IMPLICATIONS FOR G ROUNDWATER POLLUTION BY LANDFILLS ................................ ................................ ................................ ............. 46 Introduction ................................ ................................ ................................ ............. 46
6 Materials and Methods ................................ ................................ ............................ 49 Vadoze Soil Sample Collection and Handling ................................ .................. 49 Characterization of Collected Vadose Zone Soil Samples ............................... 51 Organic Compounds Used as Energy Source and Microbial Reductive Dissolution of Soil Fe oxide Minerals ................................ ................................ ... 53 Iron Reductive Dissolution Experiments ................................ ........................... 55 Iron Analysis ................................ ................................ ................................ ..... 56 Resul ts and Discussion ................................ ................................ ........................... 57 Physicochemical Characteristics of Collected Soils ................................ ......... 57 Fe reductive Dissolution in Soils Treated with Glucose as Model Organic Compound ................................ ................................ ................................ ..... 58 The ph E h Theory on the Conditions of Formation of FeS/FeS 2 and FeCO 3 .... 59 Characteristics of Used Land fill Leachate and its Ability to Promote Fe R eductive Dissolution ................................ ................................ .................... 63 Relevance of the Chemistry of Used Sources of Organic Carbon in the Reductive Dissolution of Soil Iron ................................ ................................ .. 64 Conclusions on Fe Reductive Dissolution Experiments with Organic Matter .......... 66 4 ASSESSING THE POTENTIAL OF ABIOTIC IRON REDUCTIVE DISSOLUTION IN VADOZE ZONE SOILS USING SULFIDE AS ELECTRON DONOR ................................ ................................ ................................ .................. 82 Introduction ................................ ................................ ................................ ............. 82 Materials and Methods ................................ ................................ ............................ 85 Vadoze Soil Sample Collection and Handling ................................ .................. 85 Characterization of Collected Vadose Zone Soil Samples ............................... 86 Iron Reductive Dissolution Experiments with Dissolved Sulfide ....................... 87 Preparation of Sulfide Solutions ................................ ................................ ....... 88 Sulfide Analysis and Sulfide Speci ation ................................ ........................... 89 Iron Analysis ................................ ................................ ................................ ..... 90 Statistical Analyse s ................................ ................................ .......................... 91 Results and Discu ssion ................................ ................................ ........................... 92 Disappearance of Dissolved Sulfide from Solution ................................ ........... 92 Sulfide Induced Iron Reductive Dissolution ................................ ...................... 94 Conclusions ................................ ................................ ................................ ............ 97 5 PREDICTING THE POTEN TIAL OF VADOSE ZONE SOIL IRON TO UNDERGO REDUCTIVE DI SSOLUTION ................................ ............................. 109 Introduction ................................ ................................ ................................ ........... 109 Materials and Methods ................................ ................................ .......................... 111 Hematite as Source of Iron Characterization ................................ ............... 111 Fe Reductive Dissolution Experiments ................................ ........................... 112 Analysis of Fe ................................ ................................ ................................ 112 Predictive Tools ................................ ................................ .............................. 113 Results and Discussion ................................ ................................ ......................... 114 Hematite as Source of Iron Characterization ................................ ............... 114
7 Fe Reductive Dissolution from Hematite as a Function of Key Environmental Parameters ................................ ................................ .......... 114 Conclusions ................................ ................................ ................................ .......... 118 6 GENERAL CONCLUSIONS AND RE COMMENDATIONS FOR FUTURE WORK ................................ ................................ ................................ ................... 125 APPENDIX A SUPPLEMENTARY TABLES ................................ ................................ ............... 130 B SUPPLEMENTARY FIGURES ................................ ................................ ............. 132 LIST OF REFERENCES ................................ ................................ ............................. 141 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 153
8 LIST OF TABLES Table page 3 1 Physiochemical characteristics of collected vadose zone soil sample s ............. 69 4 1 pH and analytically determined initial concentrations of sulfide in the soil slurries prepare d in triplicates for each of the tested soils. ................................ 99 4 2 Rates of disappearance of dissolved sulfide (mg S.L 1 .day 1 ) in soil slurries treated with Na 2 S.9H 2 O ................................ ................................ ................... 100 4 3 Rates of iron released to the aqueous phase and present as dissolved Fe(II) 2 S.9H 2 O .... 101 4 4 R concentrations (i.e. sum of Fe(II)aq and Fe(II)S) in the aqueous phase of soil mixed with reduce d sulfur added as Na 2 S.9H 2 O ................................ ............. 102 5 1 Concentrations of iron released per kg of soil treated with landfill leachate as amorphous (AAO) iron using the 12 tested soil samples. ................................ 120 A 1 Rates of dissolved Fe(II) (mg/kg .day) released in soil slurries ........................ 130 A 2 Rates of dissolved Fe(II) (mg/kg.day) released in soil slurries expressed in percentages (%) o f the soil total Fe concentrations. ................................ ......... 131
9 LIST OF FIGURES Figure page 1 1 Temporal trends of iron (Fe) and chloride (Cl ) in groundwater samples at t he Aucilla Area Solid Waste F acility ................................ ................................ ....... 26 1 2 Vadose zone soils impacted by unlined C & D landfills ................................ ..... 27 1 3 Vadose zone soils impacted by lined municipal a nd hazardous solid wastes ... 28 3 1 Illustration of vadose zone soil sample collection strateg y ................................ 68 3 2 XRD spectra of soils samples collected from the Klondike Landfill Site ............ 70 3 3 XRD spectra of soils. A) Sample 3 collected from the site of the New River Regional Landfill. B) Sample 4 collected from the site of the Alachua County Landfill ................................ ................................ ................................ ............... 71 3 4 Rates of Fe(II) release from the tested soils as a function of organic carbon (added as gl ucose) concentrations ................................ ................................ ..... 72 3 5 Rates of Fe(II) release from the tested soils when treated wi th 36g of organic carbon/L, added as glucose. ................................ ................................ .............. 73 3 6 Rates of Fe(II) release from the tested soils when treated with 18g of organic carbon/L, added as glucose ................................ ................................ .............. 74 3 7. Rates of Fe(II) release from the tested soils when treated with 9g of organic carbon/L, add ed as glucose ................................ ................................ .............. 75 3 8: Example XRD spectra of control and glucos e (18g C/L) treated soil sample 3 after incubation under anaerobic condition s ................................ ....................... 76 3 9 Fluorescence excitation emission matrix (EEM) spectrum of the landfill leachate used as source of organic ca rbon in Fe(II) release experi ments ......... 77 3 10 Rates of Fe(II) release from different soils, each exposed to a final concentration of 1.125g of organic carbon per liter, added as la ndfill leachate . 78 3 11 Concentration of Fe(II) released by soil treated with landfill leachate (1.125g C/L) as a function of d ifferent soil Fe portions ................................ ................... 79 3 12 Correlation between rates of Fe(II) (expressed as fraction of total Fe content) released by each of the 12 tested soils following the addition of 9g C/L glucose and 1.125g C/L landfill leachate. ................................ ........................... 80 3 13 Relationship between total (CDB) and amorphous (AAO) iron concentra tions in analyzed soils ................................ ................................ ................................ 81
10 4 1 Temporal concentration trends of the different fractions of Fe(II ) released from soil treated with a solution of reduced sulfur added as Na 2 S.9H 2 O ......... 103 4 2 Rates of iron reductive dissolution from zone 1 soils expressed as mg Fe.kg 1 day 1 and rates of disapp earance of reduced S from solution in mg S.L 1 day 1 ................................ ................................ ................................ ................ 104 4 3 Rates of iron reductive dissolution in zone 3 soil samples expressed as mg Fe.kg 1 day 1 and rates of disappearance of sulfide in mg S .L 1 day 1 ................ 105 4 4 Rates of iron reductive dissolution calculated using data obtained after a r eaction time of t=30 minutes ................................ ................................ .......... 106 4 5 Rates of iron reductive dissolution calculated using data obtained after a reaction time of t=60 minu tes ................................ ................................ .......... 107 4 6 Rates of iron reductive dissolution calculated using data obtained after a r eaction time of t= 90 minutes ................................ ................................ .......... 108 5 1 XRD spectra of the hematite sample used in reductive dissolution experiments to help develop a Fe dissolution model as a function of organic matter, pH, an d ionic strength. ................................ ................................ .......... 121 5 2 Fe released from hematite exposed to aqueous solutions containing increasing concentrations of organic carbon added as glucose in the presence of a consortium of bacte ria from an anaerobi c digester ................... 122 5 3 Fe released from hematite exposed to aqueous solutions of increasing concentrations of protons (i.e. decreasing pH) and containing bacteria from a n anaer obic digester ................................ ................................ ...................... 123 5 4 Fe released from hematite exposed to aqueous solutions of increasing ionic strengths and containing bacteria from an a naerobic digester ........................ 124 6 1 Comparative flow chart of the experimental (left) and modeling (right) studies conducted in this research ................................ ................................ ................ 129 B 1 XRD spectra of Sample 5 collected at loca tion 1 of the UF/IFAS Plant Science Research and Education U nit ................................ ............................. 132 B 2 XRD spectra of samples collected at location 2 of the UF/IFAS Plant Science Rese arch and Education Unit ................................ ................................ .......... 133 B 3 XRD spectra of samples collected from location 1 at the Au stin Carey Memorial Forest ................................ ................................ ............................... 134 B 4 XRD spectra of sample 10 collected from the Ord way Swisher Biological Station ................................ ................................ ................................ ............. 135
11 B 5 XRD spectra of sample 11 collected from horizon E3 at location 1 of the Aus tin Carey Memorial Forest ................................ ................................ ......... 136 B 6 XRD spectra of sample 12 collected from horizon E3 at location 2 of the Austin Carey Memorial F orest ................................ ................................ ......... 136 B 7 Rates of iron reductive dissolution from zone 1 soils expres sed as mg Fe.kg 1 day 1 and rates of disappearance of reduced S from solution in mg S.L 1 day 1 ................................ ................................ ................................ ....................... 137 B 8 Rates of iron reductive dissolution from zone 2 soils expressed as mg Fe.kg 1 day 1 and rat es of disappearance of reduced S from solution in mg S.L 1 day 1 ................................ ................................ ................................ ....................... 138 B 9 Rates of iron reductive dissolution from zone 2 soils expressed as mg Fe.kg 1 day 1 and rates of disappearance of reduced S from solution in mg S.L 1 day 1 ................................ ................................ ................................ ....................... 139 B 10 Rates of iron reductive dissolution from zone 2 soils expressed as mg Fe.kg 1 day 1 and rates of disappearance of reduced S from solution in mg S.L 1 d ay 1 ................................ ................................ ................................ ........................ 140
12 LIST OF ABBREVIATIONS ACMF Austin Carey Memorial Forest AMD Acid Mine Drainage A s Arsenic C a Calcium Cd Cadmium CDB Citrate Dithionite Bicarbonate Cu Copper Cr Chromium DOC Dissolved O rganic Carbon EPA Environmental Protection Agency FDEP Florida Department of Environmental Protection Fe Iron FeS Iron Monosulfide HMWOM High molecular weight organic matter LMWOM Low molecular weight organic matter LOI Los s o n Ignition MCL Maximum Contaminant Level O Oxygen OC Organic Carbon OM Organic Matter PA Pine Acres Pb Lead S Sulfur SOC Soil Organic Carbon
13 SOM Soil Organic Matter TOC Total Organic Carbon XRD X ray Diffraction Zn Zinc
14 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy INSIGHTS INTO THE MECHANISMS OF IRON REDUCTIVE DI SSOLUTION IN VADOSE ZONE SOILS AND IMPLICATIONS FOR LANDFILL ACTIVITIES: PREDICT ING THE POTENTIAL FOR GROUNDWATER POLLUTION By Akua Bonsu Oppong Anane May 2014 Chair: Jean Claude J. Bonzongo Major: Environmental Engineering Sciences Anomalously high iron (Fe ) concentrations have been measured in groundwater samples collected from monitoring wells downstream of several landfill units in Florida. Based on monitoring data, vadose zone soils and aquifer sediments impacted by landfill s could be sources of Fe that pollutes the groundwater and not landfilled wastes The refore, the objectives of this study were to investigate the reductive dissolution of soil Fe as a function of biotic and abiotic processes ; and to lay the groundwork for the development of a geochemi cal solubility model for Fe Soil samples were collected from different locations in North Florida using a sampling strategy that include d samples with a gradient in the degree of Fe mineral crystallization. After characterization, soil samples were used in biotic ( using bacteria from an anaerobic digester ) and abiotic (using sulfide as electron donor) Fe reductive dissolution batch studies Fe(II) was released from all soils in concentrations that exceeded the secondary drinking water limit of 0.3 mg/L w hen treated with bacteria and organic carbon (OC) or with sulfide but under abiotic conditions M icrobial respiration of OC in soils appears to be a major pathway and could lead to Fe reductive dissolution in
15 Fe rich soils that interact with OC rich waters or leachates The r ates of Fe reductive dissolution were p ositively correlated with soil Fe content when glucose was used as the OC source while no relationship was obvious when landfill leachate was used as OC Sulfide driven Fe reductive dissolution wa s positively correlated with soil Fe content However, despite the initial high release of Fe(II) into the aqueous phase in soil spiked with sulfide temporal trends of Fe reductive dissolution rates suggest that this pathway may not be as significant as t he biotic process due likely to the precipitation of solid FeS species Finally, batch studies were conducted using hematite a n Fe oxide mineral, to investigate the effects of pH, OC and ionic strength on Fe reductive dissolution rates O btained data were used to develop a geochemical solubility model for soil Fe; which was then validated using soil samples. However, an efficient fine tuning and validation of the model was not fully accomplished as more experimental and field data are still needed.
16 CHAP TER 1 IRON REDUCTIVE DISSOLUTION IN SOILS /SEDIMENTS AND GROUNDWATER CONTAMINATION: RESEARCH PROBLEM STATEMENT Introduction Metals are naturally present in soils in background levels that originate from metal concentrations of the parent rock from which the soil was formed (McLean and Bledsoe, 1992) But anthropogenic inputs from pesticides, fertilizers, biosolids, sewage sludge, manure elevate and landfill leachate elevate the soil metal content to contaminant levels, exceeding established regulatory limits (Silveira et al., 2003) In addition, anthropogenic activities that pollute natural systems with toxic metals could alter key physicochemical properties of soils such as pH, redox potential, and organic dissolution and mobility of metals previously bound to different mineral and organic phases (Charlatchka and Cambier, 2000; Chuan et al., 1996; del Castilho et al., 1993; Kashem and Singh, 2001) This would then increase the bioavailability of metals, especially sparingly soluble metal s (Schwertmann, 1991) But an adverse consequence is migration of the metals found in the soil solution through the soil vadose zone to the aquifer below resulting in groundwater contamination (Dube et al., 2001; Kashem and Singh, 2001; Silveira et al., 2003) This process could lead to increased concentrations of toxic metals such as arsenic (As) and ch romium (Cr) in groundwater, which are considered harmful to both human health and ecosys tem functions. Lead (Pb), cadmium (Cd), calcium (Ca), zinc (Zn) and copper (Cu) are examples of land applied metals that have been studied to determine their partitioning between solid and aqueous phases as well as their distribution (or speciation) amongs t the fractions. T he leachability and mobility of these
17 metals has been used for the assessment of environmental risk including studies on the potential of these metals to contaminate groundwater, become bioavailable to organisms, and cause adverse biological effects (Beesley et al., 2010; del Castilho et al., 1993; Tipping et al., 2003) Although most of these metals are bound onto soil iron (Fe) oxide minerals (Cajuste et al., 2000; Sastre et al., 2001) and are released into the soil solution with Fe as a result of the reductive dissolution of the Fe oxide minerals, there is limited research on the mechanisms of Fe reductive dissolution in vadose zone soils, and on the potential of this biogeochemical process to behave as a significant source of metal contamination of aquifers, namely in soils impacted by landfill activities. Problem Statement Iron (Fe) is an important micronutrient for humans and plants (Kraemer, 2001) In humans, it is found in the blood as hemoglobin which aids in the transport of oxygen throughout the body (Wessling Resnick, 1999) and in plants, it aids in the production of chlorophyll (Subcommittee on Iron, 19 79) crust, Fe is naturally abundant in soils, rocks, and in both marine and freshwater sediments (Iron, 1979; Ussher et al., 2004) Fe naturally occurs in soils predominantly as Fe(III) oxide minerals or as a component of clay minerals (I ron, 1979; Minyard and Burgos, 2007; Subcommittee on Iron, 1979) and the reported average concentration of Fe in soils is 38 g kg 1 or 38 000 ppm (Lindsay, 1979). Studies conducted in the United States tend to suggest that the typical soil Fe concentratio n is around 40 g kg 1 (Ma et al., 1997) Fe(III) oxides have important functions in soils, they provide surfaces for the sorption of metals and other inorganic compounds, as well as organic compounds (Pedersen et al., 2005) They serve as the most abundant terminal electron acceptor
18 during the degradation of organic matter (Larsen and Postma, 2001) But of concern is the reductive dissolution of Fe(III) oxide minerals and the mobility o f released Fe(II) as well as that of several other toxic metals previously locked in the structure of Fe oxide minerals (Davranche and Bollinger, 2000) A case in point is the issue of arsenic ( As ) contamination of groundwater resources, particularly pronounced in the Indian sub continent, and the severe health epidemic that resulted fro m the use of As contaminated water for drinking purposes. Reported health effects included skin lesions and skin cancer and Fe reductive dissolution was believed to be main mechanism responsible for the As that was release d in to the groundwater (Nickson et al., 2000; Sracek et al., 2004) A major problem associated with the discharge of Fe(II) containing groundwater into surface waters including lakes, streams, rivers and springs is the oxidation of Fe(II) and the formation of Fe (hydr)oxides a phenomenon common when reduced water containing dissolved Fe(II) comes into contact with oxygen (Parisio et al., 2006) This process would result in colored surface waters with implications for domestic uses such as stained clothes from the use of Fe contaminated water during laundry, deposition of Fe oxides in pipes and plumbing fixtures which may result in clogging, decreased water quality due to development of a metallic taste, odor and turbidity and the creation of Fe (hy dr)oxides Ultimately, the formed amorphous Fe (hydr)oxides could behave as a sink for dissolved toxic trace metals, building up over time, with long term toxic implications to aquatic o rganisms. In fact Fe concentrations as high as 60 mg/L to 2680 mg/L have been reported in surface waters (Gurzau et al., 2003)
19 Due to these problems, Fe is regulated by the Environmental Protection Agency (EPA) as a secondary drinking water contaminant at a maximum contaminant level (MCL) of 0.3 mg/L (USEPA, 2013) Additional Fe standards enforced in Florida include 1 mg/L i n surface freshwater water, 0.3 mg/L in mar ine surface waters and 3 mg/L in groundwater of low yield and poor quality which is not used for drinking purposes (F.A.C., 62 777) Although Fe is required as an essential micronutrient by living organisms, excess consumption of Fe has recognized toxicity. In humans, Fe is stored in the kidney and liver and ingestion of excess amounts of Fe results in damage to these organs, acu te Fe poisoning, and death in severe cases Due to these effects, a health criterion of 4.2 mg/L for Fe in groundwater has been adopted by the Florida Department of Environmental Protection (FDEP) (Ochoa et al., 2004) A part from humans, a quatic species, including periphyton, fish and invertebrates, have also been negatively impacted by high concentr ations of Fe in the groundwater, limited quantities and varieties of aquatic species have been observed and the high mortality of fish has been attributed to the ingestion and blockage of fish gills by Fe (hydr)oxides precipitates which limit respiration (Linton et al., 2007) Fe reducing zones within leachate plumes and elevated Fe(II) concentrations have been observed in groundwater in different parts of the world, inc luding watersheds impacted by the Vejen and Grindsted Landfills in Denmark, the Borden Landfill in Canada (Christensen et al., 2001) the Norman Landfill in Okl ahoma, and the United States Geological Survey (USGS) Toxic Substances Hydrology site near Bemidji, Minnesota (Cozzarelli et al., 2000; Lyngkilde and Christensen, 1992b; Tuccillo et al.,
20 1999) In these settings, t he Fe contamination observed in groundwater has been attributed mainly to landfill leachates. The latter is a liquid solution which is produced as rainwater passes through landfill waste and is composed of the following categories of compounds: dissolved o rganic matter (e.g. CH 3 COOH and fulvic acids), inorganic compounds (e.g. Fe and Mg), heavy metals (e.g. Pb and Zn) and xenobiotic organic compounds (e.g. benzene and phenols) (Christensen et al., 2001; Wizniowski et al., 2006) In fact, landfill leachates ca n contain anywhere between 3 to 5500 mg/L of Fe, with an average concentration o f 780 mg/L in the acid phase or only 15 mg/L of Fe in the methanogenic phase (Kjeldsen et al., 2002) High concentrations of Fe in groundwater have been attributed to the failure of lea chate collection systems designed to retain the leachate or the use of unlined landfills (Lyngkilde and Christensen, 1992a) A second source of Fe found in groundwater is Fe(II) from aquifer sediments, released by microbial reductive d issolution fuelled by the organic compounds present in landfill leachates (Lyngkilde and Christensen, 1992a) At the Vejen Landfill in Denmark, lower concentrations of Fe(III) were extracted from sediments near the landfill compared to the higher concentrations of Fe(III) in the sediments down gradient of the landfills proving that the availability of higher amounts of landfill leachate at the source of contamination resulted in increased production of Fe(II) as compared to the lower Fe (II) production at distances further away from the landfill, where the availability of landfill leachate is not as high (Heron and Christensen, 1995) In Florida, the occur rence of high conce ntrations of Fe in groundwater h as been determined by the analysis of water samples collected from monitoring wells at various landfill sites. At locations where Fe contaminated groundwater and surface water get
21 mixed, a thick muddy wate rs are formed. But groundwater monitoring data taken from the monitoring wells at these landfill sites gave no indication that landfill leachate was the source of Fe since dissimilar spatio temporal trends were observed between the concentrations of Fe and conservative tracers such as chloride (Cl ), a landfill leachate contamination indicator. Chloride is commonly used as an indicator of landfill leachate contamination in groundwater since it is not affected by redox and any contaminant which originates fr om landfill leachate should show the same temporal behavior as chloride (Christensen et al., 2001; Lyngkilde and Christensen, 1992a) An example of this incident occurred at the Aucilla Area Solid Waste Facility in Madison County, a lined Class 1 landfill unit, where the groundwater data taken from the monitoring wells showed a decreasing trend in the Cl concentration while the Fe concentration increased temporally as shown in Figure 1 1. The lack of correlation be tween these two parameters is a likely indication that landfill leachate is not the source of the Fe measured in the groundwater. Methods used and/or proposed for the removal of Fe from contaminated groundwater include the combination of aeration or chemic al oxidation processes (using chlorine or potassium permanganate) with filtration (Choo et al., 2005) ion exchange and the use of limes tone (Wang et al., 2012) Unfortunately, some of these methods cause fouling (or clogging) of the Fe removal medium (such as the ion exchange membranes) which then require further treatment or regeneration to remove retained Fe contaminants prior to reuse (Chaturvedi and Dave, 2012) Furthermore, some of these m ethods do not completely eliminate Fe from treated groundwater and may have to be used repeatedl y to remove Fe
22 are labor intensive, time consuming and costly. Most of the research on Fe groundwater c ontamination has focused primarily on the optimization of these remedial methods but these methods do not prevent the reoccurrence of Fe contamination in groundwater but rather serve as late corrective measures. In addition, t here is limited understanding on the mechanisms of Fe reductive dissolution in vadose zone soils or on the impact of landfill activities on the release of Fe(II) Fe reductive dissolution has mostly been studied with synthetic Fe(III) oxides but these studies do not consider the addit ional effect of the soil matrix or biogeochemical parameters (soil pH, organic matter content, soil Fe quantity, soil Fe mineralogy) on the release of Fe(II) These parameters are important and n eed to be taken into consideration to gain a better understan ding of how interactions between Fe(III) oxides and the other soil components affect the extent of Fe reductive dissolution. Based on studies conducted at several landfill sites in Florida, the source of Fe measured in groundwater could be Fe (III) oxide m inerals that are found in either vadose soils underneath these landfills or/and aquifer sediments This study focuses on the potential role of vadose zone soils only as source of Fe(II) detected in groundwater impacted by landfill units. The rationale of t his assumption is as follows. First, in vadose zone soils impacted by unlined C & D landfills, organic matter (wood being the main source of organic matter) and reduced sulfur compounds ( from reduction of SO 4 in drywall) would drive the reductive dissoluti on of soil iron minerals (Figure 1 2). In these locations impacted by unlined landfill activities, microbial processes occurring in these soils during the degradation of organic matter from inputs of landfill leachate results in development of anoxic condi tions and utilization of
23 available terminal electron acceptors, including soil Fe(III) oxides which would result in the reductive dissolution of Fe(III) and release of Fe(II) to the soil solution and subsequently to groundwater (Heron and Christensen, 1995) In addition to the above redox reaction, microbial respiration of organic matter by sulfate reducing bacteria ( SRB ) can lead to the formation of reduced sulfur compounds such as S 2 depending on the pH of solution The latter could behave as electron donor and reduce soil Fe(III) oxide minerals ( Figure 1 2 ). Overall, t hese two processes would lead to Fe contamination of groundwater ; and in this specific case of C&D landf ills Fe(II) water pollution can be linked to the landfill using geochemical tracers o f landfill leachate pollution. More intriguing is groundwater pollution by Fe at sites impacted by lined municipal and h azardous solid wastes landfills with no detectable evidence of the liner malfunction. It is hypothesized here that the presence of the liner cuts off soil aeration and rainwater/leachate infiltration underneath the landfill unit resulting in a strongly reduced soil environment that develop s underneath th e liner over time ( Figure 1 3A ). Next, upward changes in the water table level would bring organic rich groundwater in contact with terminal electron acceptors (e.g. Fe(III) oxide minerals) present in previously unsaturated soils ( Figure 1 3B ) This contac t would stimulate anaerobic microbial respiration in the newly saturated anoxic soils le ading to the reduction of soil Fe oxide minerals and dissolution of Fe(II) by iron reducing bacteria (FeRB) and that of sulfate by SRB Accordingly, Fe(II) measured in polluted groundwater under these specific landfill conditions could originate from (i) the biotic reduction of soil Fe(III) oxide minerals by FeRB and/or (ii) the oxidatio n of sulfide coupled with soil Fe(III) oxide
24 mineral reduction. The latter is conside red an abiotic reduction pathway since inorganic reduced sulfur compounds act as an electron donor. The ultimate and long term goal of this research is the development of a predictive model which can help determine the potential of soils t o release Fe and to contaminate groundwater under conditio ns of landfill operations. This dissertation, however, is limited to laboratory studies designed to lay the groundwork for the development of such a predictive model. S pecific objectives are therefore : (i) to invest igate the mechanisms of Fe reductive dissolution in soils under landfill imposed conditions and as a function of key biogeochemical parameters impacting Fe reductive dissolution in vadose zone soils ; and (ii) to initiate work on predicting the potential of a vadose zone soil to behave as a source of Fe that contaminates groundwater due to landfill imposed conditions. The overarching hypotheses is that Fe reduction dissolution occurs in soils due to the interactions o f landfill leachate (containing organic matter and reduced sulfur compounds) with soil Fe(III) oxide mineral s in landfill impacted soils. The working hypotheses dr iving this research are that Fe reductive dissolution of soil Fe(III) minerals occurs (i) as microbial catalyzed react i on s coupled wi th the oxidation of organic matter (microbial respiration); and/ or (ii) through reactions of landfill produced reduced sulfur compounds with soil Fe (III) oxide mineral s. Organization of Dissertation This dissertation is made of six separate chapters a nd the content of each the chapters is briefly summarized below. Chapter 1 is a general introduction. It presents the research problem statement objectives, and hypotheses; and it summarizes the need for this proposed research.
25 Chapter 2 is a literature review and presents the current state of knowledge with regard to the research questions raised. The focus is primarily on Fe oxide mineral s in soils and the different mechanisms that lead to their reduction and subsequent release of Fe(II) to aqueous sol utions K nowledge gaps are pointed out whenever necessary. Chapter 3 is the first chapter presenting the results of laboratory work related to the potential role of organic matter in Fe reductive dissolution as visually illustrated in Figures 1 2 and 1 3 This section of the dissertation includes the description of sample collection and handling and a presentation of the results of different laboratory studies on the reductive dissolution of iron as a function of organic matter concentrations and types, in cluding synthetic (e.g. glucose) and natural (e.g. landfill leachate ) as source s of organic carbon added to the soils to stimulate the microbial catalyzed reduction of Fe(III) oxide minerals While Chapter 3 focus on the role of biotic processes in the re ductive dissolution of Fe, Chapter 4 examines the significance of non biological or abiotic processes. The emphasis is on the role of dissolved sulfide compounds. In Chapter 5 results presented in Chapters 3 and 4 are used in combination with additional e xperimental data to lay the groundwork for development of predictive mathematical tools Chapter 6 is a general concl usion. It summarizes the mai n findings of this dissertation and outlines future research avenues that could help in the development of an efficient and accurate predictive tool
26 Figure 1 1 Temporal trends of iron (Fe) and chloride (Cl ) in groundwater samples at the Aucilla Area Solid Waste Facility. A) Collected from well 2S. B) Collected from well 7S. Adapted from Bonzongo, J ean Claude and Townsend, Timothy. 2008 Reductive Dissolution of Iron in Impacted Soils and Aquifers: Mechanisms and Development of a Predictive Model (Page.3, Figure 1 ). H ink ley Center for Solid and Haz ardous Waste Management. Florida. A B
27 CH n O + H n O CO 2 + n H + + n e Organic and Sulfide rich leachate Figure 1 2. Vadose zone soils impacted by unlined C & D landfills. Organic matter (wood being the main source of organic matter, represented above by the simplified general formula CH n O) and reduced sulfur compounds (with CaSO 4 in drywall being the main s ource of reduced sulfur compounds) would drive the reductive dissolution of iron. The combination of these biotic and abiotic processes leads to Fe contamination of groundwater. In this case of unlined C & D landfills, the observed water pollution by Fe ca n easily be linked to the landfill using geochemical tracers of landfill leachate pollution.
28 Figure 1 3. V adose zone soils impacted by lined munic ipal and hazardous solid wastes. T he presence of a liner cuts off soil aeration and leachate/ rainwater infil tration underneath the landfill unit A) A reducing envir onment would develop over time. B) In Florida, where aquifers ar e often not too far below the soil surface, changes in the water table level of a groundwater could enhance the anaerobic microbial res piration of soil microorganisms in previously unsaturated zones le ading to the reduction of soil Fe (III) oxide minerals and dissolution of FeII.
29 CHAPTER 2 BIOGEOCHEMISTRY OF I RON OXIDES IN SOILS: A REVIEW Iron as an Important Nutrient Iron (Fe) is an imp ortant micronutrient for humans, plants, bacteria and fungi (Kraemer, 2001) and is required for a variety of vital processes in living organisms including respiration and photosynthesis (Silver, 1993) In humans, it is found in the hemoglobin s tructure of the red blo od cells which are vital for transport of oxygen to various parts of the body and in plants, F e aids in the production of chlorophyll, an essential component of photosynthesis (Kim and Guerino t, 2007; Thomine and Vert, 2013; Yadavalli et al., 2012) A major health problem of Fe deficiency is anemia, a which is defined by low hemoglobin (or low red blood cell) levels and a decreased ab ility to transfer oxygen throughout the body (Barbosa et al., 2013; McLean et al., 2008) In plants, Fe deficiency results in chlor osis, a disease which causes yellowing and wilting of plant leaves and severely aff ects crop production with resultant low yield s or poor quality crops (Martnez Cuenca et al., 2013) Occurrence of Iron in the Environment Fe is widespread in the environment and can be found in the air as a result of emissions from either Fe metal or steel manufacturing plants and volcanic eruptions, in rocks and soils as Fe ores and Fe silicate/oxide minerals and in oceans and aquifers as Fe minerals and sediments including pyrite, FeS 2 (I ron, 1979; Subcommittee on Iron, 1979) The biogeochemical cyc ling of Fe is important for supply ing Fe to air, soil and water environments and it is also intrinsically connected to the cycling of sulfur (S),
30 carbon (C), oxygen (O) and nitrogen (N) in thes e environments also (Viollier et al., 2000; Zachara et al., 2001) Fe with the atomic number 26 and the atomic weight of 55.847, exists as a transition metal with several oxidation states but in soils and in natur al waters, the two most common oxidation states are Fe(II) [or ferrous iron] and Fe(III) [or ferric iron] which are prevalent u nder anoxic or oxic conditions respectively (Silver, 1993; Viollier et al., 2000) Fe like manganese (Mn) and aluminum (Al ) are the minerals which commonly form oxide minerals in soils (Dixon and Weed, 198 9) Fe(III) exists in soils mostly as insoluble Fe (IIII) oxide or Fe(III) silicate minerals. At pH 7, the majority of Fe(III) occurs in the form of Fe(III) oxides which have a reported solubility of ~10 9 M (Chipperfiel d and Ratledge, 2000; Luu and Ramsay, 2003) However, the solubility of Fe(III) increases as it forms complexes with ligands such as oxalate and citrate (Suter et al., 1988; Benner et al., 2002). Fe(II) on the other hand is usually found as the dissolved species in solution under anoxic conditions (Sulzberger et al., 1989; Viollier et al., 2000) but Fe(II) also has the ability to form insoluble minerals with carbonates, s ulfides and phosphates as siderite [FeCO 3 ] iron monosulfide [FeS] and vivianite [Fe 3 (PO 4 ) 2 .8H 2 O] respectively, under weakly acidic to neutral anoxic conditions (Shaked et al., 2004; Straub et al., 2001) Iron Oxides Fe oxides are natural soil and sedim ent constituents but synthetic forms of these oxides have also been produced in the laboratory (Cornell and Schwertmann, 2000) In the literature, the Fe oxide nomenclature is used when referring collectively to Fe oxides, hydroxides and oxyhydroxides (Pedersen et al., 2005) Fe oxides minerals contain an arrangement of a central Fe(III) [and/or Fe(II) in some cases] atom
31 surrounded by O and/or OH atoms in a basic octahedral [Fe (O/OH) 6 or FeO 6 ] or tetrahedral [FeO 4 ] unit structure (Schwertmann, 1991; Schwertmann and Cornell, 2003) Sixteen (16) types of Fe oxides have been identified bu t each mineral differs in terms of its structural arrangement (Schwertmann and Cornell, 2003; Straub et al., 2001) Fe oxides have characteristic coloring, sorption and redox properties which give s them widespread use in industry as pigments, sorbents, catalysts, additives in fertilizers or as raw materials for Fe metal or steel production (Dixon and Schulze, 2002; Schwertmann et al., 1999) Of environmental significance are the sorption and redox properties of Fe oxides which play an important role in controlling soil and water chemistries (Lovley, 1991; (Luu and Ramsay, 2003) Properties of Iron Oxides A wide array of Fe oxides exist with variations in the mineral crystallinity, stability, solubility, particle size and specific surface area (SSA) (Bonneville et al., 2009) The crystallinit y of Fe oxides ranges from the low crystalline or amorphous forms such as ferrihydrite Fe(OH) 3 FeOOH) and Fe 2 O 3 ) (Roden, 2004) The degree of crystallin ity of Fe oxides affects the extent of the Fe reductive dissolution, and although amorphous Fe oxides are more easily reduced because of their availability as terminal electron acceptors in microbially catalyzed reactions during anaerobic respiration; the reduction of crystalline Fe oxides has also been reported (Roden et al., 2000) Fe oxides undergo transformation to more stable mineral s upon aging. For example, ferrihy drite which is amorphous transforms to hematite, a highly stable crystalline mineral (Benner et al., 2002; Makris et al., 2005) Fe oxide transformation may also occur thro ugh che mical reactions that change the
32 mineral crystal structure, for example goethite forms through precipitation after the dissolution of ferrihydrite (Baltpurvins et al., 1996; Makris et al., 2005) Fe oxides have hig h specific surface areas due to the small mineral size which ranges from 5 to 150 nm (Schwertmann, 1991). The Fe oxide mineral size has a positive correlation with the mineral crystallinity and a negative correlation with SSA. An increase in the Fe oxide s ize has a resultant increase in the mineral crystallinity of the Fe but this is accompanied by a resultant decrease in the SSA of the mineral (Makris et al., 2005; Roden, 2004) It has be en demonstrated that the SS A of the smaller ferrihydrite mineral of 214 m 2 /g is about ten (10) times higher than the SSA of the bigger hematite mineral of 19 m 2 /g (Pedersen et al., 2005) Due to their hi gh SSA and reactive surfaces acquired from the presence of surface hydroxyl groups (Roden et al., 2004) iron oxides play an important role in th e sorption of compounds including metals such as Zn, Pb (Pedersen et al., 2005) inorganic anionic complexes (such as AsO 4 3 PO 4 3 CrO 4 3 ) and organic anions/ compounds (such as humic and fulvic acids) (Stipp et al., 2002) Unfortunately some of the sorbed compounds are toxic contaminants and the sorption capability of Fe oxides therefore plays an instru mental role in contaminant transport by increasing the mobility of sorbed contaminants which would have otherwise been considered to be immobile due to their low solubility in water (Thompson et al., 2006) This is due to the release of contaminants into the soil solution as Fe oxides are reductively dissolved, Fe oxides therefore have a huge impac t on water and soil chemistries (Larsen et al., 2006; Luu and Ramsay, 2003)
33 Fe oxides have high pigmentation power and soils derive their colors from the characteristic red, brown and yellow colors of Fe oxides; soil colors are therefore indicative of the presence of Fe oxides and hematite, for example, is responsible for the red color of soils (Roden, 2004; Stucki and Schwertmann, 1988) The ability of Fe(III) oxide minerals to convert to the aqueous Fe(II) species makes Fe(III) an important electron acceptor in redox processes occurring in the vadose zone of soils and aquifers under conditions when oxygen is deficient and Fe(III) has be en reported to be most abundant electron acceptor during organic matter degradation in soils and sediments. Iron Oxides in Soils Background Sources of Iron Oxides primary m inerals including magnetite, titanomagnetite, pyrite, olivine, pyroxene and ilmenite in which Fe occurs in the Fe(II) oxidation state with the exception of magnetite and titanomagnetite (Cornell and Schwertmann, 2003) The Fe concentration in the crust is reported as 5.1 % but a wide variation exists in different rock types; in basaltic and ultramafic igneous rocks the Fe content is about 9.0%, in sedimentary rocks the total Fe content is 3.9% (Stucki and Schwertmann, 1988) and in metamorphic rocks the Fe content is 3.3% (Subcommittee on Iron, 1979) P rimary Fe containing minerals undergo weathering reactions (including oxidation and hydrolysis) to form soils which contain Fe(III) oxi de minerals, an example is shown for the formation of goe thite from olivine in E qua tion 2 1 (Cornell and Schwertmann, 2003) :
34 (2 1) The Fe oxide concentration that is found in a particular soil therefore originates from the concentration of Fe in the parent rock material and this ranges from < 1 to several hundred g kg 1 (Cornell and Schwertmann, 2003) Lindsay (1979) cited that soil Fe concentrations range from 7 to 550 g kg 1 with a reported a verage Fe concentration of 38 g kg 1 Ma (1997) in a study on 40 Florida surface (A horizon) soils reported Fe in the concentration range of 0.084 to 4.5 g kg 1 with an average concentration of 1.2 g kg 1 and in that same paper, the typical Fe concentration in United States soils was reported as 40 g kg 1 The presence of Fe in soils is not regulated as a hazardous compound in Florida for example, the soil clean up target level (SCTL) is only regulated for residential soils at a value of 53 g kg 1 and there are no limits fo r Fe in commercial soils or for soil Fe exceedances for leachability into groundwater (F.A.C., 62 777) In soils, Fe oxides exist as mineral coatings on quartz or clay minerals (Stipp et al., 2002), as ceme nting agents in the cracks and veins of minerals (Kabata Pendias and Pendias, 2000) or as particles such as nodule s and concretions (Gasparatos, 2013). Although f errih ydrite is the initially precipitated Fe (III) oxide, hematitie and goethite are the most common Fe oxide minerals in soils (Benner et al., 2002) Fe oxides may be evenly distributed within a soil horizon or may form concentrations or depletions due t o Fe accumulation or deficiency respectively as a result of fluctuating oxidizing or reducing zones and Fe mobilization/oxidation within the soil horizons (Vaughan et al., 2009)
35 Anthropogenic Sources o f Iron Oxides in Soils Fe oxides are anthropogenically introduc ed into soils through inputs from acid mine drainage sludge, industrial waste products from Fe metal, steel and bauxite manufacturing as well as fertilizer use. In the Fe manufacturing industry metallic Fe is extracted from Fe ores by leaching with acid and the waste product from this process contains Fe oxides (Cornell and Schwertmann, 2003) In the steel manufacturing industry, carbon and metals such as chromium, vanadium and nickel are added to Fe to obtain steel and Fe oxides are produced either as the scale by product during heat t reatment or as rust (Silver, 1993) Removal and discard of these Fe oxides waste products on soils immobilizes the Fe oxides in the soil matrix and weathering of extra Fe in the waste also generate s additional amounts of Fe oxides in the soil (Subcommittee on Iron, 1979) Alumina is extracted from bauxite in the Bayer process through digestion with sodiu m hydroxide and the waste produced called red mud, contains 30 60% of the Fe(III) oxide mineral, hematite (Fe 2 O 3 ) and other minerals like aluminum oxide (Al 2 O 3 ) and silicon dioxide (SiO 2 ) (Brunori et al., 2005; Deba datta and Pramanik, 2013; Summers et al., 1996) Red mud is stored either in soil pits or in lagoons and this has the potential to increase the concentration of Fe oxides in the soil or in groundwater respectively (Feigl et al., 2012) Fe is a micronutrient required by plants for photosynthesis but the low solubility of Fe oxides solubility in soils of about 10 9 M (Luu and Ramsay, 2003) retards the uptake of Fe by plants resulting in Fe deficiency which is manifested by the yellowing and wilting of plant leaves (Subcommittee on Iron, 1979) To increase the bioavailability of soluble Fe to plants, fertilizers containing Fe(II) in the form of ferrous sulfate (FeSO 4 )
36 have been applied to the soil solution but Fe(II) easily oxidizes to for m Fe oxides in the aerobic soil environment and newer methods such as the foliar application of Fe are being investigated and used instead to increase Fe uptake in plants (He et al., 2013; Khoshgoftarmanesh et al., 2010) Acid mine drainage (AMD) contains Fe oxide minerals such as goethite (Ko et al., 2013) and Fe(II) as a result of the oxidation of pyrite containing rocks in subsurface mines (Johnson and Hallberg, 2005) as shown in E quation 2 2 : (2 2) AMD sludge is used as a soil amendment for remediation of arsenic but this increases the content Fe oxides in soils. Mechanisms of Dissolution of Iron Oxides Fe(III) oxide dissolution is important for the supply and transport of Fe in soils, sediments and in natural waters and also aids in the cycling of Fe between these environments (Banwart et al., 1989) Three Fe(III) oxide dissolution mechanisms cited in literature include protonation, complexation and reductive dissolution (Schwertmann and Cornell, 2003) and these reactions involve the adsorption of protons, chelating ligands or reductants onto the surface of Fe(III) oxides respectively (Schwertmann, 1991; Suter et al., 1991) Although similar steps are involved in all three dissolution mechanisms, dissolution by protonation and complexat ion both result in the release of Fe(III) whereas reductive dissolution results in the release of Fe(II) into soil solution. Generally, Fe(III) oxide dissolution occurs when adsorbed dissolution promoting species (protons, ligands or reductants) form compl exes with the surface Fe(III) atom which weakens the Fe O bonds in the Fe(III) oxide mineral so that detachment of the
37 surface Fe(III) into solution occurs (Schwertmann, 1991; Suter et al., 1991) In the dissolutio n process, the adsorption step is fast compared to the slow Fe detachment step (Sulzberger et al., 1989). Dissolution processes are surface controlled reactions since the dissolution rate depends on the sorbent (proton, ligand or reductant) concentration and it is the interaction of the sorbent with the reactive groups [OH (hydroxyl) or O 2 (oxide) ions] on the oxide surface and the reactions that occur that facilitates the detachment of Fe(II) from the oxide surface (Suter et al., 1988; Banwart, Davies e Stumm, 1989). Generally, the dissolution rate is affected by properties of the overall system (including temperature and the presence or absence of UV light), solution phase properties including pH, redox potential and sorbent concentration and properties of the oxide including the specific surface area (SSA), stoichiometry, crystal chemistry and presence (or absence) of defects or guest ions (Cornell and Schwertmann, 2003) Dissolution by Protonation In this dissolution mechanism, adsorbed protons (H + ) are involved in the detachment of Fe(III) from the Fe oxide mineral surface (Sulzberger et al., 1989) The equation for the proton dissolution reaction is given in E quation 2 3 as (Schwertmann and Cornell, 2003) : (2 3) The proton dissolution process involves the adsorption of three (3) protons onto the oxide surface, where each central Fe(III) atom is bonded to neutral OH/OH 2 pairs to 2 Adsorption of a proton (H + ) results in a positive surface charge on the 2 ) 2 + group and a positive charge of +1 on the
38 oxide surface; adsorption of the two (2) additional protons onto neighboring O/OH groups adjacent to the protonate 2 ) 2 + increases the oxide surface charge to +3 (Cornell and Schwertmann, 2003) Proton adsorption decreases the attractive force between the surface Fe(III) and the O 2 atom bonded to it and weakens the covalent Fe O bond which promotes the detachment of Fe(III) 3+ into solution (Sidhu et al., 1981) Restoration of the Fe(III) oxide surface occurs as further protons are adsorbed of protons restores before additional proton dissolution reactions occur. In acidic media, anions such as chloride and sul fate (from HCl and H 2 SO 4 respectively) associated with the adsorbed proton(s), also aid in the dissolution of the Fe(III) oxides by forming additional surface complexes by replacement of surface hydroxyl (OH ) groups with chloride and sulfate anions which either weakens the oxide Fe O bonds or lowers the positive charge on the oxide surface (for further proton adsorption) for additional Fe(III) detachment (Cornell and Schwertmann, 2003) Dissolution by Complexation In this mechanism, a complex is formed between the oxide surface Fe(III) atom and an adsorbed ligand (anion or a weak acid) results in a n increase in the solubility of Fe(III) (Sulzberger et al., 1989). The equation for the dissolution of Fe(III) oxides by complex ation is given by the E quation 2 4 below where L is the ligand denotation (Schwertmann, 1991) (2 4) The complexation dissolution mechanism occurs in three (3) steps : (i) ligand adsorption, (ii) metal detachment and (iii) surface restoration by proton adsorption (Schwertmann, 1991) Initially, Fe(III) complexes are formed by replacement of the
39 surface hydroxyl (OH ) groups with adsorbed ligands, such as oxalic acid or citric acid (Cornell and Schwertmann, 2003) The formed Fe (IIII) complex (Fe ligand) weakens the Fe O bonds in the oxide mineral which results in detachment of the Fe(III) complex into solution and as in the ca se of the proton dissolution mechanism, restoration of the Fe oxide surface by adsorption of protons in solution occurs before further complexation dissolution reactions proceed (Sulzberger et al., 1989) Reductive Dissolution of Iron Oxides Reductive di ssolution of Fe oxide minerals occurs under reducing conditions (when oxic conditions changes to anoxic conditions) and in the presence of a reductant, an electron donor which may be an organic ligand or a metal ligand complex (Suter et al., 1991) According to Davranche et al (2000), Fe reductive dissolution is a three step surface, (ii) reduction of surface Fe(III) and adsorbed contaminants and (iii) diffusion of Fe(II) Fe reductive dissolution specifically involves the transfer of electrons from the reductant, which is the electron donor, to the surface Fe(III ) the terminal electron acceptor (TEA) which then becomes reduced to Fe(II) as shown in the simplified reduction reaction in Equation 2 5 which involves electro n transfer An example is the reduction of Fe(III) with ascorbate as the electron donor (or org anic reduc tant) as shown in Equation 2 6 Changes in the oxide structure from the loss of charge as Fe(III) converts to Fe(II) and the space occupied by the larger size of the newly produced Fe(II) results in weakened Fe(II) O bonds in the oxide mineral an d the eventual detachment of Fe(II) into the soil solution (Suter et al., 1991)
40 (2 5) (2 6) According to Stucki (1988), a small fraction of the released Fe(II) remains in the soil solution as the free dissolved species but the rest either precipitates out with phosphate, sulfide, etc as a solid mineral or becomes exchangeable Fe(II) which displaces Ca 2+ and Mg 2+ in the structure of soil minerals (Stucki and Schwertmann, 1988) Reductive dissolution is also affected by pH since adsorption of a charged reductant either becomes increased or lessened with pH (Schwertmann and Cornell, 2003) Under the same pH conditions, studies have shown that reductive dissolution is a faster mechanism than protonation because of the faster detachment of Fe(II) compared to that of Fe(III) due to the fact that Fe(II) O bonds are much weaker than Fe(III) O bonds (Cornell and Schwertmann, 2003; Larsen et al., 2006) A second mechanism of reductive dissolution is reductively catalyzed dissolution which involves the combined use of a reductant and a complexing agent, an example is the Fe(I I) oxalate complex, where the Fe(II) serves as the reductant and oxalate, the organic ligand, serves as the complexing agent (Suter et al., 1991) The first step involves adsorption of the Fe complex to the oxide surface through attachment of the ligand which serves as a bridge for electron transfer to the oxide surface (Sulzberger et al., 1989; Suter et al., 1991) In the case of Fe(II) oxala te, electrons are transferred from the Fe 2+ oxalate complex to the oxide surface, reduction of Fe(III) occurs followed by detachment of the produced Fe(II) and restoration of the oxid e surface as shown in E quation 2 7 (Cornell and Schwertmann, 2003) Metal ligand complexes significantly
41 increase the rate of Fe(III) oxide dissolution, for example in the study on the reduction of goethite by oxalate, a substantial amount of Fe(III) dissolved with 10 M Fe 2+ and 20 M Fe 2+ compared to the amount of Fe(III) dissolved in the absence of Fe(II) (Suter et al., 1991) The enhanced dissolution with a combined reductant ligand complexing agent has been used utilized in the extraction of Fe in soils such as in the citrate dithionite bicarbonate (CDB) extraction of total Fe where dithionite serves as the reductant and citrate serves as the complexing agent. (2 7) Biotic and Abiotic Reductive Dissolution of Iron Oxides The factors that regulate the rate and extent of Fe reductive dissolution have been extensively studied and chemical and biological sources have both been suggested. Iron reducing bacteria responsible for the biotic reduction of Fe(III) oxides have been identified. Chemical Fe reduction has be en studied with reductants including ascorbate (Banwart et al., 1989; Larsen et al., 2006; Suter et al., 1991) and hydroxylamine (Davranche and Bollinger, 2000) .The rate of the chemical reduction of Fe has been shown to be faster than the rate of biological reduction but contrasting results have been obtained for the percentage of Fe(III) reduced through biotic processes compared to that obtained with chemical reductants. Hycainthe et al. (2006) showed that microbes reduced only 65% of the total amount of Fe(III) that was reduced by buffered ascor bate but Roden (2004) demonstrated that a higher percentage reduction of Fe(III) was obtained with the abiotic reductant, ascorbate (90 100%) of Fe(III) compared to the 13 39% reduction obtained through biotic (enzymatic) reduction. A similar observation w as shown by Lovely et al. (1991) who examined the enzymatic
42 versus non enzymatic reduction of some selected short chain fatty acids and aromatic organic compounds including acetate and benzene respectively and showed that the concentration of Fe(II) releas ed as a result of enzymatic Fe(III) was an order of magnitude higher than the Fe(II) released as a result of on enzymatic mechanisms. Biotic Reductive Dissolution of Iron Oxides In non sulfogenic environments, biological oxidation of organic matter coupled with the reduction of Fe is the dominant mechanism (Minyard and Burgos, 2007) and the factors that control the rate of bacterial reductive dissolution include the quantity and quality of the carbon source, re adsorption of the released Fe(II) (Roden et al., 2000) adsorpti on of inorganic anions such as phosphate (Davranche et al., 2013) Additionally, the surface area of Fe(I II) oxides has been shown to have a strong correlation on bacterial Fe reduction (Roden, 2004; Roden and Zachara, 1996) and slower reduction rates have been observed with increasing oxide crystallinity (Dollhopf et al., 2000) Bacterial driven reductive dissolution occurs when Fe oxides act as terminal electron acceptor during microbial respiration processes in soils and sediments. (Roden, 2004) During this pro cess, organic matter (abbreviated as OM and CH 2 O in the chemical reactions below) in soils/sediments is generally degraded through microbial catalyzed decomposition processes, where OM is oxidized with oxygen acting as the terminal electron acceptor as sho w n in E quations 2 8 and 2 9 respectively. (2 8) (2 9) In a simplified manner, the overall reaction is shown in E quation 2 10 and is therefore given as:
43 (2 10) As organic matter decomposes, if the oxygen s upply is limited in soils and sediments, then the depletion of oxygen results in development of anoxic conditions in these environments and the use of other terminal electron acceptors in the process according to the following redox sequence: NO 3 Mn (IV) oxides, Fe(III) oxides, SO 4 2 and CO 2 (Cozzarelli et al., 2000) and the equations for the reduction half reactions are given below in E quations 2 11 to 2 17 (Heron et al., 1994; Luu and Ramsay, 2003) : (2 11) (2 12) (2 13) (2 14) (2 15) (2 16) (2 17) In soils and sediments, Fe reduction occurs after nitrate reduction and before sulfate and CO 2 (methangenosis) re duction since Fe reduction is thermodynamically favorable compared to sulfate reduction and methangenosis but it yields less energy than O 2 and nitrate reduction (Luu and Ramsay, 2003) In these environments, Fe(III) oxides serve as important electron acceptors since the Fe(III) oxide concentration exceeds that of nitrate and sulfate and there are distinct zones where the oxidation of OM is coupled to the re duction of Fe(III) to Fe(II) (Larsen and Postma, 2001; Lovley, 1991) In soils and sediments, microorganisms known dissimilatory Fe reducing bacteria (FeRB) have been identified which gain energy for growth by cata lyzing the oxidation of
44 organic compounds with the use Fe(III) as electron acceptors (Lovley and Phillips, 1988) Studies have been performed using differen t FeRB bacteria include Geobacter metallireducens, which was first discovered in 1993 by Lovley (Lovley et al., 1993; Luu and Ramsay, 2003) Geothrix fermentans (Nevin and Lovley, 2002) Shewanella putrefaciens, Shewanella alga and Ferrimonas balearica Abiotic Reductive Dissolution Fe(III) reduction due to abiotic or non e nzymatic reactions have also been demonstrated with organic compounds such as ascorbate, fructose, benzaldehye (Lovley et al., 1991; Poulton et al., 2002; Roden, 2004; Sorensen, 1982) and reduced sulfur compounds s uch as hydrogen sulfide which act as electron donors during the reductive dissolution of Fe(III) oxides (Poulton et al., 2004; Slowey and Brown, 2007; Yao and Millero, 1996) The use of humic substances as electro n shuttler for Fe(III) reduction has also been reported by Lovley (1997) I n this mechanism, humic su b stances serve as electron acceptors for the oxidation of organic matter and become reduced, then the reduced humic substances abiotically transfer electro ns to Fe(III) oxides to initiate Fe reduction (Lovley, 1997) Pot ential for Soil Fe ( III ) Oxides to Act as an Fe Source of Groundwater Contamination Since the Fe reducing bacteria, Geobacter metallireducens, was first discovered by Lovley (1993) the interaction of microbes with Fe(III) oxides minerals and the mechanisms of Fe reductive dissolution have drawn much research interest in various fields including microbiology, geochemistry and soil scien ce and separate studies in have been conducted in each of these fields. Much of the research on Fe reductive dissolution h as been conducted with synthetic Fe(III) oxides and various organic carbon
45 sources in culture media. The use of iron oxides (or iron s alts) to remove H 2 S gas from sewers and aquaculture systems has also been investigated. Fe(III) oxides are common m inerals in soils and come into contact various sources of organic matter either from that which is native to the soil and/or landfill gases. These are important factors which also have the potential to induce Fe reduction in the soils but the existing research on Fe reductive dissolution does not take into account the effect on soils. T here is a gap in the knowledge of how interactions from the soil matrix or how the soil environment affects the extent of Fe reduction in soils The aim of this research is to investigate Fe (III) reductive dissolution in soils and to identify the parameters responsible for this phenomena. Finally, incorporation o f the recognized data generated from this research can be used by researchers in these fields to foresee the potential for F reductive dissolutio n in various soil environments.
46 CHAPTER 3 IRON REDUCTIVE DISSO LUTION IN SOILS AS A FUNCTION OF ORGANIC CARBON TYPES AND DEGREE OF CRYSTALLIZATION OF IRON OXIDE MINERALS : IMPLICATIONS FOR G ROUNDWATER POLLUTION BY LANDFILLS Introduction Landfills are the most widely used method for waste disposal (Cozzarelli et al., 2000) and the occurrence of groundwater contamination in landfi ll impacted sites has been in some cases problematic as well as the focus of several studies (Baun and Christensen, 2004; Christensen et al., 2001; Lyngkilde and Christensen, 1992b) Landfill leachate which origina tes from the decomposition of solid wastes and then drain s downward by infiltrating rainwater (Bolton and Evans 1991), is made of a wide variety of dissolved chemical components including iron (Christensen et al., 2 001; Wizniowski et al., 2006) Landfill leachates are also very rich in organic matter (OM) and depending on the age of the landfill, the proportions of low (e.g. organic acids such as acetate and sugars) and high (fulvic and humic acids) molecular weigh t compounds vary over time (Lyngkilde and Christensen 1992), with an overall temporal decrease in the concentration of low molecular weight organic compounds; as they are preferentially used as source of energy for bacteria. The total organic carbon (TOC) content of landfill leachates has been reported to range from 0.03 to 29g C/L (Kjeldsen et al., 2002) It is therefore likely that contact between landfill leachate on one hand and soil on the other could result in increased soil microbial respiration, stimulated by the high concentration of dissolved organi c carbon (DOC) in the leachates and the presence of electron acceptors in soils. In fact, m icrobial respiration of organic matter in natural systems is a redox process in which bacteria harvest the energy stored in chemical bonds by oxidizing organic matte r; a reaction that is coupled with the reduction of terminal
47 electron acceptors (TEAs) specific to the group of microorganisms involved. With regard to this study, iron oxide minerals present in soils can act as TEAs during anaerobic respiration o f organic matter by iron reducing bacteria (FeRB), a process which would result in the reduction of insoluble Fe(III) compounds to produce water soluble Fe(II) species. Accordingly, landfills could impact the concentration of Fe in groundwater through one or a comb ination of the above described processes. Indeed, Fe can occur at high concentrations in landfill leachates (Kjeldsen et al., 2002) and it has been observed in a few cases linked to incidences of Fe groundwater contamination at various landfill im pacted locations in the world (Albaiges et al., 1986; Christensen et al., 2001; Lyngkilde and Christensen, 1992b; Rugge et al., 1995) In recent years, however, excessively high concentrations of dissolved iron hav e been measured in groundwater samples collected from sites impacted by either lined or unlined landfills in Florida (Geosyntec, 2005) And in these specific cases, spatio temporal data obtained from monitoring wells have shown little to no link between different geochemical tracers of landfill leachate pollution and levels of Fe measured in impacted groundwater (Geosynt ec, 2005) ; implying that sources other than Fe initially in landfill leachate could be the cause of the observed contamination. In fact, Fe reductive dissolution has been documented as one of the most dominant redox processes in anaerobic soil environment s impacted by organic matter including rice paddy and wetland soils. It has been demonstrated that in rice paddy fields and in wetlands, the flooding of these soils results in microbial processes which utilize TEAs including soil Fe(III) oxide minerals fol lowing oxygen depletion in soils (Ponnamperuma, 1972; Yao et al., 1999) Overall,
48 the above observations raise questions on the significance of the role of soil Fe oxide minerals as a source of the Fe measured in la ndfill impacted groundwater. Iron oxide minerals are commonly found in soils (Fredrickson et al., 1999) where they can have a profound impact on soil solution chemistries (Larsen et al., 2006; Luu and Ramsay, 2003) and give soils their characteristic yellow, brown, or red colors (Roden, 2004; Stipp et al., 2002) adopted in this paper includes oxides, hydroxides and oxyhydroxides as reported previously by other researchers (Alloway 1995) Their concentrations can range from 1 to hundreds of g/kg of soil (Cornell and Schwertmann 2003). They can be present in soils as coatings on sand and clay particles (Stipp, Hansen et al. 2002), as nodules or as fillings in the cracks and veins of minerals (Kabata Pendias and Pendias 2000). Variations exist in the (i) degree of crystallinity, (ii) stability, (iii) solubility, (iv) particle size, and (v) specific surface area (SSA) of these minerals. With regard to crystallinity, iron oxide minera ls are found in non crystalline/amorphous forms such as in ferrihydrite Fe(OH) 3 and in highly crystal FeOOH) and hematite Fe 2 O 3 ) (Roden, 2004) It is also known that the solubility of iron oxides decreases as the stability of the mineral increases (Benner et al., 2002) Iron oxide particles are present in soils in nano size, ranging from 5 to 150 nm, resulting in high SSA (Schwertmann 1991). Iron oxide surfaces can sorb metal cations such as Ca 2+ (Pedersen, Postma et al. 2005) as well as anionic species such as AsO 4 3 and PO 4 3 (Stipp, Hansen et al. 2002 ) through a combination of pH dependent electrostatic forces and chemical bonding. But when iron oxide minerals undergo reductive dissolution, previously sorbed compounds get released into solution, increasing their mobility,
49 bioavailability, and biologica l impacts (Makris, Harris et al. 2005, Thompson, Chadick et al. 2006). In soils and aquatic systems, Fe occurs predominantly in two thermodynamically stable oxidation states as Fe(III) and Fe(II) under oxic and anoxic conditions respectively (Viollier et al., 2000) During Fe reductive dissolution, Fe(III) contained in insoluble iron oxide minerals is reduce d to the Fe(II) species, in reactions that can involve organic matter when catalyzed by bacteria (Lovley et al., 1991) or other reducing agents such as sulfide compounds in abiotic reduction reactions (Dos Santos and Stumm, 1992) In this study we investigated the combined effects of organic mat ter the study is to demonstrate that soil iron oxides can be significant sources of Fe that could lead in pollution of groundwater as observed in some landfill impacte d watersheds. The underlying assumption is that contact exists between the vadose zone soil underneath the landfill unit and leaking organic rich landfill leachate. It has been hypothesized that the biological reductive dissolution of soil Fe(III) oxide mi nerals would depend on both the quantity and type of organic matter that acts as electron donor during the anaerobic respiration by FeRB; and the degree of crystallization of minerals used as TEA s Materials and Methods Vadoze Soil Sample Collection and Ha ndling Vadose zone soils used in this study were collected in North Florida from different locations including (i) sites of current landfill Fe impacted groundwater, and (ii) sites with no landfill activities. Figure 3 1 illustrates the strategy used for collection of the different vadose zone soil samples. Briefly, soil samples were obtained from selected
50 locations in North Florida, and along a hypothetical transect defined to provide differences in the degree of crystallization of the Fe oxide minerals. The rationale for site selection was based on differences in the depth of the water table and the impacts of its fluctuations on the mineralogy of iron found in these soils. In this approach, samples obtained from zones 1 and 3 (Figure 3 1) were anticipate d to represent the crysta lline and amorphous end members respectively; and bracketing samples from zone 2 characterized by intermediate degrees of Fe oxide mineral crystallization. The actual sampling locations of the samples are as fo llows: s amples 1 and 2 were collected from two different locations within the site of the Klondike Landfill in Escambia County, Florida. This is the site of a n un lined landfill which closed in 1982 after six years of operation. Sample 3 was collected from the site of the New R iver Regional Landfill (NRRL), a lined landfill in Union County, which started operation in 1992 (Jain et al., 2006) and is still active. Sample 4 was collected from the Alachua County Southwest Landfill, a site of a lined landfill in Alachua County, which closed in 1998 after a 10 year period of operation (Comstock et al., 2010) Therefore, zone 1 samples (Figure 3 1) came from sites with past or ongoing l andfill activities. In contrast, samples from zone 2 and zone 3 came from sites with no existing landfill activities. These sites are loc ated in UF/IFAS Plant Science Research and Education Unit in Marion County, the Ordway Swisher Biological Station in Pu tnam County and the Austin Carey Memorial Forest in Alachua County. At each sampling site, samples were collected from deeper soil horizons (~2 meters below the surface) using an auger or using a shovel for samples obtained in locations with pre existing t renches. Following the collection process, soil samples were
51 placed in covered high density polyethylene contain ers and transported to our laboratory at the University of Florida where these soil samples were air dried, homogenized using a pestle and morta r, passed through a 2 mm sieve to remove large particles and gross organic debris (Al Abed et al., 2006) ; and then stored in sealed HDPE containers at room temperature (Kashem and Singh, 2001) until use for determination of physicochemical char acteristics and then in biotic and abiotic Fe reductive dissolution experiments. Characterization of Collected Vadose Zone Soil Samples Soil particle size fractions, defined as sand (0.05 to 2 mm), silt (0.002 to 0.05 mm) and clay (< 2 mm) (USDA, 1992) were determined using a combination of the pipet method and gravimetric measurement of each size fraction after soil pretreatment with hydrogen peroxide (H 2 O 2 ) to remove organic matter (USDA, 1992, 2004) For each of the collected soil samples, pH was measured using an Accument Basic AB15 pH meter on a 1:1 (m/v) soil water slurr y which had been left to equilibrate 846 Method for Soil and Waste pH (USEPA, 2004) The organic matter (OM) conte nt of each soil sample was approximated as loss on ignition (LOI) following a 4 hour combustion at 550 o C of previously oven dried soil (at 105 o C for 16 h ou rs), the calculations are shown in (Eq uation 3 1), where DW stands for dry weight of the sample measu red at the temperature indicated by the subscript number. The LOI at 550 o C is assumed to be eq u ivalent to the % OM content in E quation. 3 2 and the % of organic carbon (OC) is evaluated as half of the % LOI as shown in E quation 3 3 (Santisteban et al., 2004; Veres, 2002) (3 1) (3 2)
52 (3 3) The amorphous Fe conce ntration of the soils was determined by Inductively Coupled Plasma Atomic Emission Spectroscopy (ICP AES) after extraction and centrifugation of 0.5g of each soil sample with an acidified ammonium oxalate (AAO) solution composed of a mixture of 0.20 M ammo nium oxalate ((NH 4 ) 2 C 2 O 4 )and 0.20 M oxalic acid (H 2 C 2 O 4 ) (Begin and Fortin, 2003; McKeague and Day, 1966) The total Fe oxide concentration (i.e. the combined crystalline and amorphous Fe content) of the soils was determined on citrate dithionite bicarbonate (CDB) extract s of the soil samples conducted in a water bath at 80 0 C, followed by centrifugation and analysis by ICP AES. In the CDB procedure, the sodium dithionite (Na 2 S 2 O 4 ) solution acts to reduce Fe(III) in the soil samples to Fe(II) the sodium bicarbonate (NaHCO 3 ) solution buffers the solution (pH 7 9) and the sodium citrate (Na 3 C 6 H 5 O 7 .2H 2 O) solution chelates the released Fe(II) to prevent sorption on solid phases ( Bera et al., 2005; Golden et al., 1994; Mehra and Jackson, 1960) Finally, the difference between CDB Fe and AAO Fe concentrations allowed for the determination of Fe present in the sample in crystalline forms (Bera et al., 2005) S oil mineralogy was determined on clay size fractions mounted on tiles or slides which underwent x ray diffraction (XRD) using Cu K radiation ( 1.53 ) as generated by a computer controlled X ray diffractometer equipped with a stepping motor, a graphite monochromator and a scintillation d etector (Harris and White, 2008; Soukup et al., 2008) S amples were scanned from 2 60 0
53 Organic Compounds Used as Energy Source and Microbial Reductive Dissolution of Soil Fe oxide Minerals Since all of the collected soil samples had rather low OM content (Table 3 1) and preliminary Fe reductive dissoluti on experiments showed that native soil OM and microflora could not support detectable Fe reductive dissolution processes (data not shown) ; soil slurries were spiked with either low molecular weight organic matter (LMWOM) glucose in this case to account fo r the impact of LMWOM abundantly present in the leachates from young landfills; or with aliquots of an old landfill leachate, characterized by an organic matter content dominated by high molecular weight organic matter (HMWOM) such as fulvic and humic acid s In fact, OM in landfill leachate is made of complex polymers including proteins, lipids and polysaccharides which undergo hydrolysis to form amino acids, fatty acids and monosaccharides (Vymazal and Krpfelov, 2009) Therefore, besides the commonly cited acetate, landfill leachate contains other organic compounds of variable and unknown proportions and could be composed of different carboxylic acids of longer carbon chains (such as lactate) and sugars (such as glucose, C 6 H 12 O 6 ). Municipal solid waste (MSW) contains cellulose and hemicellulose (polysaccharides) which constitute the larges t portion (40 to 60 %) of the waste on a dry weight basis and these compounds primarily undergo hydrolysis by microbes to produce monosaccharides (Kjeldsen et al., 2002) Glucose was therefore selected as synthetic model LMW OM (Lovley et al., 1991; McLean et al., 2006 ) for the Fe reduction dissolution conducted in this study. Using glucose as a LMWOM, Bennett and Dudas (2003) evaluate d the reduction of Fe oxide minerals in soils by spi king soil slurries with 1 g of C/L Additionally glucose concentrations ranging from 0.2 to 20 g C/L have been used to study the reduction of
54 Fe oxide minerals in aquatic sediments (McLean et al., 2006) The approach used in this study was therefore adapted from the above references and glucose concentrations expressed as organic carbon concentrations of 9, 18 and 36 g C /L (equivalent to 0.125M, 0.25M, and 0. 5M glucose ) were used. Finally concentrations of organic carbon obtained after addition of glucose to the soil slurries in this study are comparable to TOC level characteristics of landfill leachate (Kjeldsen et al., 2002) A simplified redox reaction involv ing glucose as source of carbon can be summarized as illustrated in E quation 3 4; where the oxidation of glucose releases electrons that reduce Fe(III) in ( solid) to Fe(II) aq (3 4) To investigate the effect of HMWOM, leachate collected from an aged landfill ( NRRL ) and dominated by fulvic and humic acids was used as the source of organic carbon Before use, the leachate was centrifuged, vacuum filtered (0.45m) and characteriz ed for relevant physicochemical parameters. The pH was determined with an Accument Basic AB15 pH meter, the dissolved organic carbon (DOC) content was determined using a total organic carbon (TOC) analyzer (Shimadzu TOC VCPH, Japan). The chemical oxygen de mand (COD) was obtained through measurement of the absorbance ( = 620nm) after sample digestion using the dichromate method (HACH Method 8000). 254 ) was measured on a Hitachi U 2900 spectrophotometer, and the fluorescence excitation emission matrix (EEM) spectrum for the qualitative det ermination of the different groups of organic compounds was measured with a Hitachi F 2500 fluorescence spectrophotometer. Detailed descriptions of th e characterization procedure s can be found in Comstock et al. (2010)
55 Using CH 2 O as a general and simplified chemical formula for OM, and assuming complete mine ralization, the redox equation involved can be summarized as follows in E quation 3 5 (3 5) Iron Reductive D issolution E xperiments Prelim inary laboratory tests were conducted first to determine the minimum incubation time necessary for the establishment of strict anoxic conditions in the soil slurries using resazurin dye. The latter acts as an indicator of anoxic conditions and changes colo r from blue to pink under reducing conditions as it becomes reduced first to resofurin and then colorless as it is reduced to dihyhdroresofurin (Miller and Wolin, 1974; O'Brien et al., 2000) Resazurin dye also aid s in the detection of oxygen contamination of the slurries if the non colored aqueous solution turns back to pink upon re oxidation of the dihydroresofurin compound (Tratnyek et al., 2001) Based on prelim inary experiments, a minimum of a two week incubation time was adopted and samples were incubated for even longer times (up to a month) Fe reductive dissolution experiments were conducted by preparing soil slurries in which soil and the tested organic com pound solutions were mixed in a 1: 10 ratio (mass/volume). Serum bottles were used as test microcosms in which slurries were prepared under zero headspace and incubated in the dark at room temperature. The mixture was de oxygenated by bubbling with N 2 and then spiked with an aliquot from a consortium of anaerobic digester bacteria ( including FeRB ) The latter were used to microorganisms induced no Fe reductive dissolution afte r several weeks of incubation
56 under anaerobic conditions It is worth to emphasize that Fe released in these experiments are considered to be from the biological reductive dissolution of Fe oxide minerals due to the following reasons. First, soil slurries plus bacteria from the anaerobic digester, but with no organic matter addition (i.e. native OM content) gave no detectable dissolved Fe(II) after weeks of incubation under anoxic conditions. Second, soil slurries with OM added but no bacteria showed no Fe reductive dissolution. Third, soil slurries with no OM and no bacteria addition resulted in no Fe reductive dissolution. Accordingly, the above three treatments were treated as control incubations emphasizing the importance of the simultaneous additions of OM and bacteria from the anaerobic digester. All treatments were run in triplicates and in a few cases repeated to deal with the large variability in the obtained data A the end of the incubation period and within an anaerobic chamber, aliquots of the aq ueou s phase were withdrawn using sy ringes and filtered (0.45mm) using syringe filters. The obtained filtrate was then added to the Ferrozine HEPES buffer solution for determination of the Fe(II) released as described below. Iron Analysis Fe(II) released in to solution from the Fe reductive dissolution experiments was analyzed using ferrozine, a disodium salt of 3 (2 pryidyl) 5,6 bis(4 phenylsulfonic acid) 1,2,4 triazine (C 20 H 13 N 4 NaO 6 S 2 Acros Organics), which reacts with Fe(II) to form a soluble magenta com plex with a maximum absorbance at =562nm (Stookey, 1970) .The ferrozine reagent was prepared by dissolving 0.1% (w/v) of ferrozine in 0.05M HEPES buffer solution ( 4 (2 hydroxyethyl) 1 piperazineethanesulfonic acid C 8 H 18 N 2 O 4 S pH 7 Fisher Scientific Chemicals). Fe(II) standards were prepared from a ferrous ammonium sulfate stock solution (FeSO 4 .7H 2 O Fisher Scientific Chemicals ).
57 For analysis, filtered sample aliquots were added to 5 mL of the ferrozine HEPES buffer solution and analyzed for Fe(II) or acid extractable Fe(II) (0.5N HCl) at 562 nm by colorimetry (Roden, 2004; Roden and Zachara, 1996; Royer et al., 2002a) All Fe extractions and sample preparation for analyses using the ferrozine method were conducted inside an anaerobic chamber which had been previously vacuumed to remove oxygen and then filled with nitrogen, an inert gas, to ensure operation under an anoxic atmosphere to prevent re oxidation of the extracted Fe(II) Results and Discussion Physico chemical Characteristics of Collected Soil s Data obtained from the different physicochemical analyses of the soil samples are shown in Table 3 1. The particle size distribution (PSD) data showed sand as the dominant fraction in all soil samples, results which are consistent with the prevalence of sandy soils in Florida The majority of the soils could be classified as acidic soils and sample 1 with a pH of 4.16 was the most acidic. The one exception was sample 4, which had a measured pH value of 7.1. The % OM content of the collected soils ranged fr om 0.08% to 2.51 %; equivalent to an OC content of 0.04% and 1.26%. The low %OC content in these soil samples make them meet the classification of mineral soils containing < 12 % OC (Essington, 2003) The AAO and CDB extraction results confirmed to some extent the gradient in the degree of Fe crystallization in the collected soil samples. However, sample 4 classified in zone 1 (Figure 3 1) exhibited characteristics of soils classified in zone 3. However, the XRD spectra of these soil samples revealed the presence of crystalline Fe mine rals in only samples1, 2 and 4; with hematite (He) detected in sample 1, goethite (Go) in samp le 2, and possibly magnetite (Ma) in sample 4 ( Figures 3 2 and 3 3 ). No other crystalline Fe minerals were
58 detected in the rest of the soil samples using XRD analysis The lack of total overlap between the results from the chemical extraction and XRD analy sis can be explained by differences in detection limits as well as the accuracy (i.e. the capacity to truly discriminate between amorphous and crystalline Fe fractions) of these methods. Overall, zone 1 soils had high total Fe but low amorphous Fe content, zone 2 soils had both high total Fe and high amorphous Fe content, and zone 3 soils had both low total Fe and low amorphous Fe content. Fe reductive Dissolution in Soils Treated with Glucose as Model Organic Compound The results obtained from the incubati ons with all of the three tested glucose concentrations (9, 18 and 36g C/L) indicated that the chemistry of iron in the soil could affect the dissolution of Fe(III) oxide minerals, regardless of the tested glucose concentration. For each soil, comparative statistical analysis of the determined rates of Fe(II) release at each of the tested glucose concentrations was conducted using the SigmaPlot statistical software package. For each soil, significant differences between the means of the determined rates of Fe(II) release obtained at each of the 3 glucose con centrations were determined by analysis of variance (ANOVA ); and if differences existed, multiple comparisons of the rates of Fe(II) release were performed (Beesley et al., 2010) Based on the statistical analyses, 3 different trends were identified ( Figure 3 4 ): (i) a positive trend in which rates of Fe reductive dissolution increased with increasing glucose concentration ( Figure 3 4A ), (ii) a negative trend in which rates of Fe reductive dissolution decreased while glucose concentrations increased (Figure 3 4B); and (iii) finally a relatively flat trend ( Figure 3 4C ), in which rates of Fe reductive dissolution remained statistically not different
59 regardless of the in c rease in glucose concentration These trends can be attributed to several factors and the most relevant are discussed below. In the soil slurries, while the release of Fe (II) from solid to aqueous phase is driven b y the redox reactions shown in E quation 3 4; the net concentration of Fe(II) in the aqueous phase is actually the product of two concurrent reactions, Fe reductive dissolution and Fe(II) removal from solution via both sorption and precipitation. The ph E h Theory on the C onditions of F ormation of F eS / FeS 2 and FeCO 3 In anaerobic soil slurries or landfill leachates contai ning organic matter, carbonate and sulfide, the reductive dissolution of Fe(III) oxide minerals occurs in parallel with competitive reactions that affect the concentrations of produce d Fe(II). Therefore, measured concentrations of dissolved Fe(II) in such cases represent the balance of the co occurring reactions including precipitation reactions and sorption onto solid surfaces. The literature is quite abundant with regard to the geoc hemistry of Fe and Fe species distribution in aqueous systems when assuming thermodynamic equilibrium conditions. For instance, for a system at 25 o C under a pressure of 1 bar, and containing 1M of total dissolved carbonate, 10 6 M of total dissolved iron a nd 10 6 M of total dissolved sulfur; siderite (FeCO 3 ) would prevail as the stable species for solution pH values ranging from 6 to 11 and E h values ranging from 0 to 0.6 V (Garrels and Christ, 1965; Whiteley and Pearce, 2003) Siderite occurs therefore in the presence of high concentrations of dissolved carbonate (Muehe et al., 2013) and at low E h and near neut ral pH values. In fact, when organic matter is added to a biologically active system simi lar to the one described above in the pre sence of added dissolved carbonate, organic compounds can drive the formation of siderite through microbial catalyzed
60 oxidatio n of organic matter (CH 2 O) which is coupled to the reduction of Fe(III) oxide minerals and is summarized in E quation 3 6 (3 6 ) With regard to sulfur, the presence of sulfide in solution favors the formation of iron sulfide minerals (Bell et al. 1987) such as the transient species iron monosulfide ( FeS ) and the more stable pyrite (FeS 2 ). The latte r would prevail under solution pH values that range from 3 to 9 and E h values ranging from 0.1 to 0.4 V (Garrels and Christ, 1965) Accordingly, one can speculate on the basis of the rationale presented above on the potential fate of Fe(II) released into the soil solution as a result of the reduction of solid Fe(III) oxide minerals. Obviously, conditions that favor the formation of siderite and Fe sulfide minerals as well Fe(II) sorption on solid surfaces would re sult in decreased concentrations of dissolved Fe(II) following the reductive dissolution reaction. It is worth to note that under short term laboratory experiments such as those conducted in this study, formed siderite and Fe sulfide minerals are in amorph ous forms and are not detectable by XRD analysis. Additionally, the concentrations of formed crystalline minerals, if any, would likely be below the detection limits of the available analytical instruments (e.g. XRD). Assuming that the above conditions of low Eh and high dissolved carbonate concentrations were satisfied and d epending on the pH value reached (e.g. if ranging from 7 to 10), ideal thermodynamic conditions for the formation of siderite (FeCO 3 ) could have be en met for dissolved Fe(II) removal fr om a queous phase via precipitation as siderite as shown in E quation 3 7 In fact, in these hermetically sealed anaerobic
61 batch reactors, the oxidation of glucose produces carbon dioxide (CO 2 ), and at the same time Fe reductive dissolution in the presence o f glucose consumes protons (E quation 3 4) which raises the pH of the slurries. These findings therefore suggest the possibility of siderite formation in the soils in which the negative trend in Fe(II) release were observed In fact, previous studies have r eported on the formation of siderite in anoxic sediments with prevalent bacterial iron reduction activities (Bell et al., 1987; Ellwood et al., 1988; Haese et al., 1997) (3 7 ) Accordingly, in the soils exhibiting the positive trend (samples 3, 7, 9 and 10 ) it is likely that the rates of Fe(III) reduction were higher tha n the rates of Fe(II) sorption and Fe(II) precipitation as FeCO 3 ; which lead to an accumulation of Fe(II) in these reactors Consequently, the observed negative trends (samples 4, 11 and 12) are suggestive of Fe(II) sorption and Fe precipitation rates (as F eCO 3 ) being gr eater than rates of Fe reductive dissolution. Additionally, the low soil total Fe concentration s of these samples could result in the exhaustion of Fe(III) from these samples and as glucose concentrations increased, Fe(II) removal from soluti on through sorption and siderite formation continued Finally, soil samples exhibiting a flat trend (i.e. no discernable trend in slurries made with soil samples number 1, 2, 5, 6 and 8) point to the potential of Fe in soils to control the reductive dissol ution due to either quantity or speciation aspects rate of Fe (II) release, correlation analyses between the rates of Fe(II) release at each of the tested glucose concentratio ns on one hand and either the total Fe, crystalline Fe or
62 amorphous Fe concentrations on the other were conducted using data from all 12 soils. At each glucose concentration, a positive linear correlation was observed between the rates of Fe(II) release an Figures 3 5, 3 6, and 3 7 ) and the results indicated that the rate of Fe(II) release was proportional to the initial amorphous, crystalline an d t otal concentrations as was reported earlier by Roden (1996) Each of the individual Fe portions therefore had an effect on the rate of Fe(III) release. For instance with 9 g C/L glucose, the rate of Fe(II) release were positively correlated with the total Fe content (r 2 =0.744) (Figure 3 7A), crystalline Fe content (r 2 =0.471) (Figure 3 7B) and amorphous Fe content (r 2 =0.960) (Figure 3 7C). Positive correlat ions were also obtained with 18 g C/L glucose of r 2 =0.888, r 2 =0.742 and r 2 =0.878 (Figure 3 6) and 36g C/L glucose of r 2 =0.803, r 2 =0.742, r 2 =0.821 (Figure 3 5) for Total Fe, crystalline Fe and amorphous Fe respectively. However, the correlations had the most i nfluence on the rate of Fe(II) release in the soils. To determine which of the soil Fe fractions (amorphous or crystalline) had the greater impact on the rates of Fe reductive dissolution using the approach proposed by Royer et. al (2002b) at each glucose concentration the slope values obtained through the linear regression (s) of the rate(s) of Fe(II) release versus soil Fe fraction(s) were used to calculate the ratios of the slope obtained with the amorphous Fe to that of the crystalline Fe. The determin ed ratios give an indication of how the amorphous Fe fraction impacts the reductive dissolution of Fe Based on the 9 g C/L glucose linear regressions (Figure 3 7), a value of 2.9 was obtained with the ratio of the slope of the crystalline Fe fraction (6.02 x 10 5 day 1 ) to the slope of the amorphous Fe fraction (211
63 x 10 5 day 1 ). This means that the rate of Fe(II) release in the crystalline Fe portion was about 3 times the rat e of Fe(II) release in the amorphous Fe portion. Hence with 9g C/L glucose, the greatest rate of Fe(II) release was observed in the crystalline Fe fraction. With the 18g C/L and 36g C/L glucose concentrations, the determined ratios were 4.4 and 0.65. Hence the rate of Fe(II) release was 4 times greater in the crystalline fraction compared to the amorphous fraction using 18 g C/L glucose and 0.65 times greater in the c rystalline Fe fraction using 36 g C/L. Finally in the XRD analysis of sample 2 which showed p eaks corresponding to goethite prior to use in batch experiments in the control sample (without glucose treatment), after treatment with a glucose concentration of 18g C/L (equivalent to 0.25M glucose) a decrease in the intensity of these goethite peaks in the soil sample were observed in the treated sample (Figure 3 8 ) confirming the occurrence of Fe reductive dissolution and the removal of crystallin e Fe (III) oxide minerals S imilar observation s were seen in sample 1 (not shown) and would be antici pated f or the other samples with detectable XRD crystalline Fe minerals. Characteristics of Used Landfill Leachate and its Ability to Promote Fe R eductive Dissolution The pH of the leachate was 8.23; which is characteristic of aged landfill leachate, typically i n the methanogenic phase (Kjeldsen et al., 2002) The DOC content was 1.125 g C / L and the UV absorbance measured at 254 nm (UV 254 ) was 42.2 cm 1 corresponding to a specific UV absorbance or SU VA ( SUVA 254 =UV 254 /DOC x 100 ) of 3.8 L.mg 1 .m 1 This value should be indicative of the presence of DOM originating fr om both terrestrial and microbial sources (Boyer et al., 2011; Comstock et al., 2010) T he fluorescence EEM spectrum (Figure 3 9) indicated that the leachate was dominated by
64 fulvic and humic acid like organic comp ounds based on work by Chen et al (2003) Finally, the measured COD was 2,300 mg/L, this value is very typical of aged landfill leachates containing high amounts of refractory organic compounds (Kjeldsen et al., 2002) The results of the incubation of the soils with landfill leachate are shown in (Figure 3 10) a ll of the soils with the exception of sample # 7 showed Fe release above 0.3 mg/L the secondary drinking water limit of Fe but the response of the tested soils to leachate exposure was not uniform The result s indicate that the soils are not affected by DOC in leachate and this could be due to the fact that the leachate was collected from an aged landfill which contained mostly recalcitrant humic acid like and fulvic acid like organic compounds which could n ot be easily utilized by FeRB. No clear trend was observed between the rates of Fe(II) release with landfill leachate and any of the soil amorphous or crystalline Fe fractions ( Figure 3 11 ). Relevance of the Chemistry of Used Sources of Organic Carbon in the Reductive Dissolution of Soil Iron Glucose (LMWOM) and filtered landfill leachate (primarily HMWOM) were used in conducted batch experiments. The results of these experiments have been presented and discussed above. However, looking at the same resu lts in a comparative manner after normalizing the obtained Fe reductive dissolution rates per gram of organic carbon and expressing the results as percent of total Fe in the soils (i.e. CDB extract), the following is observed (Figure 3 1 2 ). First, using normalized data obtained by spiking soil slurries with glucose at a concentration corresponding to 9g C/L on one hand and those obtained by spiking the slurries with leacha te on the other, 8 of the 12 soils gave rates
65 that fall almost on a 1:1 line (Figure 3 1 2 ). This observation suggests that for these soils, glucose can be used as surrogate for OM contained in landfill leachates to experimentally reproduce the impact landf ill s might have on the vadose zone soils in these locations On the other hand, and given the refractory character of fulvic and humic acids, one could argue that besides the prevalence of HMWOM confirmed by fluoresce nce EEM in the landfill leachate used; LMWOM compounds were present, although likely in much lower concentrations than the humics. Second, samples 4, 10, 11, and 12 released more Fe(II) when exposed to leachate landfill compared to when exposed to glucose (Figure 3 1 2 ). In fact, despite the fac t that sample 4 was placed in zone 1 in the sampling rationale detailed in Figure 3 1, its physicochemical parameters are rather similar to those of samples in zone 3 (samples 11 and 12). These samples have the lowest Fe content of the 12 soils tested, the y are very poor in OM content and are dominated by sand. Based on Fe content, Figure 3 13 suggests that sample 4 should indeed be included in zone 3 instead of zone 1. This graph helps explain the similarity observed between sample 4 and samples 11 and 12. Sample 10 on the other hand is quite different from the above three primarily in terms of AAO Fe and CDB Fe, as well as the %OM content (Figure 3 13) However, based on the data at hand it is not obvious why these samples respond better to landfill leacha te than glucose additions. Possible reasons include but are not limited to differences in rates and amount of CO 2 produced by these soils in the presence of the 2 different types of OC and the resulting impacts with regard to Fe(II) removal through sorptio n or by precipitation as siderite (FeCO 3 ). Being more biodegradable, the oxidation of glucose to form CO 2 likely overlaps with Fe(II) removal through siderite formation resulting in a small net
66 accumulation of Fe(II) in aqueous phase. This observation is s upported by the fact that increasing glucose concentrations from 9g C/L to 18 and 36g C/L led to even lower daily rates of Fe reductive dissolution and further deviations from the 1:1 line (data not shown) based on normalized data as described earlier Ove rall, this study tends to suggest that excess LMWOM could delay the bleeding of Fe(II) due to siderite formation in contrast to the HMWOM containing leachate for which the slow biodegradation would favor the buildup of Fe(II) in soil water. Accordingly, ag ed landfills might be more problematic than young ones with regard to this issue of groundwater pollution by the iron in the vadose zone of soils. Finally, in these laboratory experiments, all soils incubated with either glucose or landfill leachate as sou rce of organic carbon released Fe(II) in concentrations that exceed the secondary drinking water limit of 0.3 mg/L. Conclusions on Fe Reductive Dissolution Experiment s with Organic Matter Landfill l eachate is a complex matrix composed of organic substrates consisting of both high and low molecular compounds which are utilized as substrates for microbial respiration with soil Fe(III) oxide minerals acting as TEAs. It has been oxid es could lead to Fe reductive dissolution and release s Fe (II) in quantities that exceed the Fe secondary drinking water limit of 0.3 mg/L. Accordingly, s oil Fe(III) oxide minerals could be the sources of Fe contaminating groundwater in landfill impacted si tes. I n this study, all of the collected soils were do minated by high sand fractions and therefore characterized by high hydraulic conductivities which would allow migration of released Fe(II) from vadose zone soils to aquifers. The age of the landfill aff ects the quality of OM present in leachates and therefore the rates of microbial respiration. This has implications on the mobilization of Fe(II), with the abundance of small organic molecules favoring Fe(II) removal as siderite while HMWOM supports Fe(II) release due to slow biodegradation rates and low CO 2 production compared to LMWOM. The presence of CO 2 within the soil matrix as a result of organic matter degradation affects the amount of Fe(II) accu mulation in soil solution through siderite formation.
67 Fe(II) release rates obtained with landfill leachate and with glucose are affected not only by the type of organic matter but also by the physicochemical characteristics of soils. The release of Fe(II) were higher in soils with high sand contents compared to soils with substantial clay contents. Reduction of the soil Fe fractions were not controlled by the amount of OM added as landfill leachate. However, using glucose as a proxy for landfill leachate with prevalent LMWOM, the rate of Fe(II) release of the soils showed high correlations with both the amorphous and crystalline Fe fractions.
68 Figure 3 1. Illustration of vadose zone soil sample collection strategy. Vadose zone samples were obtained from selected locations along a transect, which emphasize d differences in soil water saturation levels (increasing from zone 1 to zone 3) and degree of crystallization of Fe minerals (decreasing from zone 1 to zone3). Based on this hypothesis, samples from zones 1 and 3 represent the two end members with 1 being the most oxidized and 3 the most affected by fluctuations of the water table. All samples were collected in North Florida. Zone 1 sites (i.e. 1, 2, 3, and 4) are current landfill sites.
69 Table 3 1. Physiochemical characteristics of collected vadose zone soil samples. Zones 1, 2, and 3 are illustrated and described in Figure 3 1 Vadose Zone Sample # pH (%) OM a AAO b Fe extract (mg Fe /kg soil ) CDB c Fe extract (mg Fe /kg soil ) (%) Clay (%) Sand (%) Silt 1 4.16 0.61 242 9,632 5 .20 84.0 10.8 Zone 1 2 6.33 0.18 53.7 2,873 2.56 94.2 3.22 3 4.40 1.25 11.7 5,594 34.6 57.4 8.08 4 7.17 0.08 69.7 490 1.20 98.3 0.50 5 4.92 2.51 3,880 7,481 24.9 69.3 5.80 Zone 2 6 5.78 1.40 1,030 3,607 5.79 87.1 7.11 7 5.6 5 2.31 1,820 6,363 13.9 81.0 5.10 8 5.31 1.32 706 1,256 3.37 92.3 4.33 9 5.52 0.79 800 1,683 4.17 92.5 3.33 10 4.29 1.51 339 1,011 1.76 96.2 1.84 Zone 3 11 5.50 0.20 66.0 183 2.55 97.5 0.00 12 5.61 0.51 87.0 173 2.56 92.8 4.64 a OM = Organic Matter b AAO = Acidified Ammonium Oxalate solution used for solubilization of the amorphous fraction of Fe c CDB = Citrate Dithionite Bicarbonate used for extraction of all Fe forms. Allows the determination of total Fe
70 Figure 3 2. XRD spectra of soils samples collected from the Klondike Landfill Site. A) Samples 1 c ollected near a monitoring well. B) Sample 2 c ollected from the pit Crystalline hematite (He) and goethite (Go) are detected in sample 1 and 2, res pectively. The abbreviations are defined as follows: HIV= hydroxyl interlayered vermiculite, K= kaolinite, Go = goethite, M= mica, He = hematite, Gi = gibbsite, and Ma = magnetite A B
71 Figure 3 3. XRD spectra of soils. A) S ample 3 collected fro m the site of the New River Regional Landfill B) S ample 4 collected from the site of the Alachua County Landfill. While no crystalline was found in sample 3, traces of magnetite (Ma) were identified in sample 4. The abbreviations on these spectra are defi ned as follows: HIV= hydroxyl interlayered vermiculite, K= kaolinite, Go = goethite, M= mica, He = hematite, Gi = gibbsite and Ma = magnetite. A B
72 Figure 3 4. Rates of Fe(II) release from the tested soils as a function of organic carbon (added as glucose) concentrations. The responses of these soils to increasing glucose concentration can be di vided into 3 distinct groups: A) Fe dissolution increases with i ncreasing glucose concentration. B) Fe dissolution decreases with i ncreasing glucose concentration. C) Fe dissolution not dependent on glucose concentration. Al l data points are averages (n=3) 1SD
73 Figure 3 5. Rates of Fe(II) release from the tested soils when treated with 36g of organic carbon/L, added as glucose. A) Rates of Fe(II) release as a function of total Fe concentrations (Citrate Dithionite Bicarbonate fraction). B) Rates of Fe(II) release as a function of the crystalline Fe fraction (determined as the difference between the CDB and the amorphous (AAO) fr actions. C) Rates of Fe(II) release as a function of the acidified ammonium oxalate (AAO) Fe fraction. Plotted data are averages (n=3) 1SD.
74 Figure 3 6. Rates of Fe(II) release from the tested soils when treated with 18g of organic carbon/L, added as glucose. A) Rates of Fe(II) release as a function of total Fe concentrations (Citrate Dithionite Bicarbonate fraction). B) Rates of Fe(II) re lease as a function of the crystalline Fe fraction (determined as the difference between the CDB and the amorphous (AAO) fractions. C) Rates of Fe(II) release as a function of the acidified ammonium oxalate (AAO) Fe fraction. Plotted data are averages (n=3 ) 1SD.
75 Figure 3 7. Rates of Fe(II) release from the tested soils when treated with 9g of or ganic carbon/L, added as glucose. A) Rates of Fe(II) release as a function of total Fe concentrations (Citrate Dithionite Bicarbonate fraction). B) Rates of Fe(II) release as a function of the crystalline Fe fraction (determined as the difference between t he CDB and the amorphous (AAO) fractions. C) Rates of Fe(II) release as a function of the acidified ammonium oxalate (AAO) Fe fraction. Plotted data are averages (n=3) 1SD.
76 Figure 3 8: Example XRD spec tra of control and glucose (18g C/L) treated s oil sample 3 after incubation under anaerobic conditions. These spectra show the presence of crystalline goethite (Go) in the absence of glucose treatment, while peaks corresponding to this mineral disappear after incubation in the presence of glucose. HIV =hydroxyl interlayered vermiculite, K=kaolinite, Go=goethite, M=mica, He=hematite, Gi=gibbsite
77 Figure 3 9. Fluorescence excitation emission matrix (EEM) spectrum of the landfill leachate used as source of organic carbon in Fe(II) release experiments. T he leachate was obtained from the New River Regional Landfill, located in Union County, Florida. The peaks in the EEM spectrum indicate that the leachate is dominated by humic acid like and fulvic acid like organic compounds.
78 Figure 3 10. Rates of Fe(II) release from different soils, each exposed to a final concentration of 1.125g of organic carbon per liter, added as landfill leachate. Incubations were conducted under anaerobic and non sterile conditions. Resul ts are averages of 3 replicates (n=3) and the bars correspond to 1 standard deviation (SD). Bars are not visible for samples with very small SD (e.g. sample 3, 5, and 12).
79 Figure 3 11. Concentration of Fe(II) released by soil treated with landfill lea chate (1.125g C /L) as a function of different soil Fe portions. A) S oil to tal Fe concentration s B ) S oil crystalline Fe concentrations C ) S oil amorphous concentrations Each point represe nts the average of 3 replicates (n=3) 1SD Bars are not visible for samples with very small SD
80 Figure 3 12. Correlation between rates of F e(II) (expressed as fraction of total Fe content) released by each of the 12 tested soils following the addition of 9 g C/L glucose and 1.125 g C/L landfill leachate.
81 Figure 3 13 Relationship between total (CDB) and amorphous (AAO) iron concentration s in analyzed soils. Numbers shown next to each point 1). Based on the sampling strategy illustrated in Fig ure 3 1: Samples 1, 2, and 3 have low Fe amorphous fractions but high total Fe concentrations (zone 1). Sample s numbered 5 to 10 (zone 2) have high Fe amorphous fractions and high total Fe concentrations. Finally, samples 11, 12 (zone 3) and 4 (zone 1) are characterized by low total Fe and low Fe amorphous concentrations. Sample 4 would be better classified as a z one 3 sample, i.e. soil s from regions affected by water table fluctuations. 9632 5594 2873 490 7481 3607 6363 1683 1256 1001 183 173 100 1,000 10,000 10 100 1,000 10,000 Total Iron (mg Fe/kg) Amorphous iron (mg Fe/kg ) 9 5 3 4 8 6 7 11 12 10 2 1 Zone 2 Zone 1 Zone 3
82 CHAPTER 4 ASSESSING THE POTENTIAL OF ABIOTIC IRON REDUCTIVE DISSOLUTION IN VADOZE ZONE SOILS USING SULFIDE AS ELECTRON DONOR Introduction Anomalously high iron (Fe) concentratio ns have been measured in groundwater samples collected from monitoring wells in watersheds impacted by both lined and unlined landfill facilities in the state of Florida. Based on water quality monitoring data, it appears that the native vadose zone soils beneath the landfills and/or the aquifer sediments could be major source s of the observed groundw ater pollution with Fe since data monitoring wate r quality show s no clear evidence of landfill leachate contamination. In vadoze zone soils found under neath u nlined landfill s tructures reducing conditions c ould develop as a result of organic rich leachate downward migration and/or the diffusion and dissolution of reduced landfill gases in the soil pores These processes w ould ultimately lead to the reduction o f Fe previously locked in the soil solid phases as Fe oxide minerals; followed by the release of the water soluble Fe 2+ and subsequent aquifer contamination However, for soils impacted by lined landfills it is likely that the reducing conditions and the r esulting dissolution of Fe could be attributable to a combination of additional factors such as : (i) changes to the natural hydrology that result from the use of impermeable liners which cut s off water infiltration pathways through the vadose zone and henc e oxygen replenishment via rainwater; and (ii) potential leaks of enough land fill leachate over time to induce reductive dissolution in the soil but without significant leachate contamination signature in the impacted groundwater. The microbial decomposit ion of organic matter in landfills consumes available terminal electron acceptors (TEAs) and is best described by the sequence of reactions
83 involved in the environmental redox ladder. Following depletion of O 2 in the soil pores, microbial oxidation of orga nic matter proceeds by being coupled with the reduction of other TEAs including sulfate, which is relevant to this specific study. In fact the reduction of sulfate leads to the formation of sulfide compounds (Plaza et al., 2007; Yao and Millero, 1996) which can in turn behave as electron donors to drive the reductive dissolution of Fe in anaerobic environments. T he non biological or abiotic reduction of Fe oxide mineral s has been chemically proven, and a good numbe r of studies on abiotic Fe reductive dissolution mechanisms have been reported in the literature (e.g. Suter et al., 1991; Poultron et al., 2004; Peiffer and Gade, 2007). Unfortunately, abiotic processes of Fe have been overlooked with regard to the pollut ion of groundwater when impacted by engineered systems such as landfills. In general, t he significance of abiotic processes of Fe in soils and sediments remains poorly understood. It is likely that reduced sulfur species play a role in the reductive dissol ution of Fe in soils and sediments. R eactions of sulfide compounds with Fe (III) oxides include the reduction of Fe 3+ surfaces with the ultimate release of Fe 2+ into solution as shown in E quations 4 1 and 4 2; where pH determines the prevalent reduced sulfu r species taking part in the reaction (Peiffer and Gade, 2007; Poulton et al., 2002) (4 1) (4 2) In fact, sulfide oxidation is a major pathway f or the reductive dissolution of Fe oxides in sulfide rich and anoxic layers of natural waters and soils (Poulton, 2003) The s ulfide species involved in the process, depending on the pH of the system, include hydrogen
84 sulfide (H 2 S), and both bisulfide (HS ) and sulfide (S ) which are produced by deprotonation of H 2 S as pH increases ( pK 1 = 7.05 for H 2 S/ HS and pK 2 = 12.89 for HS / S 2 ) as shown in E quations 4 3 and 4 4 (Firer et al., 2008) (4 3) (4 4) An additional abi otic mechanism occurs during the interaction between organic ligands and Fe 3+ surfaces which lead s to either reductive dissolution of Fe through an electron transfer process ( E quation 4 5 ) or increased dissolution rates in the presence of low molecular wei ght organic matter (LMWOM) compounds and Fe 2+ complexes as illustrated in E quation 4 6 ; where 3+ OH (s ) symbol for Fe 3+ solid surface sites. In th e latter case, since the electron transfer occurs from the Fe 2+ organic complex to the Fe 3+ surface site, there is no net increase in Fe 2+ concentrations in solution However, the process results in a significant increase in the rate of Fe reductive dissolution (Suter et al., 1991 ). (4 5) (4 6) Overall, the above brief review illustrates the potential for abiotic reductive dissolution of Fe to occur in soils It also emphasi zes the importance of key environmental parameters such as redox, pH, an d OM. Besides the impact of organic rich leachate on the soil redox and Fe dissolution in groundwater, the combination of Fe rich soils and reduced sulfur compounds could provide another path t hat could lead to Fe reductive
85 dissolution. Accordingly, an und erstanding of the se abiotic processes and determination of their significance in the reductive dissolution of Fe oxide minerals in soils is needed and is the objective of this study Materials and Methods Vadoze Soil Sample Collection and Handling Vadose zone soils used in this study were collected in North Florida from different locations including (i) sites of current landfill Fe impacted groundwater, and (ii) sites with no landfill activities as previously described in Chapter 3 Briefly, soil samples were obtained from selected locations in North Florida, and along a hypothetical transect defined to provide differences in the degree of crystallization of the Fe oxide minerals. The rationale for site selection was based on differences in the depth of th e water table and the impacts of its fluctuations on the mineralogy of iron found in these soils. The actual sampling locations of the samples are as follows. Samples 1 and 2 were collected from two different locations within the site of the Klondike Landf ill in Escambia County, Florida. This is the site of an unlined landfill which closed in 1982 after six years of operation. Sample 3 was collected from the site of the New River Regional Landfill (NRRL), a n active lined landfill in Union County, which star ted operation in 1992 (Jain et al., 2006) Sample 4 was collected from the Alachua County Southwest L andfill, a site of an unlined landfill in Alachua County, which closed in 1998 after a 10 year period of operation (Comstock et al., 2010) Therefore, zone 1 samples (Figure 3 1) came from sites with past or ongoing landfill activities. In contrast, samples from zone 2 and zone 3 came from sites with no exis ting landfill activities. These sites are located in UF/IFAS Plant Research Site in Marion County, the Ordway Swisher
86 Biological Station in Putnam County and the Austin Carey Memorial Forest in Alachua County. At each sampling site, samples were collected from deeper soil horizons (~2 meters below the surface) using an auger or using a shovel for samples obtained in locations with pre existing trenches. Following the collection process, soil samples were placed in covered high density polyethylene contain er s and transported to our laboratory at the University of Florida. In the laboratory, soil samples were air dried, then homogenized using a pestle and mortar, passed through a 2 mm sieve to remove large particles and gross organic debris (Al Abed et al., 2006) ; and then stored in sealed HDPE containers at room temperature (Kashem and Singh, 2001) until use for determination of soil physicochemical char acter istics and in biotic and abiotic Fe reductive dissolution experiments in our study Characterization of Collected Vadose Zone Soil Samples Soil particle size fractions, defined as sand (0.05 to 2 mm), silt (0.002 to 0.05 mm) and clay (< 2 mm) (USDA, 1992) were determined using a combination of the pipet method and gravimetric measurement of each size fraction after soil pretreatment with hydrogen peroxide (H 2 O 2 ) to remove organic matter (USDA, 1992, 2004) For each of the collected soil samples, pH was measured using an Accument Basic AB15 pH meter on a 1:1 (m/v) soil water slurry which had been left to equilibrate 846 Method for Soil and Waste pH (USEPA, 2004) The organic matter (OM) content of each soil sample was approxi mated as loss on ignition (LOI). The amorphous Fe conc entration of the soils was determined by Inductively Coupled Plasma Atomic Emission Spectroscopy (ICP AES) after extraction and
87 centrifugation of 0.5g of each soil sample with an acidified ammonium oxalate (AAO) solution composed of a mixture of 0.20 M amm onium oxalate ((NH 4 ) 2 C 2 O 4 )and 0.20 M oxalic acid (H 2 C 2 O 4 ) (Begin and Fortin, 2003; McKeague and Day, 1966) The total Fe oxide concentration (i.e. the combined crystalline and amorphous Fe content) of the soils was determined on citrate dithionite bicarbonate (CDB) extract s of the soil samples conducted in a water bath at 80 0 C, and followed by centrifugation and analysis by ICP AES. In the CDB procedure, the sodium dithionite (Na 2 S 2 O 4 ) acts to reduce Fe(III) in the soil samples to Fe(II) the sodium bicarbonate (NaHCO 3 ) buffers the solution (pH 7 9) and the sodium citrate (Na 3 C 6 H 5 O 7 .2H 2 O) solution chelates the released Fe(II) to prevent sorption on solid phases (Bera et al., 2005; Golden et al., 1994; Mehra and Jackson, 1960) Finally, the calculated difference between the CDB Fe and AAO Fe concentrations allowed the determination of the crystalline Fe content of the soils (Bera et al., 2005) S oil mineralogy was determi ned on clay size fractions mounted on tiles or slides which underwent x ray diffraction (XRD) using Cu K radiation (1.53 ) as generated by a computer controlled X ray diffractometer equipped with a stepping motor, a graphite monochromator and a scintill ation detector (Harris and White, 2008; Soukup et al., 2008) Samples were scanned from 2 60 0 Iron Reductive Dissolution Experiments with Dissolved Sulfide Various dissolved sulfide concentrat ions have been us ed in Fe reductive dissolution research: 0.8 0 mg S/L (equivalent to 25 uM) was used to study the effect of H 2 S on hydrous Fe (III) oxides in sea water (Yao and Millero, 1996) 15 mg S/L in the removal of H 2 S from sewage (Firer et al., 2008) 16 32.1 mg S /L (equivalent to 0.5 1mM) was used in transformations studies of mercury, iron and sulfur during the
88 reductive dissolution of goethite (Slowey and Brown, 2007) and 0.6 36 mg S /L (equivalent to 192 1136 M) for kinetic studies with ferrihydrite (Poulton, 2003) However, an initial dissolved sulfide concentratio ns of about 3 .0 1 mg S /L was selected for use in the experiments in this study due to reagent and analytical li mitations. This concentration is within agreement of the range of sulfide concentrations that have been used in the previously mentioned studie s The approach used in this study was adapted from the methodologies of Lahav (2004a) and Poulton (2003) with the use of batch studies of prepared slurries of soil and sulf ide solutions in the ratio 1:10 (m/v). Weighed soil samples were placed in serum vials whi ch were capped and then hermetically sealed. Aliquots of the init ial dissolved sulfide solutions previously prepared in glass reactors were removed and injected into th e serum vials with the use of a syringe fitted with a 2 way valve (Heitmann and Blodau, 2006; Lahav et al., 2004b) Control samples, consisting of serum via ls filled with additions of the initial dissolved sulfide solutions but with no soil addition were incubated in p arallel with the soil sulfide slu rries (Yao and Millero, 1996) After reaction times of 30 minutes, 60 mi nutes and 90 minutes, aliquots were removed from the serum vials with a syringe and added to vials containing either methylene blue reagents for sulfide analysis or with HCl for Fe(II ) extraction and analysis The colored solutions formed were analyzed at the wavelengths 665 nm for sulfide determination (Cline, 19 69) and 562 nm for Fe(II) determination (Stookey, 1970) respectively. Preparation of Sulfide Solutions All of the sulfide solutions used in this study were prepared with Nanopure water purged with ultra high purity (UHP) nitrogen gas (N 2 ) (Gu enther et al., 2001;
89 Poulton, 2003; Yao and Millero, 1996) Stock sulfide solutions were prepared with sodium sulfide crystals (Na 2 S.9H 2 O, Fisher Scientific) in volumetric flasks with Nanopure water T he initial (starting) dissolved sulfide solutions use d in the Fe reduction experiments were prepared in one liter (1L) glass reactor bottles with appropriate volumes of stock sulfide solutions added with a valve fitted syringe to achieve the desired sulfide concentrations The reactors were fitted with cust om made inlet ports for degassing, stock sulfide input, acid or base input ( 0.10M HCl / 0.1M NaOH), pH probe, magnetic stirring rod a nd outlet ports These ports helped to maintain anoxic conditions inside the reactor s to prevent sulfide oxidation upon sam ple removal and stock sulfide input. Sulfide Analysis and Sulfide Speciation C oncentrations of the stock sulfide solution s sulfide standards, the initial dissolved sulfide solution s and sulfide samples extracted from the aqueous phase of the Fe reductive dissolution experiments were analyzed by colorimetry using the methylene blue method (HACH Method 8131) (Cline, 1969; Slowey and Brown, 2007) This method involves a dditions of Reagent 1 ( sulfuric acid HACH Chemical Co. ) and Reagent 2 ( N,N dimethyl p phenylenediamine sulfate and potassium dichromate HACH Chemical Co. ) to vials containing sulfide samples In this process, N,N dimethyl p phenylenediamine s ulfate reacts with the sulfide sample to produce a methylene blue sulfide complex which has a ma ximum absorbance at =665 nm (Clesceri et al., 1989) Dissolved sulfide was measured on filtered samples (0.45 m) and total sulfide ( dissolved and FeS ) on unfiltered samples (Poulton et al., 2004) FeS concentrations w ere calculated as the differe nce between the total and dissolved sulfide
90 concentrations Similarly, at the end of each reaction time, the concentration of sulfide that had been oxidized was calculated as the difference between the in itial (starting) sulfide concentration and total sulfide concentration left unreacted (or remaining in solution) at that time (Poulton, 2003) Me asurements of the other sulfide oxidation end products were no t conducted in this study Iron Analysis Fe(II) released into solution from the Fe reductive dissolution experiments was analyzed using ferrozine, a disodium salt of 3 (2 pryidyl) 5,6 bis(4 phenylsulfonic acid) 1,2,4 triazine (C 20 H 13 N 4 NaO 6 S 2 Acros Organics). The latte r reacts with Fe(II) to form a soluble magenta complex with a maximum absorbance at =562nm (Stookey, 1970) .The ferrozine reagent was prepared by dissolving 0.1% (w/v) of ferrozine in 0.05M HEPES buffer solution ( 4 (2 hydroxyethyl) 1 piperazineethanesulfonic acid C 8 H 18 N 2 O 4 S pH 7 Fisher Scientific Chemicals). Fe(II) standards were prepared from a ferrous ammonium sulfate s tock solution (FeSO 4 .7H 2 O Fisher Scientific Chemicals ). For sample analysis, Fe in the aqueous suspensions w as extracted with a 2 way valve fitted syringe into hermetically sealed vials. Different Fe(II) fractions were determined either analytically or ca lculated as follows: Total Fe(II ) : Total Fe(II) is operationally defined as the sum of the truly dissolved Fe(II) or Fe(II) aq Fe(II) present as Fe(II) sorbed on suspended solids and FeS For the determination of the total Fe(II) concentration non filtere d aqueous samples withdrawn from static batch reactors were transferred into vials containing 5 mL of 1N HCl solution. The acid allows for the extraction of Fe(II) (also called acid extractable Fe(II) ) and it is a measure of the total amount of Fe(II) redu ced and present in the aqueous phase
91 (Chacon et al., 2006; Poulton et al., 2004) The vials were purged for 5 min with ultra high purity N 2 to remove residual sulfide and the n the mixture was left to react for 1 ho ur. Aliquots of these acidified solutions were then reacted with solutions of ferrozine HEPES as described above to determine the concentration of Fe(II). Dissolved Fe(II) F raction : Samples for the determination of the dissolved Fe(II) fraction were filter ed through 0.45m syringe filters and the filtrate was collected into hermetically sealed and nitrogen flushed vials containing 0.5N HCl (Poulton, 2003) The acidified filtrates were then analyzed with the ferrozine method as outlined above Fe(II) P resent as FeS : T he formation of FeS was calculated using experimental data obtained from the analysis of total sulfide (dissolved sulfide and FeS sulfide) The FeS Fe(II) fraction was calculated based on a stoichiometric approach derived from the concentration o f FeS sulf ide assuming a composition of Fe 1.05 S common in soils and sediments (Poulton, 2003) Fe(II) Sorbed onto Suspended S olids : Fe(II) sorbed was estimated simply as the difference between Total Fe(II) and the sum of dissolved Fe(II) and Fe(II) present as FeS (or Fe(II) FeS). Statistical Analys e s For each of the reaction times (30, 60 and 90 minutes), comparative statistical analysis of the daily rates of dissolved Fe(II) of the 12 soils were conducted using the SigmaPlot statistical software package. Significant differen ces between the means o f the rates of Fe(II) release from the soils (Table 4 3) was determined by analysis of variance (ANOVA (Beesley et al., 2010) For each soil separately comparative statistical analyse s of the Fe(II) release rates obtained at each of the reaction time s were conducted using the SigmaPlot
92 statist ical software package. F or each soil, significant differences between the means of the determined rates of Fe(II) release at each of the three rea ction times were also determined by analysis of variance (ANOVA s test (Beesley et al., 2010) Results and Discussion The average concent rations of the initial sulfide solution s and the pH of the prepared s oil water mixtures are presented in Table 4 1 The prepared mixtures had pH>9 which is adequate for the occurrence of reduced ionic sulfur species ( such as bisulfide (HS ) ) which are relevant for the reduction mechanism hypothesized in this study Measurements of the rate s of sulfide disappearance and Fe(II) appearance in the aqueous phase were conducted at times 30, 60, and 90 minutes The results obtained from these experiments are as follows. D isappearance of D issolved Sulfide from S olution Table 4 2 shows the rates of disappearance of dissolved sulfide from solution bas ed on reaction times of 30, 60 and 90 minutes. Overall, rates expressed on a daily basis show that the longer the reaction time, the slower the rate of dissolved sulfi de release (Table 4 2). This decreasing temporal trend in the rate values suggests that the net concentration of sulfide in solution is aff ected by more than one reaction First and hypothesized earlier, the highest rates recorded were based on the reactio n time of 30 minutes which correspond s to the highest rates of iron reductive dissolution shown in Tables 4 3 and 4 4 (to be dis cussed in the next subsection) This point s to the reduction of soil Fe (III) oxide mineral s coupled with the oxidation of sulfid e. In addition to Fe(II) release t he major sulfur oxidation product formed upon reaction of sulfide with Fe(III) oxide s is elemental sulfur (S 0 ) as shown in E quation 4 7 (Lahav et al., 2004b) but other
93 products which may be formed include sulfate as shown in E quation 4 8 (Firer et al., 2008) thiosulfate (S 2 O 3 2 ) or sulfite (SO 3 2 ) (Lahav et al., 2004b) as shown in E quation s 4 9 and 4 1 0, res pectively (Dos Santos Afonso and Stumm, 1992) (4 7) (4 8) (4 9) (4 10) It has been shown that iron oxide minerals display a wide variability in terms of their reactivity towards dissolved sulfide (Poulton, 2003) and at pH of 7.5 for instance the rate of reduction of Fe(III) oxide minerals by sulfide incr eases in the order goethite< hydrous ferric oxides (Peiffer and Gade, 2007) Second, in addition to the consumption of sulfide in redox reactions that lead to t he reductive dissolution of Fe the precipitation of chalcophilic metal cations as sulfide minera ls constitute ano ther sink for Fe Given the h igh concentration of Fe in the soils used it is likely that this mechanism is dominated by the formation of iron monosulfide ( FeS ) which would therefore result in the removal of both Fe(II) and reduced sulfur from solution and as shown in E quation s 4 11 (Poulton et al., 2004) and 4 12 (Peiffer et al., 1992) depending on the pH of the system (4 11) (4 12) Therefore the main pathways are (i) the production of dissolved Fe (II) by reduction of Fe(III) oxide minerals as sulfide undergoes oxidation, (ii) precipi tation of released Fe(II) as FeS ; and (iii) in microbial active environments (which is not the case under these
94 experimental conditions) a likely reaction of de novo produced sulfide by sulfate reducing bacteria with additional Fe(III) (Berner, 1970) While t he main focus of the research on sulfide generation in landfills has been on the production of H 2 S and its impact on the environment research on the implications of the occurren ce of dissolved sulfide in landfill leachate has been very limited despite the detection of dissolved sulfide in leachate plumes in concentrations as high as 33 mg/L in sulfidogenic zone s (Lyngkilde and Christensen, 1992b) In addition, laboratory experiments on synthetic C&D leachate generated dis solved sulfide in concentrations ranging from 12.8 mg/L to 21.6 mg/L (Jang and Townsend, 2003) These observations raise questions on the potential role and significance of dissolved sulfide present in landfill leachate s with regard to the current issue of groundwater pollution with Fe in landfill impacted watersheds Sulfide Induced Iron Reductive D issolution The results of Fe analysis in the aqueous phase of the different batch react ors have been tabulated (Tables 4 3 and Table 4 3) and are shown in Figures 4 1, 4 2 and 4 3. Similar to the trends of the calculated rates of sulfide disappearance from solution, the rates of iron release on a per day basis showed an overall decreasing tr end with increasing reaction time. The operationally defined Fe(II) fractions defined in the Materials and Methods section were used to track the fate of Fe(II) released to the aqueous phase. Measured Fe(II) concentrations were therefore subdivided into (i ) total Fe(II), dissolved Fe(II), iron mono sulfide or Fe(II) FeS and Fe(II) sorbed on to solid suspended soil particles. Using soil sample #8 as an example, Figure 4 1 shows the distribution trend of the different Fe(II) species in mg/L this trend was com mon to all the tested soils. It can be seen that most of Fe(II) produced by reductive dissolution process
95 occurred as the operationally defined surface sorbed Fe(II) which is calculated as the difference between Total Fe(II) concentrations and the sum of the Fe(II) dissolved and Fe(II) FeS fractions. This dominant fraction is followed by the dissolved Fe(II) fraction and the Fe(II) FeS (Figure 4 1) Using selected soil samples from zone 1 (samples 1 and 2; Figure 4 2) and zone 3 (samples 11 and 12; Figure 4 3) which are the end members of degree of crystallization of Fe oxide minerals rates of production of the different Fe(II) fractions and rates of dissolved sulfide disappearance are plotted comparatively as a function of reaction time Of interest is t he formation FeS during these experiments. With regard to sulfur, the presence of sulfide in solution favors the formation of iron sulfide minerals (Bell et al., 1987) such as the tr ansient species iron monosulfide ( FeS ) and the more stable pyrite (FeS 2 ) (Bell et al., 1987) The latte r would prevail under solution pH values that range from 3 to 9 and E h values r anging from 0.1 to 0.4 V (Garrels and Christ, 1965) Assuming that the E h and pH conditions were satisfied in the experiments and in the presence of high dissolved sulfide concentrations (Ellwood et al., 1988) it is f easible that FeS was formed in some of the soil sulfide slurries tested FeS concentrations were calculated as proposed by Poulto n, assuming a composition of Fe 1.05 S (Poulton, 2003) Based on this approach, FeS was supposed to be formed in some of the reactors, but none was visually detectable. This could be due either to the occurrence of FeS as either a transient phase (Figure 4 2 B ) or the precipitation of only small amounts (e.g. Figure 4 2 A and Figure 4 3). It is also possible that FeS was formed as amorphous FeS (aq) the precursor of FeS (s) and pyrite, FeS 2(s).
96 The amorphous FeS (aq) is a rather clear compound in solution and not a black precipitate (Rickard and Morse, 2005) While differences were observed in the absolute values of the calculated rates the overall trend remains nearly t he same for all of the samples. The near perfect overlap of the rates of Fe(II) release to that of dissolved sulfide suggests that sulfide oxidation is the main mechanism leading iron reductive dissolution These results are in agreement with decreased (di ssolved) sulfide concentrations observed in sulfide rich sedime nts in the p resence of Fe (III) oxide mineral s (Jacobson, 1994) For each soil, w hen considering the rates obtained from each of the three reaction times (30, 60 and 90 minutes) and b ased on statistical analyses, only 7 of the 12 soil sample were characterized by a signific ant decreasing trend in the determined rates of iron reductive dissolution these were samples 1,3,5,6,10,11 and 12 In contrast and while showing a decreasing trend as for the above 7 samples, rates calculated based on data obtained from the remaining 5 s oil samples were not signific antly different based on the three reaction times These samples were 2,4,7,8 and 9. Also comparative analyses of all of the 12 soils were conducted at each reaction time. At the end of the 30 min reaction time, no significant differences were found between the rates of iron reductive dissolution of any of the 12 soils. In contrast, at the end of the 60 min reaction time, significant differences exist ed between soil sample s 2 and 11; which exhibited the highest and lowest determ ined rates respectively. At the end of the 90 min reaction time, no significant differences were found in any of the soils Finally, a correlation analysi s of the determined Fe(II) release rates and the soil Fe (total Fe crystalline Fe or amorphous Fe) w as conducted with data from the 12 soil
97 samples The results are presented in Figures 4 4, 4 5 and 4 6 for rates calculated based reaction time of 30, 60, and 90 min, respectively The correlation coefficients (r 2 ) obta ined at the end of the 30 min reactio n time (Figure 4 4) were r 2 = 0.321 (total Fe), r 2 = 0.106 (crystalline Fe) and r 2 = 0.481 (amorphous Fe). The r 2 obta ined at the end of the 60 min reaction time (Figure 4 5) were 0.0483 (total Fe), 0.00151 (crystalline Fe) and 0.434 (amorphous Fe). The r 2 obta ined at the end of the 90 min reaction time ( Figure 4 6) were 0.00785 (total Fe), 0.00733 (crystalline Fe) and 0.184 (amorphous Fe). While no substantive correlation was obtained it is worth to note how the relationships become weaker with increasin g reaction time suggesting that over time, sulfide induced reductive dissolution of iron might result in significant losses of previously released Fe(II). Finally, the amor phous Fe fraction tends to be the most prone to reductive dissolution (Jacobson, 1994) Conclusion s L ined landfills are used for the disposal of municipal solid was te (MSW) and hazardous waste, in which accidental leaching could result in consequences as those simulated and discussed in Chapter 3. In such cases, organic rich leachate could also lead to the use of sulfate as TEA by sulfate reducing bacteria resulting in the production oxide minerals. On the other hand, construction and demolition (C&D) debris consists of items like wood, drywall, concrete and m etals (Jang and Townsend, 2003) Gypsum drywall is a C&D waste component which contains hydrated calcium sulfate (CaSO 4 2H 2 O), a known precursor for reduced sulfur species Rainfall infiltration through C& D wast e would dissolve calcium and sulfate from gypsum and dissolved sulfide could be formed with the use of sulfate as T EA by sulfate reducing bacteria (Plaza et al., 2007; Yao and
98 Millero, 1996) Dissolved sulfide coul d then behave as an electron donor to drive the no n biological reduction of soil Fe(III) oxides. The main findings of this research can b e summarized as follows. The expe rimental results confirm the ability of dissolved sulfide from landfill leachate to dr ive Fe reductive dissolution in soils Although the initial dissolved sulfide concentration (about 3 mg S /L) used in these experiments was on the lower end of what can be found in landfill leachates, especially when compared to sulfide concentration s f ound in C& D landfill leachate ; the tested concentration s were sufficient to release detectable amounts of Fe(II). Given the complexity of the aqueous chemistry of sulfide, it is likely that high sulfide concentrations could actually limit the concentrations of dissolved Fe(II) that can be carried downward in the aqueous phase to the aquifer by infiltrating leachate and/or rain water. The trends of iron reductive dissolution observed in this study do suggest that at sulfide concentrations as high as 12 mg S /L to 30 mg S /L common to landfill leachates Fe(II) removal from solution would likely prevail over F e reductive dissolution rates, resulting in a negligible Fe(II) pollution impact The rate of dissolved Fe(II) release in the soils was independent of the type or concentration of the soil Fe portion although the amorphous Fe fraction seemed to be more prone than the crystalline Fe to dissolve through sulfide induced reduction Sulfide disappearance from dissolution and Fe (II) release occurred in a short amount of time (30 minutes in these experiments) then decreased with increasing reaction time suggesting that the removal of dis solved sulfide and Fe (II) from the soil solution could possibly as a result of th e formation of FeS Longer reaction times are neede d to obtain the full extent of Fe reduction with dissolved sulfide since the initial precipitation of FeS was observed after 90 minutes reaction time. Also a plateau in the Fe (II) release was not o bserved but higher sulfide concentrations would likely make levels of dissolved Fe(II) negligible
99 Table 4 1. pH and analytically determined initial concentrations of sulfide in the soil slurries prepared in triplicates for each of the tested soils Soil Sample # pH of soil slurries Initial Sulfid e a Concentration (mg/L) 1 10.04 3.74 0.3 2 10.05 3.36 0.3 3 10.14 3.55 0.1 4 9.98 3.69 0.3 5 6 7 8 9 10 11 12 10.05 10.04 9.98 10.03 10.06 10.05 10.05 10.04 3.63 0.4 2.93 0.1 3.24 0.2 3.05 0.0 2.99 0.1 2.90 0.3 2.53 0.2 3.90 0.0 a Sulfide concentrations are m ean s (n=3) 1 SD.
100 Table 4 2. Rates of disappearance of dissolved sulfide (mg S.L 1 .day 1 ) in soil slurries treated with Na 2 S.9H 2 O. The rates of sul fi d e oxidation expressed in mg S. L 1 day 1 and calculated ba sed on reaction times of 30, 60 and 90 minutes. The initial concentrations of added sulfide are expressed as mg S/L for each of the tested soils are shown in Table 4 1. Soil Sample Rates ( mg S.L 1 .day 1 ) of disappearance of dissolved sulfide at: t 1 =30min a t 2 =60min a t 3 =90min a 1 846 310 449 74 268 87 .0 2 533 76 310 33 199 78.8 3 559 142 325 238 303 12 .0 4 1814 78 821 49 138 31 .0 5 589 147 220 27 306 57 .0 6 662 85 467 81 276 36 .0 7 1305 3 .0 564 34 270 101 8 976 139 496 28 311 19 .0 9 374 355 144 142 145 2 .0 10 1094 168 464 22 292 40 .0 11 5 .0 0 120 82 .0 14 64 .0 59 .0 12 1721 536 608 199 381 38 .0 a Rate values are means of 3 experimental replicates 1 SD
101 Table 4 3 Rates of iron released to the aqueous phase and present as dissolved Fe(II) m) in soil reacted with reduced sulfur solution added as Na 2 S.9H 2 O Results are expressed in mg of Fe per k g of soil per day ; a nd calculated based on 3 different reaction times (t) of 30, 60 and 90 minutes. The initial concentrations of sulfide used as electron donor for each of the soil s lurries are shown in Table 4 1. Soil Sample Rates of iron reductive dissolution ( mg Fe.kg 1 .day 1 ) at: t 1 =30min a t 2 =60min a t 3 =90min a 1 590 26.8 253 30.2 152 11.8 2 750 969 366 88.9 199 78.8 3 439 58.0 193 67.9 164 0.0 0 4 269 86.1 189 17.8 138 31.4 5 680 144 308 0.0 0 200 0.0 0 6 583 177 188 20.7 110 30.1 7 354 130 284 40.4 170 27.2 8 331 43.8 246 87.7 206 57.4 9 367 27.3 347 153 247 113 10 328 47.4 276 0.0 0 180 0.0 0 11 291 52.6 158 65.1 124 13.2 12 418 291 217 46.6 122 62.3 a Rate values are means (n=3) 1 SD
102 Table 4 4. Rates of iron reductive dissolution co ncentrations ( i.e. sum of Fe(II)aq and Fe(II)S ) in the aqueous phase of soil mixed with reduced sulfur added as Na2S.9H2O. The r ates expressed in mg Fe.kg 1 day 1 were calculated based on different experimental reaction times (t) of 30, 60, and 90 minutes. The initial concentrations of added sulfide for each of the tested soil s lurries are shown in Table 4 1. Soil Sample Rates of Fe(II) release ( mg Fe.kg 1 day 1 ) t 1 =30min a t 2 =60min a t 3 =90min a 1 959 9.48 480 87 .2 310 34.7 2 1125 1016 1410 0.00 811 222 3 851 199 425 67.9 341 65.3 4 939 53.6 474 98.1 450 136 5 968 5.30 705 2.65 495 177 6 814 177 454 9.59 307 93.4 7 865 9.48 630 159 387 102 8 1225 232 848 87.8 546 16.5 9 547 118 573 251 312 65.7 10 831 196 400 38.3 373 217 11 958 163 494 120 318 0.00 12 800 131 391 118 323 68.8 a Ra te values are averages of 3 replicates 1 SD
103 Figur e 4 1. Temporal concentration trends of the different fractions of Fe(II) released from soil treated with a solution of reduced sulfur added as Na2S.9H 2 O. Data obtained from soil s ample #8 is us ed as example to illustrate the temporal trend of T otal Fe(II ) concentration s and the phase partitioning of released Fe(II) m) and solid phases. C olumn s represent the average s of 3 replicates (n=3) and the error bars correspond to 1 standard deviation (1 SD)
104 Figure 4 2. Rate s of iron reductive dissolution from zone 1 soils expressed as mg Fe.kg 1 day 1 and rates of disappearance of reduced S from solution in mg S.L 1 day 1 A ) T rend s of rates calculated based on three different reaction times usi ng the results obtained with soil sample #1. B ) Similar to A ) but based on results obtained with soil sample # 2. Each data point represents the average of 3 replicates (n=3) and the error bars correspond to 1 standard deviation.
105 Figure 4 3. Rates of iron reductive dissolution in zone 3 soil samples expressed as mg Fe.kg 1 day 1 and rates of disappearance of sulfide in mg S.L 1 day 1 A ) Trends of rates calculated based on three different reaction times using the results obtained from soil sample #11. B ) Similar to A ) but based on results obtained from soil sample #12. Each data point represents the average of 3 replicates (n=3) and the error bars correspond to 1 standard deviation.
106 Figure 4 4 Rates of iron reductive dissolution calculated u sing data obtained after a reaction time of t= 30 minutes as a function of A) Total Fe concentration (Citrate Dithio nite Bicarbonate fraction). B) the concentration of crystalline Fe (determined as the difference between the CDB and amorphous (AAO) fractio ns ). C) concentration of the amorphous Fe (acidified ammonium oxalate or AAO fraction) Plotted data are averages (n=3) 1 SD.
107 Figure 4 5 Rates of iron reductive dissolution calculated using data obtained after a reaction time of t=60 minutes, as a function of (A) Total Fe concentration (Citrate Dithionite Bicarbonate fraction); (B) the concentration of crystalline Fe (determined as the di fference between the CDB and amorphous (AAO) fractions); and (C) concentration of the amorphous Fe (acidified ammonium oxalate or AAO fraction). Plotted data are averages (n=3) 1 SD.
108 Figure 4 6 Rates of iron reductive dissolution calculated using data obtained after a reaction time of t=90 minutes. A) As a function of t otal Fe concentration (Citrate Dithionite Bicarbonate fract ion). B) A s a function of the concentration of crystalline Fe (determined as the difference between the CDB and a morphous (AAO) fractions). C) As a function of the concentration of the amorphous Fe (acidified ammonium oxalate or AAO fraction). Plot ted data are avera ges (n=3) 1 SD.
109 CHAPTER 5 PREDICTING THE POTENTIAL OF VADOSE ZONE SOIL IRON TO UNDERGO REDUCTIVE DISSOLUTIO N Introduction The grow ing concern associated with the issue of Fe dissolution in certain landfill impacted subsurface waters and the search for re medial solutions is a perfect example of a reactive approach to deal with environmental pollution issues. As old landfills are closing and several new ones are being opened, a proactive approach is needed to minimize, and if possible eliminate, the complex ramifications associated with delayed prevention and remediation measures. Briefly, in Fe rich soils found under unlined landfills, reducing conditions could develop as a result of leachate migration into the vadose zone, and/or the diffusion of landfill gases into the pores of heterogeneous unsaturated soils, leading to the reduction of Fe previously locked into Fe(III ) oxide minerals and transfer to the aqueous phase as Fe(II) aq However, for sites impacted by lined landfills, it is likely that the reduc ing conditions and the resulting dissolution of Fe could be attributable to several factors includi ng: (i) changes to the natural hydrology that result from the use of impermeable liners which cut off water infiltration pathways through the vadose zone and oxygen replenishment via rainwater; and (ii) potential leaks of landfill leachate over time, sufficient enough to induce reductive dissolution, but without significant leachate contamination signature in the impacted groundwater. The ultimate goal of thi s study is to develop a modeling tool that use s aqueous and solid phase environmental parameters to predict the potential of any vadose zone soil to respond to landfill activities with regard to Fe dissolution and groundwater pollution. In this initial ph ase of the study, the primary objective is to (i) lay the groundwork for a solubility model that predict s solution concentration of Fe based on a
110 pre determined set of soil physicochemical characteristics; and (ii) test the potential of multi parameter re Multiple linear regression s are often used in the attempt to model the relationship between two or more variables and a response variable by fitting a linear equation to observed data. Ever y value of the independent variable x is in this case associated with a value of the dependent variable y The population regression line for explanatory variables x 1 x 2 ..., x n is defined as : y = o + 1 x 1 + 2 x 2 n x n (5 1) E quation 5 1 can be used to describe changes of the mean response y as a function of key variables ( x 1 x 2 ,..., and x n ). The identified variables can be investigated fu rther to generate the parameters that could be most relevant to the geochemical modeling effort. In fact, s everal models that predict the concentration or activity of metals in soil pore waters have been published (El Falaky et al., 1991; Ma and Lindsay, 1993; Jopony and Young, 1994; Sauve, 1999; Sauve et al., 2000). Parameterization of these models has been based on easily measurable soil characteristics such as total metal content, pH, and soil OM. More mechanistic models with greater input requirements have also developed as an extension to the chemical equilibrium program CHARON (Hesterberg e t al., 1993). However, there are limitations to the general and widespread application of
111 content of Al, Fe, Si, and Mn oxides and humus in soil and estimation of the specific sur face area of each constituent. Algorithms previously proposed by others for prediction of metal solubility in soils based on information commonly included in soil geochemical surveys include the activity model by Tye et al. (2003), illustrated in E quation 5 2 and where pFe 2+ = log 10 (Fe), OC = organic carbon content of the soil in %, Fe soil = iron content of the soil, Fe 2+ = activity of ferrous iron in soil pore water, N is the n th relevant parameter impacting Fe dissolution; and k i and m are model fittin g constants specific to iron. (5 2 ) The advantage of this solubility expression is the flexibility to include all relevant parameters affecting iron reductive dissolution. It can therefor e be fed by data obtained from laboratory experiments and can be used by engineers, geologists, landfill operators and regulators to determine whethe r Fe reductive dissolution could pose a problem at any given site targeted for landfill development. Materi als and Methods Hematite as Source of Iron C haracterization A synthetic Fe mineral, hematite, was used in anaerobic laboratory incubations in the presence of a consortium of bacteria obtained from an anaerobic digester to determine the rate constants ( k ) o f Fe reductive dissolution as a function of increasing organic carbon (glucose) proton (pH), and salt (ionic strength) concentrations. The hematite sample was obtained from Fisher Scientific (USA) and used in lab oratory experiments as received. Previously published X Ray diffractometer (XRD) techniques
112 (Harris and White, 2008; Soukup et al., 2008) were used to characterize the hematite sample Briefly a h ematite suspension prepared in a 1M NaCl solution (adjusted t o pH 10) was mounted on tiles and either K or Mg saturated, and washed with de ionized water and then with a 30% glycerol solution for Mg saturated samples. Samples were irradiated with a 1.53 line from a copper target using a computer controlled XRD equi pped with a stepping motor, a graphite monochromator, and a scintillation detector which scans up to 60 d spacing for mineral identification. Fe R e ductive Dissolution E xperiments Based on the disc ussion provided in earlier chapters on the reductive dissolution of iron, laboratory experiments were designed to assess the potential of solid phase iron to undergo dissolution under anaerobic conditions as a function of selected environmental parameters (OM, pH, and ionic strength). The focus was limited to the potential soil liquid interactions and microbial induced reduction of Fe(III) oxide minerals. Using the experimental approach described in Chapter 3, experiments were conducted using hematite as so lid phase mineral, and the effect of organic matter, pH and ionic strength on Fe reductive dissolution assessed using inoculum of a consortium of bacteria from an anaerobic dige ster. The results obtained from these experiments were then used to generate fi tting constants specific to Fe which can be used in a multi parameter equation to predict the potential of Fe in any of t he 12 soil samples to undergo reductive dissolution. Analysis of Fe At the end of the incubation time, samples were placed inside an an aerobic glove box, and using a syringe, aliquot samples were withdrawn from experimental bottles,
113 concentrations determined by spectrometry using the ferrozine method (Roden, 2004; Stookey, 1970) Briefly, ferrozine, a disodium salt of 3 (2 pyridyl) 5,6 bis(4 phe nylsulfonic acid) 1,2,4 triazine, reacts with Fe(II) to form a rather stable magenta complex which is soluble in water and therefore used for the direct determination of Fe(II) in water. The visible absorption spectrum of the ferrous complex of ferrozine e xhibits a single sharp peak with a maximum absorbance at 562 nm. Predictive T ools First, the physicochemical data obtained from the analysis of collected soils (Chapter 3, Table 3 1) were used in a multiple linear regression analysis to identify the param eters that correlated best with the concentrations of Fe(II) released by the tested soils. Using data presented in Chapter 3, the best fit was obtained by correlating the the oxalate Fe extract (AAO Fe concentration or CDB Fe. This exercise resulted in E quation 5 3, with r 2 values of 0.968, 0.932 and 0.863 for %0C, amorphous or AAO Fe and total or CDB Fe, respectively [Fe] = 0.744 2.778 [%OC] + 0.001 63 [Ox Fe] + 0.000569 [CDB Fe] (5 3) Based on E quation 5 3 organic carbon content was selected as a key soil parameter in addition to Fe concentrations used to predict the concentrations of iron that might be released by vadose zon e soils when exposed to landfill leachate. In addition to OM, the potential effects of pH and ionic strength were also assessed. The potential of vadose zone soils to release iron as Fe(II) as a function of the above listed parameters was estimated using a modified version of E quation 5 2. Modifications were based on our experimental observations (data discussed in in previous chapters 3 and
114 4, and E quation 5 3). Accordingly, a ( Fe soil ) and the organic c arbon content (%OC); while minimizing the weight of other parameters such as pH and ionic strength (E quation 5 4). (5 4) Finally, the model generated Fe(II) concentrations were compared to the expe rimental data obtained by iron reductive dissolution in tested soils, induced by landfill leachate. Results and D iscussion Hematite as Source of Iron C haracterization The XRD analysis of used hematite is given in Figure 5 1. The analyzed sample showed mica (M) and quartz (Q) impurities in addition to peaks of Fe minerals other than hematite (e.g. peaks shown by a question mark, which are believed to be ferromagnesian type minerals, and G for goethite). Fe Reductive Dissolution from Hematite as a Function o f Key Environmental P arameters To determine the parameters specific to iron reductive dissolution as stipulated in Equation 5 4, batch experiments were conducted. In this case, hematite instead of soil samples was used as the solid phase and depending on t he specific parameters under consideration the procedures outlined below were followed. Glucose as Source of Organic C arbon : The aqueous solution used to test the effect of organic matter on Fe reductive dissolution in these experiments consisted of N anop ure water spiked with glucose (as source of organic carbon) and a consortium of bacteria from the anaerobic digester. The aqueous phase was equilibrated with
115 hematite as solid phase in a 1:10 ratio (m/v). Each treatment was repeated 5 times including cont rol samples which were similar in chemical composition as treatments but with no glucose added. The analysis of the aqueous phase at the end of a 2 week (14 day) incubation period at room temperature and under anaerobic conditions showed a linear relations hip between the concentrations of dissolved Fe(II) and that of glucose (Figure 5 2). The constant k oc (Equation 5 4) is therefore given by the slope of this relationship and is equal to 137.58 (Figure 5 2). pH : To test the effect of pH on biotic iron reduc tive dissolution, batch experiments similar to those described for glucose were used, except that in this case, pH was the only variable, adjusted to specific values (4, 7, and 9) by use of either 0.1N NaOH or 0.1N HCl. To make sure that the Fe reductive d issolution process was not limited by other parameters other than pH, all liquid hematite mixtures were spiked with (i) a similar aliquot volume of the glucose solution to bring the final concentration of organic carbon to about 150 mg C/L; and (ii) a cons ortium of bacteria from the anaerobic digester. The aqueous phase was equilibrated with hematite as solid phase in a 1:10 ratio (m/v). Each treatment was replicated 5 times. In contrast to the above glucose based experiments, each treatment had an addition al control whose chemical composition was similar to the treatments (same pH), but with no bacteria addition to help account for the non biological dissolution. Values obtained from controls, if any, were then subtracted from that of the corresponding trea tments. The analysis of the aqueous phase at the end of a 2 week incubation period at room temperature and under anaerobic conditions showed a linear relationship between the concentration of dissolved Fe and increasing concentrations of protons (Figure 5 3).
116 Ionic strength : Batch experiments to assess the effect of increasing salt concentrations on the biotic reductive dissolution of iron were designed in a similar manner as the above pH experiments. However, in this specific case, all parameters were mai ntained constant in all reactors while the ionic strength (determined as: ) varied from 0.11, 0.22, and 0.33 M, using a macro mineral solution (NaCl, KH 2 PO 4 CaCl 2 and MgCl 2 ) containing the major mono and divalent cations commo n in soil pore waters and landfill leachates. The aqueous phase was equilibrated with hematite as solid phase in a 1:10 ratio (m/v). Each treatment was repeated 5 times. Each treatment had a control whose chemical composition was similar to the treatment ( same ionic strength), but with no bacteria addition to help account for the non biological effect. Values obtained from controls were subtracted from that of the corresponding treatment. The analysis of the aqueous phase at the end of a 2 week incubation p eriod at room temperature and under anaerobic conditions showed a negative linear relationship between the concentration of dissolved Fe and ionic strength (Figure 5 4). Based on Equation 5 3, this experiment allows for the determination of an example k n equal to 91.545 in this case. The above results (Figs. 5 2, 5 3, and 5 4) show the effects of 3 key parameters on the reduction of Fe(III) present in hematite inside the reactors The lack of Fe(II) ( aq ) in vials without bacteria added confirmed the fact that the dissolution process is microbial driven. It is worth noting that all of these experiments were conducted in hermetically sealed vials. Therefore, the buildup of CO 2 from microbial respiration of added glucose (as discussed in Chapter 3) could impa ct the speciation and phase partitioning of Fe. In fact, the increase in P CO2 in these hermetically sealed vials could potentially lead to the
117 buildup of CO 2 resulting from microbial respiration of glucose and favor the formation of siderite (FeCO 3 ) in sol utions with pH of about 6 and greater (Drever, 1997). The formed siderite would then result in Fe 2+ removal fr om solution via precipitation (E quation 5 4). Fe 2+ + CO 3 2 FeCO 3 (5 4) This would decrease the levels of dissolved Fe(II), by decreasing its concentrations in the aqueous phase. Second, for non buffered systems (similar to those used in this study), it is likely that the increase in P CO2 in sealed vials would also induce a decrease in pH due to the formation of carbonic acid (H 2 CO 3 ), and the extent of this reaction may limit the formation of siderite, or ultimately dissolve the previously formed siderite as illustrated in E quation 5 5 (Drever, 1997). FeC O 3 + 2H + Fe 2+ + H 2 O + CO 2 (5 5) monitored over time nor measured at the end of the incubation period. Althou gh the above observations do call for additional measurements (e.g. CO 2 production and temporal changes in pH, to name a few); the strong relationships observed between dissolved Fe(II) levels and concentration gradients of OM and ionic strength suggest th at the impact of CO 2 could be negligible. In contrast, the decreasing concentrations of dissolved Fe(II) with increasing pH could be explained, at least partly, by Fe(II) removal via precipitation as siderite (FeCO 3 ). Model V alidation : Using E quation 5 3 a nd constants determined from Figures 5 2, 5 3, and 5 4, the potential of solid phase iron minerals present in each of the tested soil samples to undergo biological reductive dissolut ion was calculated based on the
118 AAO and the CDB chemically extracted iron. The results are presented in Table 5 1. This preliminary model validation effort led to the following observations: In both cases (i.e. using either AAO Fe or CDB Fe), the model over predicts the Fe(II) concentrations that can be released per kg of soil, when compared to data obtained by reacting landfill leachate with tested soils The Fe AAO gives a slightly much better prediction of the concentration of Fe that can be dissolved by landfill leachate stimulated biotic reductive dissolution However, this is true only for 5 out of the 12 soil samples tested. It is obvious that a much larger data set is needed for an adequate validation and fine tuning of the model. With regard to the lab generated data, the extension of the incubation period to times that exceed 2 weeks could result in much higher Fe(II) release and produce a better match of the model prediction. Therefore, with this limited data set, the effort can only be considered preliminary and a sensitivity analysis could not be performed. Results pr esented in this chapter are relevant for conditions where landfill leachate would come in contact with soil Fe(III) oxide minerals. In absence of leachate, we hypothesized earlier that abiotic reactions involving reduced sulfur compounds could take place a nd potentially lead to iron reductive dissolution. This latter aspect has not been taken into account since relevant experimental data were not obtained with sulfide. C onclusions The process of developing a predictive tool for assessing the potential of v adoze soil Fe to undergo reductive dissolution under conditions commonly encountered in landfill impacted environments was initiated. Limited to biotic processes, the model over predicted the amount of Fe released experimentally by soils reacted with landf ill leachate. Specifically, the model over predicted the soluble Fe(II) concentrations of soils with very low capacity. Since these soils all had very high total (CDB) and amorphous (AAO) Fe concentrations it is likely, that parameters speci fic to Fe bonding to the different components of the soil matrix need to be investigated and incorporated into the model.
119 The 2 week incubation period used in this study may not be ideal for all soil types, assuming that a much longe r time might be needed for the low Fe bleeding capacity soils to significantly release Fe(II) into aqueous phase via reductive dissolution. The results of sulfide induced Fe(III) reduction resulted in a temporal decreasing trend, and therefore a negative slope. The incorporati on of a sulfide term into the model could likely help decrease the absolute values of the model predicted concentrations and improve the possibility of matching the experi mental data. Unfortunately, due to the lack of adequate laboratory settings, experime nts to determine the rate constant of iron reductive dissolution by reaction with sulfide using synthetic Fe oxide minerals were not conducted. Finally, the very small number of samples tested limited the efficiency and accuracy of the model validation ef fort. Accordingly, future fine tuning work is necessary to meet the ultimate goal of a robust predictive tool. The following could contribute new knowledge and help improve the forecast effort: Increase the number of soil samples to 50 or more Determine t he ideal incubation period for each sample type and use the longest time for all samples Analyze the soil samples for TFe, Fe AAO, and XRD not only before, but also after the reductive dissolution experiments to identify the Fe fractions affected by the re ductive dissolution.
120 Table 5 1. Concentrations of iron released per kg of soil treated with landfill leachate as amorphous (AAO) iron using the 12 tested soil samples. Soil sample mg Fe (II) released per kg of soil treated with landfill leachate Predicted mg of Fe (II) released based on CDB Fe Predicted mg of Fe (II) released based on AAO Fe 1 2.65 29.31 12.86 2 1.83 22.58 5.68 3 12.60 35.45 14.55 4 3.47 30.76 19.89 5 1.83 21.56 1 8.62 6 2.94 22.48 16.98 7 0.11* 22.56 17.05 8 3.06 21.06 17.84 9 1.83 17.27 15.19 10 11.10 14.59 11.45 11 8.87 17.41 13.86 12 9.12 14.07 12.07 Values corresponding to the analytical method detection limit or MDL
121 Figure 5 1 XRD spectra of the hematite sample used in reductive dissolution experiments to help develop a Fe dissolution model as a function of organic matter, pH, and ionic strength.
122 Figure 5 2 Fe released from hematite exposed to aqueous solutions containing increasing con centrations of organic carbon added as glucose in the presence of a consortium of bacteria from an anaerobic digester. Produced slurries were incubated anaerobically for 2 weeks. Each point on the graph represents the average of 5 replicates (n=5) and t he bars correspond to 1 standard deviation. [Fe 2+ ] aq = 137.58[OC] r = 0.9687 0 5000 10000 15000 20000 25000 30000 35000 0 50 100 150 200 250 Dissolved Iron (mg Fe 2 /L)*10 3 Organic Carbon (mg C/mL)
123 Figure 5 3 Fe released from hematite exposed to aqueous solutions of increasing concentrations of protons (i.e. decreasing pH) and containing bacteria from an anaerobic digester. Mixtures were incubate d anaerobically for 2 weeks. Each point on the graph represents an average of 5 replicate s (n=5) and the bars correspond to 1 standard deviation. y = 0.0483e 0.85pH R = 0.9947 0 20 40 60 80 100 120 3 4 5 6 7 8 9 10 Dissolved iron (mg Fe(II)/L) pH
124 Figure 5 4 Fe released from hematite exposed to aqueous solutions of increasing ionic strengths and co ntaining bacteria from an anaerobic digester. Mixtures were incubated anaerobically for 2 weeks. Each point on the graph represents the average of 5 replicate s (n=5) and the bars correspond to 1 standard deviation (1SD) 60 65 70 75 80 85 90 95 100 105 0 0.1 0.2 0.3 0.4 Dissolved iron (mg Fe 2+ /L) Ionic Strength (M)
125 CHAPTER 6 GENERAL CONCLUSIONS AN D RECOMMENDATIONS FOR FUTURE WORK This study initiated the process of developing a predictive tool for assessing the potential of soil Fe to undergo reductive dissolution under conditions anticipated in landfill impacted soil environments. The experimental and modeling approaches used are summarized in Figure 6 1 The s pecific objectives were to investigate both the biotic and abiotic mechanisms of iron reductive dissolution in vadose zone soils with emphasis on the role of the degree of crystallization of Fe(III) oxide minerals. The experimental results revealed that regardless of the degree of crystallization and phase distribution oxide minerals, both biotic and abiotic iron reductive dissolution experiments could potentially lead to dissolved Fe(II) concentrations above the Fe secondary drinking water limit of 0.3 mg/L. Theoretically, these results point to the possibility of both microbial and chemical catalyzed iron reductive dissolution in landfill impacted soils and the potential oxides to act as source of Fe(II) that is contaminating groundwater as observed in specific sites in Florida. However, the significance and the extent of the above processes with regard to the observed Fe pollution of groundwater remain unclear. For instance, in studies focusing on types and quantities of organic matter used as source of energy by FeRB it has been shown that leachate from old landfills favor the accumulation of Fe(II) in aqueous phase as compared to leachate from young landfills which contain abundant LMWOM. Much h igher percentages of Fe were released as dissolved Fe(II) when soil were treated with the leachate of the relatively old New River Regional Landfill. In fact, a ged landfill leachate is dominated by high molecular weight organic
126 matter (HMWO M ) w hich do not degrade readily to produce large concentrations of CO 2 as would LMWOM (e.g. glucose lactate, or acetate); and to remove as much Fe(II) from soil solution as would LMWOM The c hemical (or abiotic ) reduction of soil Fe(III) oxide mineral s has been demonstrated with the use of dissolved sulfide in hermetically sealed batch reactors T his could have implications for Fe rich soils coming to contact with sulfide rich leachate such those expected from C &D landfills. However, the abiotic release of Fe(II) was observed by spiking soils with relatively low concentrations of sulfide ( ~ 3 mg/L on average ), and the obtained temporal trends of iron reductive dissolution in soils suggested that the occurrence of high sulfide concentrations could actually eliminate the accumulation of Fe(II) in the aqueous phase due primarily to chemical precipitation with chalcophile metal cations, primarily iron (e.g. formation of transient FeS species ). This is validated by the overlap of the temporal trends of Fe(II) and dissolved sulfide in tested soils However, formed colloidal FeS could be prone to downward transport to reach the aquifer, and likely to remain stable under anaerobic conditions characteristics of most aquifers The ability of the cryst alline Fe or amorphous Fe to be released through reactions with organic matter or sulfide was assessed using linear regression analyses. The experimental results obtained indicate d that the amorphous Fe was more susceptible to Fe reduction. So far, the model tends to over predict the soluble Fe (II) concentrations that can be released by natural soils namely soils with very low capacity. I t is likely, that parameters specific to Fe bonding to the different components of the soil
127 matrix as well as the significance of the type of organic matter used in lab experiments need to be investigated ; and the model updated accordingly Alternatively, the 14 day incubation period used in this study may not be ideal for all soil t ypes, assuming that a much longe r time might be needed for the low Fe bleeding capacity soils to significantly release Fe into aqueous phase. Finally, the small number of samples and inadequate Fe solubility potentials limited the efficiency and accura cy of the model validation effort. Accordingly, future fine tuning work is necessary to meet the ultimate goal of a robust predictive tool. Future research avenues should focus on: Increasing the Number of Soil S amples (up to 100 ) : To develop a more accur ate prediction model, the number of samples must be increase d significantly for both validation and sensitivity analysis studies In this study, 12 samples were used and although variation existed in the degree of crystal lization of Fe(III) oxide minerals, there was less variety in terms of types of Fe minerals present in soils. Addressing the above listed weaknesses in future research is necessary. Determining the Ideal Incubation Period for Each Sample Type and Using Longer R eaction T ime s for All S amples : P eriod s of 2 to 4 weeks were used for the experiments conducted in this study, but temporal measurements were not made to make sure that equilibrium and the maximum amount of Fe that could be released from each soil got dissolved. Physicochemical A nalyse s : Analyzing the soil samples for TFe, Fe AAO, and XRD not only before, but also after the reductive dissolution experiments to
128 identify the Fe fractions affected by the reductive dissolution. This will help to get a better understanding of the extent of F e release in these soils and whether the predominant mechanism of Fe reductive dissolution is occurring. XRD or ion exchange could be used to determine the predominant redox process at the beginning and end of the experiments. Sulfide as an Electron D onor : Evaluate the significance of Fe reductive dissolution by reduced sulfur compounds using concentration gradients and determine if it is necessary to account for this abiotic process in the modeling effort. Looking F orward : Columns should be used for hydro dynamic studies, which are much closer to mimicking in situ conditions as compared to batch experiments. In the later, equilibrium will be reached much faster, conditions that limit further iron reductive dissolution reaction to occur. Column studies bette r imitate the in situ conditions of leaching by intermittent leaching by rainwater
129 Figure 6 1. Comparative flow chart of the experimental (left) and modeling (right) studies conducted in this research. Batch studies were used to assess the significa nce of biotic and abiotic in Fe reductive dissolution. Hematite was used as source of Fe(III) in biotic batch experiments designed to generate rate constants of chemical reactions driven by selected environmental parameters to be used in the model. Abiotic experiments with hematite and a concentration gradient of sulfide was not completed within the time period of this study. Therefore, the predicted Fe(II) concentrations are based only biotic process, resulting in an over estimation of the levels of Fe(II) released by natural soils.
130 APPENDIX A SUPPLEMENTARY TABLES Table A 1. Rates of dissolved Fe(II) (mg/kg.day) re leased in soil slurries. The rates expressed in mg Fe.kg 1 day 1 are normalized to 1g C/L and we re calculated based on the organic carbon con centrations of 1.125 g C /L of leachate or 9, 18 or 36 g C /L of glucose. Soil Sample # Leachate (1.125 g C/L) Glucose concentration (9 g C/L) Glucose concentration (18 g C/L) Glucose concentration (36 g C/L) 1 1.68 6.06 3.49 3.09 2 8.03 3.63 2.27 1.80 3 1.16 3.07 1.67 3.43 4 2.20 0.59 0.22 0.12 5 1.16 9.32 3.14 2.14 6 1.86 3.17 1.84 1.01 7 0.07 3.31 2.26 2.00 8 1.16 1.28 0.55 0.26 9 1.94 1.30 0.74 0.44 10 7.05 0.02 0.75 0.37 11 5.63 0.26 0.09 0.03 12 5.82 0.40 0.26 0.05 Rate values are avera ges of 3 replicates (n=3) 1SD
131 Table A 2 Rates of dissolved Fe(II) (mg/kg.day) released in soil slurries expressed in percentages (%) of the soil total Fe concentrations with NRRL a Leachate (at 1.12 g C/L ), Glucose ( at 9,18 and 36 g C/L) and Sul fide ( with 3 1 mg S/L at 30, 60 and 90 minute reaction times). Soil Sample # Leachate (1.12 g C/L) b Glucose Concentration (9 g C/L) b Glucose Concentration (18 g C/L) b Glucose Concentration (36 g C/L) b Sulfide t 1 =30min b Sulfide t 2 =60min b Sulfide t 3 =90min b 1 0.02 0.06 0.04 0.03 0.13 0.11 0.10 2 0.28 0.12 0.08 0.06 0.54 0.53 0.43 3 0.02 0.05 0.03 0.06 0.16 0.14 0.18 4 0.45 0.12 0.05 0.02 1.14 1.61 1.77 5 0.02 0.12 0.04 0.03 0.19 0.17 0.17 6 0.05 0.09 0.05 0.03 0.34 0.2 2 0.19 7 0.00 0.05 0.04 0.03 0.12 0.19 0.17 8 0.09 0.10 0.04 0.02 0.55 0.81 1.02 9 0.12 0.08 0.04 0.03 0.45 0.86 0.92 10 0.70 0.00 0.07 0.04 0.68 1.14 1.11 11 3.11 0.14 0.05 0.02 3.31 3.60 4.25 12 3.38 0.23 0.15 0.03 5.03 5.22 4.42 Rate values are a verages of 3 replicates (n=3) 1SD
132 APPENDIX B SUPPLEMENTARY FIGURES Figure B 1. XRD spectra of Sample 5 collected location 1 of the UF/IFAS Plant Science Research and Education Unit. A) Sample 6 collected from the E horizon. B) Sample 7 collecte d from the Bt horizon. The abbreviations are defined as follows: HIV= hydroxyl interlayered vermiculite, K= kaolinite, Go = goethite, M= mica, He = hematite, Gi = gibbsite, and Ma = magnetite, F= feldspar, C= crandillite.
133 Figure B 2. XRD spectra of samples collected location 2 of the UF/IFAS Plant Science Research and Education Unit. A) Sample 6 collected from the E horizon. B) Sample 7 collected from the Bt horizon. The abbreviations are defined as follows: HIV= hydroxyl interlayered vermiculite, K= kaolinite, Go = goethite, M= mica, He = hematite, Gi = gibbsite, and Ma = magnetite, F= feldspar, C= crandillite
134 Figure B 3. XRD spectra of sample s c ollected from location 1 at the Austin Carey Memorial Forest A) Sample 8 collected from horizon AE. B) Sample 9 collected from horizon E2. The abbreviations are defined as follows: HIV= hydroxyl interlayered vermiculite, K= kaolinite, Go = goethite, M= mica, He = hematite, Qz = quartz, Gi = gibbsi te, and Ma = magnetite, F= feldspar C = crandillite
135 Figure B 4 XRD spectra of sample 10 collected from the Ordway Sw isher Biological Station The abbreviations are defined as follows: HIV= hydroxyl interlayered vermiculite, K= kaolinite, Go = go ethite, M= mica, He = hematite, Qz = quartz, Gi = gibbsite, and Ma = magnetite F= feldspar.
136 Figure B 5. XRD spectra of sample 11 collected from horizon E3 at location 1 of the Austin Carey Memorial Forest. The abbreviations are defined as follows: HIV= hydroxyl interlayered vermiculite, K= kaolinite, Go = goethite, M= mica, He = hematite, Qz = quartz, Gi = gibbsite, and Ma = magnetite, F= feldspar, C = crandillite. Figure B 6 XRD spectra of sample 12 collected from horizon E3 at location 2 of t he Austin Carey Memorial Forest. The abbreviations are defined as follows: HIV= hydroxyl interlayered vermiculite, K= kaolinite, Go = goethite, M= mica, He = hematite, Qz = quartz, Gi = gibbsite, and Ma = magnetite, F= feldspar, C = crandillite
137 Figure B 7. Rates of iron reductive dissolution from zone 1 soils expressed as mg Fe.kg 1day 1 and rates of disappearance of reduced S from solution in mg S.L 1 day 1 A) Trends of rat es calculated based on three different reaction times using the results obtained with soil sample #3 B) Similar to A) but based on results obtained with soil sample # 4 Each data point represents the average of 3 replicates (n=3) and the error bars corres pond to 1 standard deviation.
138 Figure B 8. Rates of iron reductive dissolution from zone 2 soils expressed as mg Fe.kg 1 d ay 1 and rates of disappearance of reduced S from solution in mg S.L 1 day 1 A) Trends of rates calculated based on three different reaction times using the results obtained with soil sample #5 B) Similar to A) but based on results obtained with soil samp le # 6 Each data point represents the average of 3 replicates (n=3) and the error bars correspond to 1 standard deviation.
139 Figure B 9. Rates of iron reductive dissolution from zone 2 soils expressed as mg Fe.kg 1 day 1 and rates of disappearance of reduced S from solution in mg S.L 1 day 1 A) Trends of rates calculated based on three different reaction times using the result s obtained with soil sample #7. B) Similar to A) but based on results obtained with soil sample #8. Each data point represents the average of 3 replicates (n=3) and the error bars correspond to 1 standard deviation.
140 Figure B 10 Rates of iron reductive dissolution from zone 2 soils expressed as mg Fe.kg 1 day 1 and rates of disappearance of reduced S from solution in mg S.L 1 day 1 A) Trends of rates calculated based on three different reaction times using the results obtained with soil sample #9 B) Similar to A) but based on results obtained with soil sample # 10 Each data point represents the average of 3 replicates (n=3) and the error bars correspond to 1 standard deviation
141 LIST OF REFERENCES Al Abed, S.R., Hageman, P.L., Jegadeesan, G., Madhavan, N., Allen, D., 2006. Comparative evaluation of short term leach tests for heavy metal release from m ine ral processing waste. Sci. Total Environ. 364, 14 23. Albaiges, J., Casado, F., Ventura, F., 1986. Organic indicators of groundwater pollution by a s anitary landfill. Water Res. 20 1153 1159. Alloway, B.J., 1995. Heavy metals in soils, 2 ed. Springer, New York. Baltpurvins, K.A., Burns, R.C., Lawrance, G.A., Stuart, A.D., 1996. Effect of pH and Anion Type on the Aging of Freshly Precipitated Iron(III) Hydroxide Sludges. Environ Sci. Technol. 30 939 944. Banwart, S., Davies, S., Stumm, W., 1989. The role o f oxalate in accelerating the Fe2O3) by asc orbate. Colloid Surf. 39 303 309. Barbosa, F.P.T., Luna, R.C.P., Monteiro, M.G.C.A., Asciutti, L.S.R., de Oliveira, R.S.S., Diniz, A.d.S., Lima, R.T., Benigna, M.J.C.d.N., de Moraes, R.M., Costa, M.J.C., Gonalves, M.d.C.R., 2013. Relationship between hemoglobin, serum retinol and habitual meat consumption in the elderly: A popu lation based study. Arch. Gerontol. Geriatr. 57 60 65. Baun, D.L., Christensen, T.H., 2004. Speciati on of Heavy Metals in Landfill Leac hate: A Review. Waste Manage. Res. 22 3 23. Beesley, L., Moreno Jimnez, E., Clemente, R., Lepp, N., Dickinson, N., 2010. Mobility of arsenic, cadmium and zinc in a multi element contaminated soil profile assessed by in situ soil pore water sampling, column leaching and sequ ential extraction. Environ. Pollut 158 155 160. Begin, L., Fortin, J., 2003. Evaluation of an Acid Ammonium Oxalate Extraction to Determine Fluoride Resident Concentrations in Soils. J. Environ. Qual 32 662 673. Bell, P.E., Mills, A.L., Herman, J.S., 1987. Biogeochemical Conditions Favoring Magnetite Formation during Anaerobic Iron Reduction Appl. Environ. Microbiol. 53 2610 2616. Benner, S.G., Hansel, C.M., Wielinga, B.W., Barber, T., Fendoerf, S ., 2002. Reductive Dissolution and Biomineralization of Iron Hydroxide under Dynami c Flow Conditions. Environ. Sci. Technol. 36 1705 1711. Bennett, B., Dudas, M.J., 2003. Release of arsenic and molybdenum by reductive dissolution of iron oxides in a soil with enriched levels of arsenic. J. Environ. Eng. Sci. 2 265 272.
142 Bera, R., Seal, A., Banerjee, M., Dolui, A.K., 2005. Nature and profile distribution of iron and aluminum in relation to pedogenic processes in some soils developed under tropical env ironm ent in India. Environ. Geol. 47 241 245. Berner, R.A., 1970. Sedime ntary pyrite formation. Am. J. Sci. 268 1 23. Bonneville, S., Behrends, T., Van Cappellen, P., 2009. Solubility and dissimilatory reduction kinetics of iron(III) oxyhydroxides: A linear f ree energy relationship. Geochim.Cosmochim. Acta 73 5273 5282. Bonzongo, J. C., Townsend, T., 2008. Reductive Dissolution of Iron in Impacted Soils and Aquifers: Mechanisms and Development of a Predictive Model. Hinkley Center for Solid and Hazardous Wast e Management, Florida. Boyer, T.H., Graf, K.C., Comstock, S.E.H., Townsend, T.G., 2011. Magnetic ion exchange treatment of stabilized landfill leachate. Chemosphere 83 1220 1227. Brunori, C., Cremisini, C., Massanisso, P., Pinto, V., Torricelli, L., 2005. Reuse of a treated red mud bauxite waste: studies on environm ental compatibility. Hazardous Waste. Hazard. 117 55 63. Cajuste, L.J., Cruz Diaz, J., Garcia Osorio, C., 2000. Extraction of heavy metals from contaminated soils: I. sequential extraction in s urface soils and their relationships to DTPA extractable metals and metal plant uptake. J. Environ. Sci. Health Part A 35 1141 1152. Chacon, N., Silver, W.L., Dubinsky, E.A., Cusack, D.F., 2006. Iron reduction and soil phosphorus solubilization in humid t ropical forests soils: The roles of labile carbon pools and an electron shuttle compound. Biogeochemistry 78 67 84. Charlatchka, R., Cambier, P., 2000. Influence of Reducing Conditions on Solubility of Trace Metal s in Contaminated Soils. Water Air Soil Po llut 118 143 168. Chaturvedi, S., Dave, P.N., 2012. Removal of iron for safe drinking water. Desalination 303, 1 11. Chen, W., Westerhoff, P., Leenheer, J.A., Booksh, K., 2003. Fluorescence excitation Emission matrix regional integration to quantify sp ectra for dissolv ed organic matter. Environ. Sci. Technol. 37 5701 5710. Chipperfield, J.R., Ratledge, C., 2000. Salicylic acid is not a bacterial siderophore: a theoretical study. BioMetals 13 165 168. Choo, K.H., Lee, H., Choi, S.J., 2005. Iron and man ganese removal and membrane fouling during UF in conjunction with prechlorination for dr inking water treatment. J. Memb r. Sci. 267 18 26.
143 Christensen, T.H., Kjeldsen, P., Bjerg, P.L., Jensen, D.L., Christensen, J.B., Baun, A., Albrechtsen, H. J., Heron, G ., 2001. Biogeochemistry of l andfill leachate plumes. Appl. Geochem. 16 659 718. Chuan, M.C., Shu, G.Y., Liu, J.C., 1996. Solubility of heavy metals in a contaminated soil: Effects of redox potential and pH. Water Air Soil Pollut. 90 543 556. Clesceri, L ., Greenberg, A.E., Trussel, R.R., 1989. Standard Methods for the examination of Water and Wastewater, 17 ed. APHA, Washington, DC. Cline, J.D., 1969. Spectrometric determination of hydrogen sulfide in natural waters. Limnol. Oceanogr. 14 454 458. Comstoc k, S.E.H., Boyer, T.H., Graf, K.C., Townsend, T.G., 2010. Effect of landfill characteristics on leachate organic matter properties and coagulation treatability. Chemosphere 81 976 983. Cornell, R.M., Schwertmann, u., 2000. Iron Oxides in the Laboratory. J ohn Wiley & Sons Incorporated. Cornell, R.M., Schwertmann, U., 2003. The Iron Oxides: Structure, Properties, Reactions, Occurences and Uses, 2nd edition ed. Wiley VCH, Weinheim, Germany. Cozzarelli, I.M., Suflita, J.M., Ulrich, G.A., Harris, S.H., Scholl, M.A., Schlottmann, J.L., Christenson, S., 2000. Geochemical and Microbiological Methods for Evaluating Anaerobic Processes in an Aquifer Contaminated by L andfill Leachate. Environ. Sci. Technol. 34 4025 4033. Davranche, M., Bollinger, J. C., 2000. Heavy M etals Desorption from Synthesized and Natural Iron and Manganese Oxyhydroxides: Effect of R eductive Conditions. J. Colloid Interface Sci. 227 531 539. Davranche, M., Dia, A., Fakih, M., Nowack, B., Gruau, G., Ona nguema, G., Petitjean, P., Martin, S., Hoc hreutener, R., 2013. Organic matter control on the reactivity of Fe(III) oxyhydroxides and associated As in wetland soils: A kinetic modeling study. Chem. Geol. 335 24 35. Debadatta, D., Pramanik, K., 2013. A Study on Chemical Leaching of Iron from Red mu d using Sulp huric Acid. Research J. Chem. Environ. 17 50 56. del Castilho, P., Chardon, W.J., Salomons, W., 1993. Influence of Cattle Manure Slurry Application on the Solubility of Cadmium, Copper, and Zinc in a Manured Acidic, Loamy Sand Soil. J Environ Q ual 22 689 697. Dixon, J.B., Schulze, D.G., 2002. Soil Mineralogy with Environmental applications, 1 ed. Soil Science Society of America, Inc., Madison, Wisconsin.
144 Dixon, J.B., Weed, S.B., 1989. Minerals in Soil Environments, 2nd ed. Soil Science Society of America, Madison, Wisconsin. Dollhopf, M.E., Nealson, K.H., Simon, D.M., Luther Iii, G.W., 2000. Kinetics of Fe(III) and Mn(IV) reduction by the Black Sea strain of Shewanella putrefaciens using in situ solid state volta mmetric Au/Hg electrodes. Mar. C hem 70 171 180. Dos Santos Afonso, M., Stumm, W., 1992. Reductive dissolution of iron(III) (hydr)oxides by hydrogen sulfide. Langmuir 8 1671 1675. Dube, A., Zbytniewski, R., Kowalkowski, T., Cukrowka, E., Buszekwi, B., 2001. Adsorption and Migration of Heavy Metals in Soils. Pol. J. Environ. Stud. 10 1 10. Ellwood, B.B., Chrzanowski, T.H., Hrouda, F., Long, G.J., Buhl, M.L., 1988. Siderite formation in anoxic deep sea sediments: A synergetic bacteria controlled process with improtant implications in pal eomagnetism. Geology 16 980 982. Essington, M.E., 2003. Soil and Water Chemistry: An integrative approach. CRC Press LLC, Boca Ranton, Florida. F.A.C., 62 777. Contaminant Clean Up Target Levels. Feigl, V., Anton, A., Uzigner, N., Gruiz, K., 2012. Red Mud as a Chemical Stabilizer for Soil Contami nated with Toxic Metals. Water Air Soil Pollut. 223 1237 1247. Firer, D., Friedler, E., Lahav, O., 2008. Control of sulfide in sewer systems by dosage of iron salts: Comparison between theoretical and experimental results, and practical implications. Sci. Total Environ 392 145 156. Fredrickson, J.K., Zachara, J.M., Kennedy, D.W., Dong, H., Onstott, T.C., Hinman, N.W., Li, S., 1999. Geochim. Cosmochim. Acta 62 3239. Garrels, R.M., Christ, C.L., 1965. Solutions, M inerals and Equilibria. Freeman, Cooper and Co., San Francisco Geosyntec, 2005. Iron Assessment Report Escambia County, Florida. Florida Department of Environmental Protection Project: FL0716. Golden, D.C., Ming, D.W., Bowen, L.H., Morris, R.V., Lauer, H. V., 1994. Acidified oxalate and dithionite solubility and color of synthetic, partially oxidized Al magnetites and their therm al oxidation products. Clays Clay Miner 42 53 62. Guenther, E.A., Johnson, K.S., Coale, K.H., 2001. Direct ultraviolet spectroph otometric determination of total sulfide and iodide in natural waters. Anal Chem 73 3481 3487.
145 Gurzau, E.S., Neagu, C., Gurzau A.E., 2003. Essential metals Case study on iron. Ecotox. Environ. Safety 56 190 200. Haese, R.R., Wallmann, K., Dahmke, A., Kretzmann, U., Mller, P.J., Schulz, H.D., 1997. Iron species determination to investigate early diagenetic reactivity in marine sedim ents. Geochim. Cosmochim. Acta 61, 63 72. Harris, W.G., White, G.N., 2008. X ray diffraction techniques for soil mineral i dentification, in: Ulrey, A., Drees, R. (Eds.), Methods of Soil Analyis: Part 5 Minerological methods. Soil Sci. Soc. Am., Madison, WI, pp. 81 115. He, W., Shohag, M.J.I., Wei, Y., Feng, Y., Yang, X., 2013. Iron concentration, bioavailability, and nutritio nal quality of polished rice affected by different forms of foliar iron fertilizer. Food Chemistry 141 4122 4126. Heitmann, T., Blodau, C., 2006. Oxidation and incorporation of hydrogen sulfide by dissolved organic matter. Chem. Geol. 235 12 20. Heron, G ., Christensen, T.H., 1995. Impact of Sediment Bound Iron on Redox Buffering in a Landfill Leachate Polluted Aquifer (Vejen, Denmark). Environ. Sci.Technol. 29 187 192. Heron, G., Christensen, T.H., Tjell, J.C., 1994. Oxidation capacity of aquifer sedimen ts. Environ. Sci. Technol. 28 153 158. Jacobson, M.E., 1994. Chemical and Biological Mobilization of Fe(III) in Marsh Sediments. Biogeochemistry 25 41 60. Jain, P., Powell, J., Townsend, T., Reinhart, D., 2006. Estimating the Hydraulic Conductivity of La ndfilled Municipal Solid Waste Using the Bo rehole Permeameter Test. J. Environ. Eng. 132 645 652. Jang, Y.C., Townsend, T.G., 2003. Effect of Waste Depth on Leachate Quality from Laboratory Construction and Demolition Debris Landfills. Environ. Eng. Sci. 20 183 196. Johnson, D.B., Hallberg, K.B., 2005. Acid mine drainage remedi ation options: a review. Sci. Total Environ 338 3 14. Kabata Pendias, A., Pendias, H., 2000. Trace Elements in Soils and Plants, 3rd edition ed. CRC Press LLC, Boca Raton, Florida Kashem, M.A., Singh, B.R., 2001. Metal availability in contaminated soils: I. Effects of flooding and organic matter on changes in Eh, pH and solu bility of Cd, Ni andZn. Nutri. Cyc. Agroecosys. 61 247 255.
146 Khoshgoftarmanesh, A.H., Schulin, R., Chaney, R .L., Daneshbakhsh, B., Afyuni, M., 2010. Micronutrient efficient genotypes for crop yield and nutritional quality in sustainable agriculture. A review. Agron. for Sustain. 30 83 107. Kim, S.A., Guerinot, M.L., 2007. FEBS Letters 581, 2273 2280. Kjeldsen, P., Barlaz, M.A., Rooker, A.P., Baun, A., Ledin, A., Christensen, T.H., 2002. Present and Long Term Composition of MSW Landfill Leachat e: A Review. Crit. Review Envi ron. Sci. Technol. 32 297 336. Ko, M. S., Kim, J. Y., Lee, J. S., Ko, J. I., Kim, K. W., 2013. Arsenic immobilization in water and soil using acid mine drain age sludge. Appl. Geochem 35 1 6. Kraemer, S.M., 2001. Iron oxide dissolution and solubility in the p resence of siderophores. Aquat. Sci. 66 3 18. Lahav, O., Ritvo, G., Slijper, I., He arne, G., Cochva, M., 2004a. The potential of using iron oxide rich soils for minimizing the detrimental effects of H 2 S in freshwater aquaculture systems. J. Aquac. 238 263 281. Lahav, O., Ritvo, G., Slijper, I., Hearne, G., Cochva, M., 2004b. The potenti al of using iron oxide rich soils for minimizing the detrimental effects of H2S in freshwater aquaculture systems. J. Aquac 238 263 281. Larsen, O., Postma, D., 2001. Kinetics of reductive bulk dissolution of lepidocrocite, ferrihydr ite, and goethite. Ge ochim. Cosmochim. Acta 65 1367 1379. Larsen, O., Postma, D., Jakobsen, R., 2006. The reactivity of iron oxides towards reductive dissolution with ascorbic acid in a shallow sandy aquifer (Rm, Denmark). Geochimica et Cosmochimica Acta 70, 4827 4835. Lint on, T.K., Pacheco, M.A.W., McIntyre, D.O., Clement, W.H., Goodrich Mahoney, J., 2007. Development of bioassessment based b enchmarks for iron. Environ. Toxicol. Chem 26 1291 1298. Lovley, D.R., 1991. Dissimilatory Fe (III) and Mn (IV) reduction. Microbiol Review, 259 287. Lovley, D.R., 1997. Microbial Fe (III) reduction in subsurface envir onments. FEMS Microb. Rev 55 305 313. Lovley, D.R., Giovannoni, S.J., White, D.C., Champine, J.E., Phillips, E.J.P., Gorby, Y.A., Goodwin, S., 1993. Geobacter metallir educens gen nov sp nov, a microorganism capable of coupling the complete oxidation of organic compounds to the reduction of iron and other metals. Arch. Microbiol. 159 336 344.
147 Lovley, D.R., Phillips, E.J.P., 1988. Novel mode of microbial energy metabolis m organic carbon oxidation coupled to dissimilatory reduction of iron or manganese Appl. Environ. Microbiol. 54 1472 1480. Lovley, D.R., Phillips, E.J.P., Lonergan, D.J., 1991. Enzymic versus nonenzymic mechanisms for iron(III) reduction in aquatic sed iments. Environ. Sci. Technol. 25 1062 1067. Luu, Y.S., Ramsay, J.A., 2003. Review: microbial mechanisms of accessing insoluble Fe(III) as an energy source. World J. M icrob. Biot. 19 215 225. Lyngkilde, J., Christensen, T.H., 1992a. F ate of organic contaminants in the redox zones of a landfill leachate pollution plume (vejen, denmark). J. Contaminant Hydr 10 291 307. Lyngkilde, J., Christensen, T.H., 1992b. Redox zones of a landfill leachate pollution pl ume (Vejen, Denmark). J. Conta minant Hydr 10 273 289. Ma, L.Q., Tan, F., Harris, W.G., 1997. Concentrations and distributions of eleven meta ls in Florida soils. J. Environ. Qual 26 769 775. Makris, K.C., Harris, W.G., O'Connor, G.A., El Shall, H., 2005. Long term phosphorus effects on evolving physicochemical properties of iron and aluminum hydroxides. J. Colloid Interf. Sci. 287 552 560. Martnez Cuenca, M. R., Forner Giner, M.., Iglesias, D.J., Primo Millo, E., Legaz, F., 2013. Strategy I responses to Fe deficiency of two Citrus rootstocks differing in their tolerance to iron chlorosis. Scientia Horticulturae 153 56 63. McKeague, J.A., Day, J.H., 1966. Dithionite and oxalate extractable fe and al as aids in differentiating va rious classes of soils. Can J. Soi l Sci. 46 13 22. Mc Lean, E., Cogswell, M., Wojdyla, D., de Benoist, B., 2008. Worldwide prevalence of anaemia, WHO Vitamin and Nutrition Information System, 1993 2005. J. Public Health Nutri. 12 2008. McLean, J.E., Bledsoe, B.E., 1992. Behavior of Met als in Soils. EPA Groun d Water Issue EPA/540/S 92/018, 1 25. McLean, J.E., Dupont, R.R., Sorensen, D.L., 2006. Iron and arsenic release from aquifer solids in response to biostimulation. J Environ Qual 35 1193 1203. Mehra, O.P., Jackson, M.L., 1960. Iron oxide removal from soil s and clays by a dithionite citrate buffered w ith sodium bicarbonate. Clay Clay Miner. 7 317 327. Miller, T.L., Wolin, M.J., 1974. A Serum Bottle Modification of the Hungate Technique for Cultivating Obligate Anaerobes. Appl. Environ. Microbiol. 27 985 9 87.
148 Minyard, M.L., Burgos, W.D., 2007. Hydrologic Flow Controls on Biologic Iron(III) Reduction in Natural Sediments. Environ. Sci. Technol. 41 1218 1224. Muehe, E.M., Obst, M., Hitchcock, A., Tyliszczak, T., Behrens, S., Schrder, C., Byrne, J.M., Michel F.M., Krmer, U., Kappler, A., 2013. Fate of Cd during Microbial Fe(III) Mineral Reduction by a Novel and Cd Tolerant Geobacter Species. Environ. Sci. Technol. 47 14099 14109. Nevin, K.P., Lovley, D.R., 2002. Mechanisms for accessing insoluble Fe(III) o xide during dissimilatory Fe(III) reduction by G eothrix fermentans. Appl. Environ. Microbiol. 68 2294 2299. Nickson, R.T., McArthur, J.M., Ravenscroft, P., Burgess, W.G., Ahmed, K.M., 2000. Mechanism of arsenic release to groundwater, Bangladesh and W est Bengal. Appl. Geochem. 15 403 413. O'Brien, J., Wilson, I., Orton, T., Pognan, F., 2000. Investigation of the Alamar Blue (resazurin) fluorescent dye for the assessment of mammalian cell cytotoxicity. Eur. J. Biochem 267 5421 5426. Ochoa, H.G.G., B., To lson, J.K., Roberts, S.M., 2004. Draft Technical Report: Development of Cleanup Target Levels (CTLs) for Chapter 62 777, F.A.C., in: Protection), F.D.F.D.o.E. (Ed.). Parisio, S., Keimowitz, A.R., Simpson, H.J., Lent, A., Blackman, V., 2006. Arsenic Rich Ir on Floc Deposits in Seeps Downgradient of Solid Waste Landfills. Soil Sediment Contamin. 15 443 453. Pedersen, H.D., Postma, D., Jakobsen, R., Larsen, O., 2005. Fast transformation of iron oxyhydroxides by the catalytic action of aqueous Fe (II). Geochim. Cosmochim. Acta 69 3967 3977. Peiffer, S., Dos Santos Afonso, M., Wehrli, B., Gaechter, R., 1992. Kinetics and mechanism of the reaction of hydrogen sulfide with lepidocrocite. Environmental Science & Technology 26, 2408 2413. Peiffer, S., Gade, W., 2007. Reactivity of Ferric Oxides toward HS at low pH. Environ. Sci. Technol. 3159 3164. Plaza, C., Xu, Q., Townsend, T., Bitton, G., Booth, M., 2007. Evaluation of alternative landfill cover soils for attenuating hydrogen sulfide from construction and demolit ion (C& D) debris landfills. J. Environ. Manage 84 314 322. Ponnamperuma, F.N., 1972. The chemistry of submerged soils. Adv. Agron. 24 29 96. Poulton, S.W., 2003. Sulfide oxidation and iron dissolution kinetics during the reaction of dissolved sulfide wit h ferrihydrite. Chem. Geol. 202 79 94.
149 Poulton, S.W., Krom, M.D., Raiswell, R., 2004. A revised scheme for the reactivity of iron (oxyhydr)oxide minerals towards dissolved sul fide. Geochim. Cosmochim. Acta 68 3703 3715. Poulton, S.W., Krom, M.D., Rijn, J .V., Raiswell, R., 2002. The use of hydrous iron (III) oxides for the removal of hydrogen sulphide in aqueous systems. Water Res 36 825 834. Rickard, D., Morse, J.W., 2005. Acid volatile sulfide (AVS). Mar. Chem. 97 141 197. Roden, E., 2004. Analysis of long term bacterial vs. chemical Fe (III) oxide reduction kine tics. Geochim. Cosmochim. Acta 68 3205 3216. Roden, E., Soboloev, D., Glazer, B., Luther III, G.W., 2004. Potential for Microscale Bacterial Fe Redox Cycling at the Aerobic Anae robic Interface Geomicrobiol. J. 21 379 391 Roden, E.E., Urrutia, M.M., Mann, C.J., 2000. Bacterial Reductive Dissolution of Crystalline Fe(III) Oxide in Continuous Flow Column Reactors. Appl. Environ. Microbiol. 66 1062 1065. Roden, E.E., Zachara, J.M., 1996. Microb ial Reduction of Crystalline Iron(III) Oxides: Influence of Oxide Surface Area and Potential for Cell Growth. En viron. Sci. Technol. 30 1618 1628. Royer, R., Burgos, W., Fisher, A., Jeon, B., Unz, R., Dempsey, B., 2002a. Enhancement of hematite bior eduction by natural organic matter. En viron. Sci. Technol. 36 2897 2904. Royer, R.A., Burgos, W.D., Fisher, A.S., Jeon, B.H., Unz, R.F., Dempsey, B.A., 2002b. Enhancement of hematite bioreduction by natural organic matter. Environ Sci Technol 36 2897 290 4. Rugge, K., Bjerg, P.L., Christensen, T.H., 1995. Distribution of organic compounds from municipal solid waste in the groundwater downgradient of a landfill (grindsted, denmark). En viron, Sci. Technol. 29 1395 1400. Santisteban, J.I., Mediavilla, R., Lo pez M.J.G., Castano, S., Martinez Alfaro, P.E., 2004. Loss on ignition: a qualitative or quantitative method for organic matter and carbonate miner al content in sediments? J. Paleolimnol 32 287 299. Sastre, I., Vicente, M.A., Lobo, M.C., 2001. Behaviour of cadmium and nickel in a soil amended with sewage sludge Land Degrad. Dev. 12 27 33. Schwertmann, U., 1991. Solubility and dissolution of iron oxides. Plant and Soil 130, 1 25.
150 Schwertmann, U., Cornell, R .M., 2003. The Iron Oxides: Structure, Properties, Reactions, Occurences and Uses, 22 ed. Wiley VCH. Schwertmann, U., Friedl, J., Stanjek, H., 1999. From Fe(III) Ions to Ferrihydrite and then to Hematite. J. Colloid Interf Sci. 209, 215 223. Shaked, Y., Er el, Y., Sukenik, A., 2004. The biogeochemical cycle of iron and associat ed elements in Lake Kinneret. G eochimica et Cosmochimica Acta 68 1439 1451 Sidhu, P.S., Gilkes, R.J., Cornell, R.M., Posner, A.M., Quirk, J.P., 1981. Dissolution of Iron Oxides and Hy droxides in Hydrochloric and Perchloric Acid. Clays Clay Miner. 29 269 276. Silveira, M.L.A., Alleoni, L.R.F., Guilherme, L.R.G., 2003. Biosolids and Heavy Metals in Soils. Scientia Agricola 60, 793 806. Silver, J., 1993. Chemistry of Iron, 1 st ed. Blacki e Academic & Professional. Slowey, A.J., Brown, G.E., 2007. Transformations of mercury, iron, and sulfur during the reductive dissolution of iron oxyhydroxide by sul fide. Geochim. Cosmochim. Acta 71 877 894. Sorensen, J., 1982. Reduction of Ferric Iron in Anaerobic, Marine Sediment and Interaction with Reduction of Nitrate and Sulfate. Appl. Environ. Microbiol. 43 319 324. Soukup, D.A., Buck, B.J., Harris, W.G., 2008. Preparing soils for minerological analysis, in: Ulrey, A., Drees, R. (Eds.), Methods of Soil Analysis: Part 5 Minerological Methods. Soil Sci. Soc. Am., Madison, WI, pp. 13 31. Sracek, O., Bhattacharya, P., Jacks, G., Gustafsson, J. P., Brmssen, M.v., 2004. Behavior of arsenic and geochemical modeling of arsenic enrichment in aqueous en vir onments. Appl. Geochem 19 169 180. Stipp, S.L.S., Hansen, M., Kristensen, R., Hochella, M.F., Bennedsen, L., Dideriksen, K., Balic Zunic, T., Lonard, D., Mathieu, H.J., 2002. Behaviour of Fe oxides relevant to contaminant uptake in t he environment. Chem Geol. 190 321 337. Stookey, L.L., 1970. Anal. Chem. 42, 779. Straub, K.L., Benz, m., Schink, B., 2001. Iron metabolism in anoxic environments at near neutral pH. FEMS Microbiol. Ecol ., 181 186. Stucki, J.W., Schwertmann, U., 1988. Iron in Soils and Clay Minerals. D. Reidel, Holland.
151 Subcommittee on Iron, Assembly of Life Sciences, 1979. Iron. University Park Press, Baltimore, Maryland. Sulzberger, B., Suter, D., Siffert, C., Banwart, S., Stumm, W., 1989. Dissolution of fe(iii)(hydr)oxides in natural wate rs; laboratory assessment on the kinetics controlled by surfac e coordination. Mar. Chem 28 127 144. Summers, R., Smirk, D., Karafilis, D., 1996. Phosphorus retention and leachates from sandy soil amended with bauxite residue (red mud). Aust. J. Soil Res Suter, D., Banwart, S., Stumm, W., 1991. Dissolution of hydrous iron(III) oxides by reductive mechanisms. Langmuir 7 809 813. Thomine, S., Vert, G., 2013. Iron transport in plants: better be safe than sorry. Curr. Opin. Plant Biol. 16 322 327. Thompson, A., Chadick, O.A., Rancourt, D.G., Chorover, J., 2006. Iron oxide crystallinity increases during soil redox oscillat ions. Geochim. Cosmochim. Acta 70 1710 1727. Tipping, E., Rieuwerts, J., Pan, G., Ashmore, M.R., Lofts, S., Hill, M.T.R., Farago, M.E., Th ornton, I., 2003. The solid solution partitioning of heavy metals (Cu, Zn, Cd, Pb) in upland soils of England and Wales. Environ. Pollut 125 213 225. Tratnyek, P.G., Reilkoff, T.E., Lemon, A.W., Scherer, M.M., Balko, B.A., Feik, L.M., Henegar, B.D., 2001 Visualizing Redox Chemistry: Probing Environmental Oxidation Reduction Rea ctions with Indicator Dyes. Chem. Educator 6 172 179. Tuccillo, M.E., Cozzarelli, I.M., Herman, J.S., 1999. Iron reduction in the sediments of a hydrocarbon contaminat ed aquifer. Appl. Geochem 14 655 667. USDA, 1992. Soil Survey Laboratory Manual. Soil Survey Laboratory Manual SSIR No. 42. USDA, 2004. Soil Survey Laboratory Manual. Soil Survey Laboratory Manual. USEPA, 2004. Test methods for evaluating solid waste, physical chemi cal methods. SW 846. USEPA, 2013. Drinking Water Contaminants. Ussher, S.J., Achterberg, E.P., Worsfold, P.J., 2004. Marine B iogeochemistry of Iron. Environ. Chem 1 67 80. Vaughan, K., Rabenhorst, M.C., Needelman, B.A., 2009. Saturation and Temperature E ffects on the Development of Reducing Conditions in Soils. Soil Sci. Am. 73 663 667.
152 Veres, D.S., 2002. A comparative study between loss on ignition and total carbon analysis on minerogenic sediments. Studia Universitatis Babes Bolyai, Geologia, XLVII 1, 171 182. Viollier, E., Inglett, P.W., Hunter, K., Roychoudhury, A.N., Van Cappellen, P., 2000. The ferrozine method revisited: Fe(II)/Fe(III) determination in natu ral waters. Appl. Geochem 15 785 790. Vymazal, J., Krpfelov, L., 2009. Removal of organic s in constructed wetlands with horizontal sub surface flow: A review of the field experience. S ci. Total Environ. 407, 3911 3922. Wang, Y., Sikora, S., Kim, H., Dubey, B., Townsend, T., 2012. Mobilization of iron and arsenic from soil by construction and d emolition debris lan dfill leachate. Waste Manage 32 925 932. Wessling Resnick, M., 1999. Bio chemistry of Iron Uptake. Crit. Rev. Biochem. Mol. Biol. 34 285 314. Whiteley, J.D., Pearce, N.J.G., 2003. Metal distribution during diagenesis in the contaminate d sediments of Dulas Bay, Anglesey, N. Wales, UK. Appl. Geochem. 18 901 913. Wizniowski, J., Robert, D., Surmacz Gorska, J., Miksch, K., Weber, J.V., 2006. Landfill leachate treatment methods: A review. Environ. Chem. Lett. 4 51 61. Yadavalli, V., Neelam S., Rao, A.S.V.C., Reddy, A.R., Subramanyam, R., 2012. Differential degradation of photosystem I subunits under iro n deficiency in rice J. Plant Physiol. 169 753 759. Yao, H., Conrad, R., Wassmann, R., Neue, H.U., 1999. Effect of soil characteristics o n sequential reduction and methane production in sixteen rice paddy soils from China, the Philippines, and Italy. Biogeochemistry 47, 269 295. Yao, W., Millero, F.J., 1996. Oxidation of hydrogen sulfide by hydrous Fe(III) oxid es in seawater. Mar. Chem. 52, 1 16. Zachara, J.M., Fredrickson, J.K., Smith, S.C., Gassman, P.L., 2001. Solubilization of Fe(III) oxide bound trace metals by a dissimilatory Fe(III) reducing bacte rium. Geochim. Cosmochim. Acta 65 75 93.
153 BIOGRAPHICAL SKETCH Akua B onsu Oppong Anane w as born in Kumasi, Ghana in 1979. She spent the first four years of her elementary school education in Adelaide, South Australia before she returned with her family via a six month stay in the United Kingdom, to Ghana to comp lete her elementary and high s chool education. She was accepted into the Kwame Nkrumah University of Science and Technology degree in chemical e ngineering. She worked as an Assistant Research Scientist at the Water Research Institute of the Council for S cientific and Industrial Research (CSIR) before gaining admission to the Chemistry Department at University of Florida Gainesville where she graduated with a m Environmental Engineering Science Department in 2008 w here she completed her Ph D.