<%BANNER%>

Effects of Activated Carbon Surface Chemistry Modification on the Adsorption of Mercury from Aqueous Solution.

Permanent Link: http://ufdc.ufl.edu/UFE0044041/00001

Material Information

Title: Effects of Activated Carbon Surface Chemistry Modification on the Adsorption of Mercury from Aqueous Solution.
Physical Description: 1 online resource (129 p.)
Language: english
Creator: Faulconer, Emily Kaye
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2012

Subjects

Subjects / Keywords: carbon -- chemistry -- mercury -- surface
Environmental Engineering Sciences -- Dissertations, Academic -- UF
Genre: Environmental Engineering Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Mercury (Hg), a naturally occurring element, is toxic and can lead to negative health impacts for humans and ecosystems. Activated carbon adsorption is effective in treating Hg-laden effluent for safe discharge. Two modifications of commercially available activated carbon were investigated: iron impregnation to allow for magnetic sorbent recapture and wet chemical oxidation to enhance aqueous Hg capture. The modified carbons were characterized by nitrogen adsorption-desorption, XRD, pHpzc, vibrating sample magnetometry, elemental analysis, and total acidity titration. The 3:1 C:Fe magnetic carbon retained a high surface area of 790 m2/g and was 95% magnetically recoverable, with the iron present primarily as maghemite. The characteristics of the surface oxygen modified carbons varied based on the nature of the modifying reagent and its concentration. The modified carbons were applied to trace level Hg solutions (100 microgram/L). The 3:1 MPAC achieved the highest adsorption capacity, reaching 91% Hg removal with 2% volatilized and 84% adsorbed. Adsorption occurs primarily as chemisorption, thus allowing for non-hazardous residuals disposal until reaching a loading of greater than 800 microgram Hg/ g MPAC. Surface area and point of zero charge were identified as primary variables influencing adsorption in this system. Hg(II) adsorption was strongly correlated with oxygen content of the C(O)-modified activated carbons. Carbons with the highest oxygen content achieved the highest Hg(II) removal. Contrary to expectations, a strong correlation with oxygen content was not seen in Hg(0) adsorption. Rather, this data best fit a four variable model that identified surface area, pore volume, pHpzc, and oxygen content, with the pHpzc being the primary variable influencing results. At the experimental loading rate, no carbons leached Hg at levels requiring disposal as a hazardous waste. Kinetic models indicated both physisorption and chemisorption adsorption mechanisms. Hg speciation and binding mechanisms was predicted using sorbent and matrix characteristics. The use of sequential chemical extraction to verify these operational binding mechanisms was unsuccessful.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Emily Kaye Faulconer.
Thesis: Thesis (Ph.D.)--University of Florida, 2012.
Local: Adviser: Mazyck, David W.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2012
System ID: UFE0044041:00001

Permanent Link: http://ufdc.ufl.edu/UFE0044041/00001

Material Information

Title: Effects of Activated Carbon Surface Chemistry Modification on the Adsorption of Mercury from Aqueous Solution.
Physical Description: 1 online resource (129 p.)
Language: english
Creator: Faulconer, Emily Kaye
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2012

Subjects

Subjects / Keywords: carbon -- chemistry -- mercury -- surface
Environmental Engineering Sciences -- Dissertations, Academic -- UF
Genre: Environmental Engineering Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Mercury (Hg), a naturally occurring element, is toxic and can lead to negative health impacts for humans and ecosystems. Activated carbon adsorption is effective in treating Hg-laden effluent for safe discharge. Two modifications of commercially available activated carbon were investigated: iron impregnation to allow for magnetic sorbent recapture and wet chemical oxidation to enhance aqueous Hg capture. The modified carbons were characterized by nitrogen adsorption-desorption, XRD, pHpzc, vibrating sample magnetometry, elemental analysis, and total acidity titration. The 3:1 C:Fe magnetic carbon retained a high surface area of 790 m2/g and was 95% magnetically recoverable, with the iron present primarily as maghemite. The characteristics of the surface oxygen modified carbons varied based on the nature of the modifying reagent and its concentration. The modified carbons were applied to trace level Hg solutions (100 microgram/L). The 3:1 MPAC achieved the highest adsorption capacity, reaching 91% Hg removal with 2% volatilized and 84% adsorbed. Adsorption occurs primarily as chemisorption, thus allowing for non-hazardous residuals disposal until reaching a loading of greater than 800 microgram Hg/ g MPAC. Surface area and point of zero charge were identified as primary variables influencing adsorption in this system. Hg(II) adsorption was strongly correlated with oxygen content of the C(O)-modified activated carbons. Carbons with the highest oxygen content achieved the highest Hg(II) removal. Contrary to expectations, a strong correlation with oxygen content was not seen in Hg(0) adsorption. Rather, this data best fit a four variable model that identified surface area, pore volume, pHpzc, and oxygen content, with the pHpzc being the primary variable influencing results. At the experimental loading rate, no carbons leached Hg at levels requiring disposal as a hazardous waste. Kinetic models indicated both physisorption and chemisorption adsorption mechanisms. Hg speciation and binding mechanisms was predicted using sorbent and matrix characteristics. The use of sequential chemical extraction to verify these operational binding mechanisms was unsuccessful.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Emily Kaye Faulconer.
Thesis: Thesis (Ph.D.)--University of Florida, 2012.
Local: Adviser: Mazyck, David W.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2012
System ID: UFE0044041:00001


This item has the following downloads:


Full Text

PAGE 1

1 EFFECTS OF ACTIVATED CARBON SURFACE CHEMISTRY MODIFICATION ON TH E ADSORPTION OF MERCURY FROM AQ U E OUS SOLUTION By EMILY KAYE FAULCONER A DISSERTATION PRESENTED 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 2012

PAGE 2

2 2012 Emily Kaye Faulconer

PAGE 3

3 To my mother and father

PAGE 4

4 ACKNOWLEDGMENTS I am thankful to my twin sister, Laura Faulconer, for blazing the trail to a Ph.D. and offering support and advice from her journey. I deeply respect her a chievements both academically and professionally. For my mother and father, with their unwavering support and unconditional love, I feel overwhelmed with thankfulness I recognize my fortune in having such dedicated, supportive, and loving parents. I would like to thank my graduate advisor, David Mazyck, for his mentorship and for pushing me to perform my best. I am also grateful for the guidance provided by my advis ory committee, Dr. Treavor Boyer, Dr. Paul Chadik, and Dr. Lena Ma. I am grateful for the support of former and present members of my research g roup : Amy Borello, Heather Byrne, Timothy English, Alec Gruss, Ana Maria Hagan, Sanaa Jaman, and Taccara Willi ams. My research was also greatly benefited by interactions with David Baum and Rick Loftis of Environmental Performance Solutions. I would like to extend warm appreciation for the laboratory assistance of Ross Beardsley, Kenneth Sherman, and particularly Natalia Hoogesteijn. I am excited to see her do wonderful things in her pursuit of a graduate degree. Without the constant support of my friends, I would never have persevered to achieve such challenging goals. I express sincere appreciation for the fri endship of Katrina Indarawis Victoria Mc C loud, Emi Lenes, Heather Byrne, Audra Kelly, Regena Hudson, Judy Walden, and Clinton Williams Their words of encouragement and emotional support when I encountered seemingly impossible challenges were invaluable In their own way, each show ed me my strength. Those close to me know that I do n o t shy away from a challenge but I cannot imagine achieving my goals without these individuals that provided support, love, and energy.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ 4 LIST OF TABLES ................................ ................................ ................................ ........... 8 LIST OF FIGURES ................................ ................................ ................................ ........ 9 ABSTRACT ................................ ................................ ................................ .................. 11 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ ... 13 Problem Statement ................................ ................................ ................................ 13 Hy potheses ................................ ................................ ................................ ........... 14 Objectives ................................ ................................ ................................ .............. 14 2 LITERATURE REVIEW ................................ ................................ ......................... 16 Mercury ................................ ................................ ................................ ................. 16 History ................................ ................................ ................................ ............. 16 Mercury Chemistry ................................ ................................ .......................... 16 Physical and chemical properties ................................ .............................. 16 Mercury speciation ................................ ................................ .................... 17 Mobility and solubility of Hg complexes ................................ ..................... 21 Health I mpacts ................................ ................................ ................................ 22 Human health impacts ................................ ................................ .............. 22 Environmental health impacts ................................ ................................ ... 24 M ercury Emissions ................................ ................................ .......................... 24 Chlor alkali industry ................................ ................................ .................. 25 Flue gas desulphurization ................................ ................................ ......... 26 Mercury emission regulations ................................ ................................ ... 27 Aqueous mercury removal technologies ................................ ................... 28 Activated Carbon ................................ ................................ ................................ ... 29 Synthesis of Activated Carbon ................................ ................................ ........ 29 Thermal activation ................................ ................................ .................... 3 0 Chemical activation ................................ ................................ ................... 31 Activated Carbon Modification ................................ ................................ ......... 31 Enhanced surface oxygen functionality ................................ ..................... 31 Iron impregna tion ................................ ................................ ...................... 32 Adsorption ................................ ................................ ................................ ....... 34 Adsorption theory ................................ ................................ ...................... 34 Isotherm theory ................................ ................................ ......................... 36 Aqueous phase metal adsorption ................................ .............................. 38 Mercury adsorption from aqueous solution ................................ ............... 39

PAGE 6

6 3 MATERIALS AND METHODS ................................ ................................ ............... 51 Chemicals and Materials ................................ ................................ ....................... 51 Materials Synthesis ................................ ................................ ............................... 51 Iron Impregnation ................................ ................................ ............................ 51 Surface Oxygen Modification ................................ ................................ ........... 53 Activated Carbon Characterization Methods ................................ .......................... 54 Porosity ................................ ................................ ................................ ........... 54 Instrumentation ................................ ................................ ......................... 54 Surface area ................................ ................................ ............................. 54 Pore volume ................................ ................................ ............................. 55 Pore size ................................ ................................ ................................ ... 56 Point of Zero Charge ................................ ................................ ....................... 56 Total Acidity Titration ................................ ................................ ....................... 57 Elemental Analysis ................................ ................................ .......................... 58 X Ray Diffraction ................................ ................................ ............................. 58 Vibrating Sample Magnetometry ................................ ................................ ..... 59 Magnetic Adsorbent Recovery ................................ ................................ ........ 59 Adsorbent Stability ................................ ................................ .......................... 60 Aqueous Mercury Removal ................................ ................................ ................... 61 Labware Preparation ................................ ................................ ....................... 61 Mercury Quantification Methods ................................ ................................ ...... 61 Test Stand ................................ ................................ ................................ ....... 62 Hg Mass Balance ................................ ................................ ............................ 63 Batch Studies ................................ ................................ ................................ .. 63 Investigation of Adsorption Mechanisms ................................ ......................... 64 Influence of pH and pCl ................................ ................................ ............ 64 Sequential chemical extract ion ................................ ................................ 64 Data Analysis ................................ ................................ ................................ ........ 66 4 CHARACTERIZATION OF MODIFIED ACTIVATED CARBON ............................. 71 MPAC Characterization ................................ ................................ ......................... 71 Porosity ................................ ................................ ................................ ........... 71 Magnetic Characteristics ................................ ................................ ................. 72 X ray diffraction ................................ ................................ ......................... 72 Vibrating sample magnetometry ................................ ............................... 74 Magnetic adsorbent recovery ................................ ................................ .... 75 Adsorbent Stability: Iron ................................ ................................ .................. 75 C(O) Modified Carbon Characterization ................................ ................................ 76 Porosity ................................ ................................ ................................ ........... 76 Surface Oxygen Functionality ................................ ................................ .......... 76 5 TRACE LEVEL AQUEOUS MERCURY REMOVAL USING MODIFIED ACTIVATED CARBON ................................ ................................ .......................... 84 MPAC Results ................................ ................................ ................................ ....... 84

PAGE 7

7 Controls ................................ ................................ ................................ ........... 84 Pseudo Equilibrium Adsorption ................................ ................................ ....... 85 Contact time ................................ ................................ ............................. 85 Batch testing of synthetic waters ................................ ............................... 85 Adsorption Isotherms ................................ ................................ ...................... 87 Kinetics Studies ................................ ................................ ............................... 88 Adsorbent Stability: Hg ................................ ................................ .................... 89 C(O) Results ................................ ................................ ................................ .......... 89 Controls ................................ ................................ ................................ ........... 89 Batch Testing of Synthetic Waters ................................ ................................ .. 90 Effect of C(O) on Hg adsorption ................................ ................................ 90 Effect of porosity on Hg adsorption ................................ ........................... 91 Adsorption Isotherms ................................ ................................ ...................... 92 Kinetic Studies ................................ ................................ ................................ 92 Adsorbent Stability ................................ ................................ .......................... 93 6 ADSORPTION MECHANISMS ................................ ................................ ............ 104 Proposed Adsorption Mechanisms ................................ ................................ ...... 104 Mechanisms of Hg(0) Adsorption ................................ ................................ .. 105 Influence of pH and pCl on Hg(II) adsorption ................................ ....................... 105 Sequential Chemical Extraction ................................ ................................ ........... 106 Protocol Verification ................................ ................................ ...................... 106 Application ................................ ................................ ................................ .... 107 7 CONCLUSIONS AND RECOMMENDATIONS ................................ .................... 111 Magnetic Powdered Activated Carbon ................................ ................................ 111 Surface Oxyg en Modified Carbon ................................ ................................ ........ 112 Contributions to Science ................................ ................................ ...................... 113 Future Recommendations ................................ ................................ ................... 114 APPENDIX A MODIFICATION OF SURFACE OXYGEN FUNCTIONALITY OF BIOCHAR FOR HG ADSORPTION ................................ ................................ ...................... 115 LIST OF REFERENCES ................................ ................................ ............................ 118 BIOGR APHICAL SKETCH ................................ ................................ ......................... 129

PAGE 8

8 LIST OF TABLES Table page 2 1 Stability constant values for Hg OH and Hg Cl compounds ............................... 43 2 2 Select anthropogenic sources of Hg ................................ ................................ .. 44 2 3 Reported ranges of chlor alkali wastewater constituents ................................ ... 44 2 4 Reported ranges of FGD wastewater constituents ................................ ............ 45 3 1 Surface Area Calculation Methods by P/P 0 range u tilized ................................ .. 68 4 1 Porosity of v arious MPACs ................................ ................................ ................ 78 4 2 Magnetic characteristics of various MPACs ................................ ....................... 78 4 3 Magnetic solid phase extraction results for various MPACs ............................... 78 4 4 Characterization of various C(O) modified carbons ................................ ........... 79 5 1 Hg leaching from various carbons under landfill conditions ............................... 94 6 1 Variation of Hg DI contact pH with pH pzc of C(O) modified carbons ................. 108 6 2 Predicted Hg speciation and SCE extraction fraction for giv en pH and pCl values ................................ ................................ ................................ .............. 109 6 3 Hg distribution in SCE extraction fractions ................................ ....................... 109 A 1 Biochar characterization data ................................ ................................ .......... 116 A 2 Adsorption of aqueous Hg(II) by raw and modified biochar ............................. 117

PAGE 9

9 LIST OF FIGURES Figure page 2 1 Hyd ration of Hg 2+ ion in water ................................ ................................ ............ 46 2 2 3 dimensional geometry of Hg 2+ hydration ................................ ........................ 46 2 3 Distribution o f Hg(II) at different pH values ................................ ........................ 46 2 4 Mercury Eh pH diagram for Hg O H S Cl syst em ................................ .............. 47 2 5 Distribution of Hg(II) at various chloride concentration s ................................ ..... 47 2 6 Hg(II) Speciation at varying pH and chlorid e concentrations .............................. 48 2 7 IUPAC gas adsorpti on isotherm classifications ................................ .................. 49 2 8 Nitrogen adsorption isotherm on micro and mesoporous carbon exhibiting a closed h ysteresis loop ................................ ................................ ....................... 49 2 9 Types of hysteresis loops observed during a dsorption ................................ ..... 50 3 1 Common acidic surface oxygen groups on activated carbon with pH above the pHpzc ................................ ................................ ................................ .......... 69 3 2 Vibrating Sample Magnetom eter Schematic ................................ ...................... 69 3 3 Hysteresis loop resulting from VSM ana lysis ................................ ..................... 69 3 4 Cold Vapor Atomic Absorption Spectroscopy Schematic ................................ ... 70 3 5 Schematic of batch adsorption test stand ................................ .......................... 70 4 1 MPAC nitrogen adsorption deso rption isotherms ................................ .............. 80 4 2 Effect of thermal oxidation on porosit y of 1:1 C:Fe ................................ ............ 81 4 3 BJH pore size distribution of select MPACs ................................ ....................... 82 4 4 Powder XRD patterns of MPAC particles before and after thermal oxidation ..... 82 4 5 BJH pore size distribution of select C(O) modified carbons ............................... 83 5 1 Background Hg mass balance ................................ ................................ ........... 95 5 2 Effect of contact time on Hg (II) adsorption ................................ ......................... 95 5 3 Effect of iron loading on pseudo equilibrium adsorption of 100 g/L Hg(II) ......... 96

PAGE 10

10 5 4 Influence of 3h oxidation at 250 C and 450 C on aqueous Hg(II) removal ........ 9 6 5 5 Mass balance distribution ................................ ................................ .................. 97 5 6 Hg mass balance for 3:1 C:Fe adsorbent ................................ .......................... 97 5 7 Hg(II) adsorption isotherm onto 3:1 MPAC ................................ ........................ 98 5 8 Kinetic models for the adsorption of Hg(II) onto 3:1 MPAC ................................ 99 5 9 Hg leaching from 3:1 C:Fe at various loading rates under landfill conditions ... 100 5 10 Background Hg(0) mass balance for a 30 s contact time ................................ 100 5 11 Hg removal through adsorption and volatilization for various surface modified carbo ns ................................ ................................ ............................. 101 5 12 Hg(II) adsorption isotherm onto NAC 1M ................................ ......................... 102 5 13 Kinetic models for the adsorption of Hg(II) onto C(O) modified carbons .......... 103 6 1 Influence of pH on aqueous Hg(II) adsorption ................................ ................. 110 6 2 Influence of pCl on Hg(II) volatilization ................................ ........................... 110

PAGE 11

11 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 EFFECTS OF ACTIVATED CARBON SURFACE CHEMISTRY MODIFICATIO N ON THE ADSORPTION OF MERCURY FROM AQUEOUS SOLUTION By Emily Kaye Faulconer May 2012 Chair: David Mazyck Major: Environmental Engineering Sciences Mercury (Hg) a naturally occurring element, is toxic and can lead to negative h ealth impacts for humans and ecosystem s Activated carbon adsorption is effective in treating Hg laden aqueous effluent for safe discharge. Two modifications of commercially available activated carbon were investigated: iron impregnation to allow for magnetic sorbent recapture and wet chemical oxidation to enhance aqueous Hg capture. The modified carbons were characterized by nitrogen adsorption desorption, XRD, pH pzc vibrating sample magnetometry, elemental analysis, and total acidity titration. The 3:1 C:Fe magnetic powdered act ivated carbon (MPAC) retained a high surface area of 790 m 2 /g and was 95% magnetically recoverable, with the iron present primarily as maghemite. The characteristics of the surface oxygen modified carbons varied based on the nature of the modifying reagent and its concentration. The modified carbons were applied to trace level Hg solutions (100 g/L). The 3:1 MPAC achieved the highest adsorption capacity reaching 91% Hg removal with 2% volatilized and 84% adsorbed. Adsorption occurs primarily as chemisorp tion, thus allowing for non hazardous residuals disposal until reaching a loading of greater than

PAGE 12

12 800 g Hg/ g MPAC. Surface area and point of zero charge were identified as primary variables influencing adsorption in this system. Hg(II) adsorption was st rongly correlated with oxygen content of the C(O) modified activated carbons Carbons with the highest oxygen content achieved the highest Hg(II) removal Contrary to expectations, a strong correlation with oxygen content was not seen in Hg(0) adsorption. Rather, th ese data best fit a four variable model that identified surface area, pore volume, pH pzc and oxygen content, with the pH pzc being the primary variable influencing results. Using the standardized EPA TCLP protocol, it was found that no carbons le ached Hg at levels requiring disposal as a hazardous waste a t the experimental loading rate Kinetic models indicated both physisorption and chemisorption adsorption mechanisms. Hg speciation and binding mechanisms was predicted using sorbent and matrix c haracteristics. The use of sequential chemical extraction to verify these operational binding mechanisms was unsuccessful due to extraction inefficiencies and phase transformation

PAGE 13

13 CHAPTER 1 INTRODUCTION Problem Statement Mercury (Hg) is a naturally occu rring elemen t found in air, water, and soil. The U.S. EPA lists Hg and Hg compounds as toxic pollutants under section 307(a) of the Clean Water Act. In aquatic ecosystems, inorganic mercury undergoes chemical and microbial transformation to methylmercury. Methylmercury is a serious environmental concern due to its high toxicity and ability to bioaccumulate and biomagnify [1] Hg enters the environment from sources such as volcanoes or anthropo genically from sources such as the chlor alkali industry, coal fired power plants, battery manufacturing, metal mining, and the pharmaceutical industry. The toxic nature of Hg ury into Minamata Bay resulted in Hg poisoning of the local population through consumption of contaminated fish and shellfish. Industrial Hg release continues today. The Toxics Release Inventory stated that the total disposal or release of Hg in the United States increased by 1.9 million pounds from 2006 to 2007 a 38% increase [1] Current Hg discharge limits for industrial effluent vary by region [2] As Hg regulations become increasingly strict, new effluent control technologies will be required to treat trace levels of aqueous Hg. The traditional technologies for aqueous Hg treatment, including precipitation and adsorption, have struggled to treat to ng/L effluent levels that are required to ensure the health of the environment and humans. Any residual Hg that remains in the wastewater upon discharge can persist in its dissolved or particulate form and may undergo transformation t o methylmercury [3]

PAGE 14

14 Thus, it is vital to control Hg discharges wherever possible in order to protect the health of humans and the environment. Activated carbon, a high surface area sorbent, has been used for many app lications in aqueous treatment. Recent research has focused on enhancing the effectiveness of activated carbon by modifying specific properties, chemically and physically. It is possible to tailor the surface chemistry of activated carbon to increase adsor ption capacity and selectivity for Hg. Modification by iron impregnation can provide the carbon with magnetic properties, allowing for magnetic capture and thus easier residuals disposal. This study focuses on understanding the surface chemistry reactions between aqueous Hg and activated carbon in order to develop a sorbent that can b e applied to water with varying characteristics, is recoverable from aqueous solution and can treat trace levels of Hg. Hypotheses 1. T he impreg nation of activated carbon with ferrimagnetic iron oxides (magnetite and maghemite) would allow for magnetic separation and thus more responsible residuals disposal 2. Iron impregnation would not significantly impact the adsorption capacity of the composite sorbent. 3. Matrix characteristic s such as pH and pCl would influence Hg speciation and thus adsorption mechanisms. 4. Wet chemical oxidation of activated carbon would increase surface oxygen functionality ; i ncreased surface oxygen functionality would increase Hg adsorption capacity. Object ives 1. Synthesize magnetic carbons that are at least 95% recoverable through magnetic separation. 2. Increase acidic C( O) on activated carbon surfaces with minimal pore degradation

PAGE 15

15 3. Characterize carbons with various techniques including nitrogen adsorption des orption, point of zero charge, and total acidity. 4. Determine which experimental conditions yield the highest removal of aqueous Hg 5. Predict the influence of matrix pH and pCl on Hg speciation ; propose Hg adsorption mechanisms.

PAGE 16

16 CHAPTER 2 LITERATURE R EVI EW Mercury History Mercury (Hg) has been used by humans throughout history for various purposes including mirror production and medicines, despite awareness of i t s poisonous properties hat appeared among workers in Idrija, Slowenia [3] As mercury toxicity became better understood, its use in dental amalgams and pharmaceuticals diminished, with a few exceptions. Even with the current understanding o f mercury toxicity, some cultures continue to use mercury for rituals as well as cosmetic and pharmaceutical purposes. A rtisanal gold mining which often uses elemental Hg Au amalgamation for gold recovery, has been increasing over the past few decades le ading to a resurgence in mercury use Mercury Chemistry Physical and chemical properties The heavy metal mercury has an atomic number of 80, an atomic mass of 200.59, and a density of 13.55 g/cm 3 Mercury has an electron configuration of [Xe]5s 2 p 6 d 10 6s 2 w ith the highest energy electron occupying a d orbital. With a melting point of 39.8 C, Mercury is the only metal that is a liquid at standard temperature and pressure (STP). Mercury has three oxidation states, Hg(0) (elemental Hg), Hg(I) (mercurous Hg), a nd Hg(II) ( mercuric Hg).

PAGE 17

17 Mercury speciation Three broad categories of Hg speciation are elemental (Hg(0)), inorganic (Hg(I) and Hg(II)), and organic mercury. These chemical forms impact its solubility and reactivity as well as its mobility, bioavailabilit y, toxicity, bioaccumulation, and biomagnification [4] Elemental mercury has a high vapor pressure (14 mg/m 3 at 20C) Inorganic mercury occurs as Hg(I) and Hg(II) salts. Many Hg(II) salts are readily soluble in water and thus are highly mobile and toxic. A notable exception is HgS which has a solubility of ~10 ng/L. Inorganic mercury has a high affinity for selenium, which can explain the protective role it plays in mercury toxicity. Inorganic mercury also has a high affinity for sulfur, including amino acids such as cysteine and methionine, which explains its high toxicity. Hg(I) is less stable than Hg(II) and is only sparingly soluble, resulting in lower toxicity. Organic mercury consists of a covalent bond between a divalent Hg atom and carbon. These compounds can react with biologically important ligands and can easily cross biological membranes. Mercury can cycle between the atmosphere (air), hydrosphere (water), and lithosphere (land) as well as transfer through the food chain. The most common forms of Hg found in the environment are metallic Hg, mercuric sulf ide, mercuric chloride, and methylmercury. The main dissolved Hg species in aquatic environments are Hg(0), Hg(II) complexes, and organic Hg forms primarily as monomethylmercury cation and dimethylmercury [3] F or the purposes of this work, the focus will be on aqueous Hg chemistry, excluding methylmercury. Formation of hydration spheres. When an Hg 2+ ion is placed in water, the hydrogen bonding network of the water is altered as the water molecules rotate so tha t

PAGE 18

18 their negative dipoles face the opposite charge of the Hg ion, thus breaking hydrogen bonds. This group of water molecules is called a hydration shell. The new orientation results in a net charge of the same sign as the ion on the outside of this hydrati on shell (Figure 2 1). This charge then tends to orient nearby water molecules, causing a second hydration shell and resulting in further disruption of the hydrogen bonding network. ole to the H electron cloud, weakening the bond and allowing for easier dissociation of the water molecule. This phenomenon results in the metal ion acting as a polyprotic acid as the complexed water deprot onates [5] Mercury complexation with H 2 O. Without complexing ligands present, hydrolysis plays a large role in speciation. At a low pH (below pH 2), the hexaqua ion, Hg(H 2 O) 6 2+ is octahedrally coordinated by water m olecules with equal Hg O bond lengths (Figure 2 2). As the pH increase s the octahedral coordination is distorted This results in two axial oxygen atoms with a shortened Hg O bond length and four equatorial oxygen atoms with lengthened Hg O bond len gths [6] Up to two protons can be released from the waters of hydration surrounding the Hg 2 + ion (Equations 2 1 to 2 2 [7] ) as Hg 2+ hydrolyzes to HgOH + and Hg(OH) 2 (Figure 2 3). In the absence of complexing ligands, Hg(OH) 2 is the dominant inorganic species at pH 6 [8] Hg 2+ + H 2 O HgOH + + H + K 1 = 10 3.4 = {H + }{HgOH + }/{Hg 2+ } (2 1 ) HgOH + + H 2 O Hg(OH) 2 + H + K 2 = 10 2.7 = {H + }{Hg(OH) 2 }/{HgOH + } (2 2 ) Hg 2+ + 2H 2 O Hg(OH) 2 + 2H + K overall = 10 6.1 = {H + } 2 {Hg(OH) 2 }/{HgOH + } (2 3 )

PAGE 19

19 Mercury Complexation with Ligands. Associati on with various ligands is strongly dependent upon environmental conditions including the type and concentration of Lewis bases present, the redox status (pE), Eh, pH, and pCl (Figure 2 4) [9,10] Th e Hg ion can react with a ligand through inner or outer sphere complexation. Inner sphere complexation (e.g. ion exchange) involves the exchange of a hydration water for the ligand (Equations 2 4 and 2 5 where L = ligand ) [10] The loss of the water molecule from the hydration sphere is often the rate determining step. Outer sphere complexation (e.g. hydrogen bonding) is an electron transfer that involves separate chemical components that remain separate during the ent ire electron transfer event, as opposed to inner sphere electron transfer, in which the two chemical components are connected via a chemical br idge [10,11] Ligands alter the adsorption of met al cations in the following ways: the formation of stable non adsorbing complexes the formation of ternary surface complexes, competitive adsorption of ligands onto the ad sorbent surface and reduction of the positive charge at the adsorbent surface throu gh adsorption of the ligand [12] Hg(H 2 O) 6 2+ + L Hg(H 2 O) 6 L + (2 4 ) Hg(H 2 O) 6 L + Hg(H 2 O) 5 L + + H 2 O (2 5 ) Mercury complexation with chloride. In aqueous solution, Hg can complex with chloride ligands to form very stable Hg Cl complexes even at very low chloride concentrations (Figure 2 5). U p to four water molecules from the hydration sphere can be exchanged for chloride ions, depending upon the chloride concentration (Equations 2 6 to 2 1 0 [7] ). The mass balance for a system containing Cl and OH as ligands is represented in Equation 2 1 1

PAGE 20

20 Hg 2+ + Cl HgCl + K 1 = {HgCl + }/{Hg 2+ }{Cl } (2 6 ) HgCl + + Cl HgCl 2 K 2 = {HgCl 2 }/{HgCl + }{Cl } (2 7 ) HgCl 2 + Cl HgCl 3 K 3 = {HgCl 3 }/{HgCl 2 }{Cl } (2 8 ) HgCl 3 + Cl HgCl 4 2 K 4 = {HgCl 4 2 }/{HgCl 3 }{Cl } (2 9 ) Hg 2+ + 4Cl HgCl 4 2 4 = {HgCl 4 2 }/{Hg 2+ }{Cl } 4 (2 1 0 ) (2 1 1 ) A overall n equilibrium constant that describes a ligand displacement equilibrium reaction The constant is derived by fitting experimental data into a chemical model of the equilibrium system so values are found to vary with t he source of the data (Table 2 1 ). As s een in equation 2 12, the differences in stability constant values can impact the predicted speciation. A large stability constant denotes a strong tendency to form a complex. Thus, based on the log K values given by Benjamin, the Hg Cl species are more li kely to form than the Hg OH species [7] Hahne and Kroontje [13] performed a thorough examination of the effect of chloride concentrations on Hg speciation Using the stabili ty cons tants provided by Benjamin [7] using concentrations rather than activities, and verified by Visual MINTEQ, the following conclusions have been drawn. At pH 2, chloride levels of just 3.5 7) result in the shift of Hg speciation from 50% Hg 2+ and 50% HgOH + to include approximately 25% of the total Hg as Hg Cl complexes HgCl + and HgCl 2 (Figure 2 6). When chloride concentrations reach 500 mg/L (pCl 1.85), Hg is present entirely as Hg Cl complexes with 85% as HgCl 2 Increasing the chloride concentration to levels

PAGE 21

21 commonly found in chlor alkali wastewater (25,000 mg/L) further alters the speciation to primarily HgCl 4 2 [13] The major difference between Hg speciation at pH 2 and pH 4 is present at 3.5 Cl At pH 2, Hg exists as 25% Hg Cl complexes while at pH 4 all Hg is hydrolyzed as mono and dihydroxy species and no chloro complexes are present. At the other chloride concentrations investigated, the speciation did not differ much between the t wo pH values [13] At pH 6, 100% of the Hg at pCl 12 and 7 exists as fully hydrolyzed Hg(OH) 2 The speciation of Hg at hi gher chloride concentrations remains similar to the distribution at the more acidic pH values o f 2 and 4 [13] Increasing to an alkaline pH value of 8, the fully hy drolyzed Hg species is dominant, constituting 100% of the total Hg for pCl 12 and pCl 7 At pCl 1.85, Hg(OH) 2 accounts for 70% of the Hg. At this p oint, HgCl 2 accounts for 28% of total Hg as opposed to the 85 89% at pH 2, 4, and 6 [13] Mercury complexation with sulfur. Mercury is sulfophilic, with a strong affinity for ligands containing sulfur [10] Mercuric sulfide, HgS, is one of the least soluble salts known and readily precipitates from aqueous solution In the presence of chloride ions and oxidizing conditions, Hg Cl complexes will predominate while reducing condit ions allow for Hg S complexes to predominate. More soluble than mercuric sulfide, HgS 2 2 forms at high pH and Hg(SH) 2 forms at low pH. Mercuric complexation with nitrate. Hg(NO 3 ) 2 completely ionizes in solution to form Hg 2+ and 2 NO 3 .In this system, no co mplexation occurs beyond Hg hydrolysis. This reaction is relevant due to the laboratory use of Hg(NO 3 ) 2 standards. Mobility and solubility of Hg complexes Speciation can determine the solubility and mobility of Hg in the environment with the degree of m obilization depending upon the degree of complexation. Hg Cl

PAGE 22

22 complexation increases solubility while Hg S complexation decreases solubility, with K s 0 values of 2.59 x 10 15 and 2 x 10 53 respectively Without chloride ions present, the mobility of Hg is re stricted both due to the solubility of Hg(OH) 2 and the potential for adsorption of Hg 2+ and HgOH + Hg(OH) 2 is soluble up to 107 mg/L (5.37 x 10 4 M) at which point precipitation will take place. But with just 0.35 g/L chloride at pH 6, most of the Hg will be present as Hg Cl complexes, which are highly soluble [8] Previous researchers have determined that the introduction of chloride ions to solution can release Hg from sedi ments into solution [14] As stated earlier regarding the stability constants the source of the equilibrium values can impact the predicted total soluble Hg concentration present at a given pH in a system in equilibriu m with Hg(OH) 2(s) OH and Cl ( Equation 2 12 ). (2 1 2 ) Health Impacts Human health impacts Toxicokinetics and toxicodynamics The chemical speciation of Hg influences its toxicokinetics (absorption, distribution, metabolism, and excretion) [15] Elemental Hg exposure occurs primarily through inhalation as it is rapidly absorbed through the lungs with a pproximately 80% of inhaled vapors absorbed by lung tissues [15] Once absorbed, elemental Hg can penetrate both the placental and the blood brain barrier to act as a neurotoxicant [4] Elemental Hg is elimin ated through urine, feces, exhalation, sweat, and saliva, dependent upon the extent of oxidation. Symptoms of elemental Hg exposure include tremors, lethargy, insomnia, memory loss, cognitive impairment, and headaches as well as kidney, pulmonary, and thyr oid effects [16]

PAGE 23

23 Absorption Hg(I) and Hg(II) occurs primarily through the gastrointestinal tract; therefore, most exposure occurs through diet. Even soluble mercury salts are not well absorbed, with uptake ranging between 7 15% [3] Because inorganic Hg is not lipid solu ble, it has very limited ability to cross both the blood brain and placental barriers. Symptoms of inorg anic Hg exposure include gastro intestinal pain, vomiting diarrhea, loosening of the teeth, and renal damage [16] Methylmercury is rapidly absorbed through the gastrointestinal tract and easily penetrates both blood brain and placental barriers in humans and animals [15] Symptoms of methylmercury exposure include blurred vision or blindness, deafness, speech impairment, headaches, tremor, and loss of coordination or memory. The developing fetus is particularly sensitive to methylmerc ury exposure Prenatal exposure can result in developmental neurological abnormalities such as delayed onset of walking or talking and cerebral palsy [4] Epidemiological studies S tudies have not reliably add ressed the effects of maternal exposure to elemental Hg on the developing fetus [15] No studies on developmental toxicity associated with inorganic Hg exposure are available. The first epidemiologic report of methylme rcury poisoning is centered on the chronic methylmercury exposure that occurred in Minamata, Japan between 1953 and 1960. The Chisso Corporation factory released wastewater with high levels of Hg into the harbor, resulting in bioaccumulation of methylmercu ry in fish and shellfish ranging from 10 to 35 mg/L Subsequent consumption of these fish resulted in neurological symptoms in adults and both neurological and developmental symptoms in prenatally exposed children [17 ] In one study of 628 human cases, 78 deaths occurred [15]

PAGE 24

24 The effects of acute high level methylmercury poisoning were demonstrated in Iraq in 1971 when methylmercury fungicide treated seed designated for planting was instead ground into flour and baked into bread for human consumption. Prenatally exp o sed children exhibited symptoms including blindness, deafness, and paralysis [18] Environmental h ealth i mpacts Methylmercury ca n be formed in aquatic ecosystems through microbial metabolism and chemical processes. Sulfate reducing bacteria take up Hg in its inorganic form and convert it to methylmercury. M ethylmercury moves through the food chain when these bacteria are consumed o r release the methylmercury into the aquatic ecosystem. Top predators in the aquatic food chain, such as large fish, otter, mink, and raptors have the highest tissue levels of Hg [19] The pro cess of Hg bioaccumulation is complex and involves biogeochemical cycling and ecological interactions [4] Natural unpolluted surface waters are reported to have total Hg levels ranging between 0.1 and 5 ng/L. Assuming 1 ng/L total Hg and recognizing that methylmercury accounts for 1 to 10% of total Hg, the methylmercury concentration will range from 10 to 100 pg/L, which could easily exceed the Wildlife Criteria [4] Mercury Emissions Mercury release can occur from natural sources such as volcanic activity and weathering of rocks and, to a greater degree, from anthropogenic activity both current and historic (Table 2 2 ) [3] Coal fired power production is the single largest global source of atmospheric Hg emissions due to both an increasing global demand for power production and decreasing intentional use of Hg in industrialized countries.

PAGE 25

25 The chemical form of r eleased Hg depends upon its source, the environment, and other minor factors. As an element, Hg is persistent and cannot be broken down to less toxic substances. It is important to recognize that local releases of Hg have a global effect. Mercury can trans port long distances through ocean and air currents. Elemental Hg has an atmospheric residence time of several months to one year. So me models suggest that up to 50% of Hg deposited in North America is from external sources [4] Major pathways of anthropogenic Hg sources to water include direct discharge, indirect discharge, atmospheric deposition, and surface run off and leachate from contaminated soil and landfills. The majority of Hg in surface waters is due to air deposition related to anthropogenic activities, both domestic and international [20] Major point sources of Hg release to water in western countries include chlor alkali facilities, pharmaceutical industries, me tal processing plants, offshore oil activities, and coal fired power p lants Chlor alkali i ndustry The chlor alkali industry manufactures chlorine, hydrogen, and sodium hydroxide (caustic soda). The manufacturing process involves electrolysis of a salt so lution to convert chloride ions to elemental chlorine Three basic process variations for electrolytic production of chlorine are diaphragm cell, Hg cell, and membrane cell, with each using a different method to keep the chlorine product separate from the hydrogen and caustic soda. In the Hg cell process, Hg is used as the cathode where elemental sodium will accumulate while the chlorine will migrate to the anode. The chlorine is treated for sale and the sodium forms an amalgam with Hg This amalgam is the n used to produce hydrogen gas and caustic soda [21] Approximately 1 kg of Hg per 1000 kg chlorine produ ced is lost from the process including atmospheric losses and effluent

PAGE 26

26 waste stream [22] Although reliance on Hg cells at chlor alkali facilities is diminishing, 5 Hg cell facilities are still in operation in the United States and contribute approximately 7.1 tons per year anthropogenic Hg release [19] The reported constituent concentration ranges for chlor alkali wastewater are listed in Table 2 3 There is potential for a portion of the total Hg in chlor alkali wastewater to be in the el emental state Due to the influence of pH on Hg speciation, it is important to note that the pH of chlor alkali wastewater tends to be either acid ic (~ pH 2 ) or basic (~ pH 12 ) [23,24] In 20 03, t he EPA lowered the Hg national emission standard for hazardous air pollutants (NESHAP) by 3,068 kg per year applicable to Hg cell chlor alkali plants, Hg ore processing facilities, and sludge incineration and drying plants. Specifically, the final ru le limited Hg emissions from Hg cell chlor alkali plants to 2.3 kg Hg /day [25] In March 2011, the EPA proposed further reduction of Hg NESHAP by either eliminating the use of Hg fuel cell technology or improving work p ractices to reduce fugitive Hg emissions from the cell room to near zero levels. Flue g as d esulphurization Hg occurs naturally in coal in varying concentrations COALQUAL, a database that contains analyses of over 7,000 coal samples, identifies the mean H g concentration in coal as 0.17 g/g [26] When the coal is burned, Hg is released as an air pollutant, contributing 13 26% of the total airborne emissions of Hg in the United States [26] This necessitates the use of pollution control devices such as activated carbon injection that directly targets Hg or flue gas desulphurization (FGD) scrubbers that target sulfur dioxide but also co capture oxidized Hg

PAGE 27

27 FGD wastewater typically conta ins 10 80 0 g/L Hg primarily in the oxidized state [ 19, 27 95] ] The wastewater also tends to contain high levels of dissolved solids, suspende d solids, and organic compounds (Table 2 4 ). The pH of FGD wastewater typically falls within the range of 4.5 to 9. The EPA is currently working to revise the effluent limitations guidelines and standards for the steam electric power generating point sour ce category This category i ncludes FGD wastewater effluent. T hese new guidelines will likely address discharge limits for a variety of metals including Hg [28] Mercury e mission r egulations Mercury discharge is regul ated under the Clean Water Act (CWA) and the Resource Conservation and Recovery Act (RCRA) Mercury is listed as a toxic pollutant under section 307(a) of the Clean Water Act. For the protection of aquatic life, th e Clean Water Act established mercury wate r quality standards (WQS) of 1.4 g/L for an acute dose and 0.77 g/L for chronic exposure. Over 8,000 bodies of water in the United State s exceed WQS for Hg [20] Some regions of the U.S. has established more strict Hg regulations. The maximum ambient water concentration is an average 1.3 ng/L, according to the Great Lakes Initiative Wildlife Criteria RCRA requires that the EPA manage hazardous waste with a cradle to grave responsibility. Because of its toxicity, Hg is considered a hazardous waste. The EPA has established standards for the generation, transportation, storage, treatment, disposal, and recycling of hazardous waste, including mercury containing waste. Land disposal restrictions exist that may require waste to be treated prior to landfilling

PAGE 28

28 Aqueous m ercury removal t echnologies Sulfide precipitation. Sulfide precipitation, capable of achieving a minimum effluent of 10 100 /L Hg, is a common remediation method for Hg laden wastewater from both chlor alkali industry and coal fired power plants utilizing FGD wet scrubbers [29] As presented in Eq. 2 13, organic and inorganic sulfides react to form insoluble Hg sulfide (K sp at 25 C is 2 x10 53 ) bu t these compounds can be difficult to remove from the waste water, necessitating additional treatments such as pH adjustment, coagulation, flocculation, gravity settling, or filtration [29] Ou tside of the ideal near neutral pH range, soluble Hg S species form. HgS 2 2 forms at high pH while Hg(SH) 2 which forms at low pH [29,30] Hg 0 + Hg 2 2+ + Hg 2+ + S 2 2 H g 0 + 2Hg S (s) (2 13 ) Disadvantages of sulfide precipitation include the potential for Hg to resolubilize in certain landfill conditions, difficulty monitoring real time sulfide levels, the presence of toxic residual sulfide in the effluent and the diff iculty of treating and disposing of Hg and sulfide laden sludges [29] The reducing conditions of sulfide precipitation are ineffective for insolubilizing elemental Hg [31] T he sludge produced often requires a treatment such as mineral encapsulation to ensure it is inert. The costs of treating chlor alkali wastewater using sulfide precipitation were reported as $1.50/1000 gal, adjusted for inflation [32] This cost is higher if additional treatments are applied. Coagulation/co precipitation. As an alternative or used in conjunction with sulfide precipitation coagulation/co precipitation using aluminum sulfate (a lum) or iron salts can be used to treat aqueous Hg in wastewater. This treatment is capable of achieving effluent H g concentrations of 5 to 10 /L using alum and 0.5 12.8 /L

PAGE 29

29 using iron salts [29] Coagulation is most efficient when used in conjunction with pH adjustment and filtration. Adsorption Processes Ads orbents have the potential to achieve high Hg removal efficiencies Activated carbon, the predominantly applied adsorbent i s known to adsorb Hg(II) from aqueous solutions and can reach effluent levels of 0.5 to 20 [22,29,33 37] However, removal levels depend highly upon the initial concentrations, the pH, and the con centration of other pollutants competing for adsorption sites [29] Due to isothermal behavior of the adsorbent, incremental adsorbent dosage results in increased treatment efficiency but, unl ess recovery of the adsorbent is feasible, this increases the wastewater treatment residuals that require ultimate disposal. Granular activated carbon (GAC) is often applied as a fixed bed unit with columns in parallel or series. Powdered activated carbon (PAC) is often applied as a slurry and requires subsequent solids separation. Modification of activated carbon such as impregnation with carbon disulfide, bromine, or ozone, have been shown to enhance Hg removal [29,38,39] In anticipation of new and more stringent water quality based standards, adsorption can be used as a polishing technique to reach lower Hg concentrations in industrial wastew ater effluent [40] Activated Carbon Synthesis of Activated Carbon Activated carbon is made in two steps by first heating a carbonace ous precursor in an inert atmosphere to eliminate light and heavy carbon based oils and non carbon elements as volatile gases and then ac tivating thermally (physically) or chemically. A fter activation, the surface of the carbon is heterogeneous with a typi cal elemental composition of 88% C, 0.5% H, 0.5% N, 1% S, 6 7% O, and ash constituents [41] The

PAGE 30

30 amount of oxygen can range from 1 20% depending on raw material, activation, and additional treatments. The heteroatoms ty pically occur at edges and corners of the graphene sheet and behave similarly to the functional groups commonly found in aromatic compounds [35,42] The properties of activated carbon, such as surfac e area and pore size, are affected by the nature of the activation method as well as the source material [35] Thermal a ctivation Porosity. Thermal activation is performed using CO 2 or H 2 O (g) at temperatures over 400C to remove carbon atoms, thus creating mes o and macro porosity according to the stoichiometry shown in Equations 2 1 4 and 2 15 [35] Porosity development occurs by the opening of previously inaccessible pores, the creat ion of new pores by selective gasification of certain structural components, and the widening of existing pores. A t temperatures over 400F, the carbon atom attached to a surface ox y gen complex is a common site for gasification. C + CO 2 (g) 2CO (g) (2 14 ) C + H 2 O (g) CO (g) + H 2 (g) (2 15 ) Surface o xygen f unctionality. At temperatures below 400C, t he reactions of CO 2 steam, and O 2 with carbon can result in chemisorbed oxygen (Equations 2 16 and 2 17 ) Su rface oxygen complex formation is selective based on carbon surface heterogeneity and results in C(O) group with wide ranges of functionality with variable stability. The se groups can influence the wettability, polarity, acidity, adsorption behavior, and c atalytic and chemical reactivity of the carbon. C + O 2 C(O) (2 1 6 )

PAGE 31

31 C + H 2 O (g) C(O) (2 17 ) Possible basic C(O) groups formed are pyrone, first proposed by Boehm and Voll in 1970, and chromene, first proposed by Garten and Weiss in 1957 [42] Although the main source of carbon basicity is a result of these basic basicity can weakly contribute to the basic nature of a carbon [42] Possible a cidic groups are carboxyl, quinone, hydroxyl, carb onyl, carboxylic anhydride, and lactone [35] Acidic surface groups cause the carbon surface to be hydrophilic and polar. [22] Usually, both a cidic and basic groups are present on t he carbon surf ace. C onsequently, activated carbon is amphoteric. Chemical a ctivation As thermal activation primarily creates meso and macro pores, controlled wet chemical activation can be used to create microporosity. Chemical activation is commonly per formed by carbonizing the precursor at 450 to 600C in the presence of ZnCl 2 KOH, or H 3 PO 4 [35] Activated C arbon M odification Modification of existing activated carbon surface chemistry features can be performed chem ical ly (acidic treatment or impregnation) or physical ly (heat treatment). Acidic treatment e nhances C(O) [43] Physical modification enhances surface area, pore volume, and C(O). Activated carbon surface chemistry can b e manipulated using these techniques to produce adsorbents that are tailored for a particular function. Enhanced s urface oxygen functionality C(O) groups can be formed through acid treatment with t he amount of oxygen gained dependent up on the method and the precursor used [44 49] W et chemical

PAGE 32

32 oxidation uses oxidizing aqueous solutions such as ozone [50] nitric acid [43,45,48,51] and hydrogen peroxide [45,51] Nitric acid is the most widely used method of increasing the total acidity in a wet chemical oxidation [48,50] Wet oxidations are generally thought to minimally alter other surface chemistry characteristics such as por e size distribution [43,45,47,49,52,52] but several researchers have found that concentrated nitric acid ox idation reduced the BET surface area and total pore volume while the pore width increased due to pore collapse [ 53 57] Salame noted a loss in mesopore volume specifically associated with oxidation using concentration nitric acid and ammonium persulfate [58] Oxidation with hydrogen peroxide increases the volume of pores havi ng a diameter of ~6A [45,47,59] Iron i mpregnation Researchers have previously experimented with magnetic adsorbents. Oliveira et al. [60] created an activated carbon/iron oxide magnetic composite via fast hydrolysis at pH 10 of a 2:1 Fe (III) : Fe (II) and 1:1, 2:1, and 3:1 C:Fe. Magnetization, X ray diffraction and Mossbauer data suggest that the main magnetic phase present in the com posite is maghemite. T emperature programmed reduction (TPR) data suggests that the iron oxides present can be reduced to magnetite, enhancing the magnetization. The experiment also determined that the surface area loss was proportional to the iron loading. The composites did not significantly lose magnetic strength in the pH range of 5 11. Oliveira et al. [61] also synthesized a magnetic zeolite for Cr 3+ Cu 2+ and Zn 2+ removal from water. Gorria et al. [62] synthesized a magnetic adsorbent by depositing nickel nanoparticles on activated carbon. Magnetism. Iron (Fe) is a malleable transition metal with an atomic number of 26, atomic mass of 55.85, and an electron configuration o f [Ar]4s 2 3d 6 Iron exists in two

PAGE 33

33 main oxidation states, Fe(II) (Fe 2+ ferrous Fe) and Fe(III) (Fe 3+ ferric Fe). Ferrous iron spontaneously oxidizes to ferric iron, reducing solubility. The 3d electrons determine magnetic properties. E ach d orbital occupie s a different orientation in space: d xy d yz d xz d z2 d x2 y2 Coordination to oxygen or hydroxyl causes unequal energy distribution in t he d orbita ls [63] Magnetic properties arise because of interactions between the spin moments of the electrons and the orbital moment. Ferromagnetic materials possess parallel electron spins, resulting in an overall net magnetic moment with large permeability (ratio of magnetic flux density to external field strength) and large positiv e susceptibility (strong attraction) to an external magnetic field. Ferromagnetic materials are spontaneously magnetic and retain their magnetic properties after the external field has been removed [63] Antiferromagn etic materials possess electron spins of equal magnetic moment with antiparallel alignment, resulting in zero overall magnetic moment, positive permeability, and a small positive susceptibility. Magnetite Magnetite (Iron (II,III) Oxide) is a naturally oc curring ferrimagnetic iron oxide with inverse spinel structure and a face centered cubic unit cell based on 32 O 2 ions. The tetrahedral sublattice (A) contains one Fe 3+ atom surrounded by four oxygen atoms while the octahedral sublattice (B) contains one iron atom, either Fe 3+ or Fe 2+ surrounded by six oxygen atoms, thus forming the two interpenetrating magnetic sublattices. The saturation magnetism of magnetite ranges from 92 to 100 Am 2 /kg. Magnetite contains eight formula units, Y[XY]O 4 (X=Fe 2+ Y = Fe 3 + ), per unit cell. The unit cell edge length is 0.839 nm and surface area ranges between 4 and 100 m 2 /g. Magnetite is frequently non stoichiometric and iron can be partly of fully replaced by

PAGE 34

34 other metal ions depending upon steric hinderance (based on atom ic radii and valence). Substitution changes the unit cell edge length and therefore can be identified via XRD analysis [63] Cation substitution of mercury for iron in the iron oxide structure can be ruled out based o mercury is too large to substitute for either ferrous or ferric ions [64] Maghemite. Maghemite, a structural polymorph of magnetite, is a natural ly occurring ferrimagnetic iron oxide with spinel ferrite structure. Maghemite has a cubic unit cell based on 32 O 2 ions and a unit cell length of 0.834 nm. Each unit cell contains 32 O 2 3+ d an Fe 2+ deficient magnetite. The iron cations are randomly distributed over 8 tetrahedral (A) and 16 octahedral (B) sublattices with randomly distributed vacancies limited to the octahedral sites. Due to the structure of maghemite, the saturation magneti sm can vary from 60 to 80 Am 2 /kg. Maghemite has a surface area ranging from 8 to130 m 2 /g [63] Adsorption Adsorption theory The current understanding identifies adsorption as a surface phenomenon that results from un saturated and unbalanced molecular forces on a solid surface that are satisfied by attracting adsorbate molecules, atoms, or ions, resulting in a higher concentration of these particles on the solid surface relative to the bulk solution. Activated carbon a dsorption can by physical or chemical. Physisorption occurs through van der Waals attraction (dispersion forces). Asymmetry of the electron distribution in the adsorbate particle causes a transient dipole moment that, when it is approaching the solid adso rbent surface, can induce an appropriately oriented dipole moment in a surface molecule, producing instantaneous

PAGE 35

35 attraction. The se forces are greater in the micropores where the adsorbate molecules can be closer to each other than in the bulk aqueous phase [35] Physisorption is a reversible exothermic process that is not site specific and can re sult in multimolecular thickness of the adsorbed phase. Chemisorption forces arise from redistribution of electrons between th e adsorbent and adsorbate, resulting in a site specific irreversible chemical bond [65] Chemisorption results in unimolecular thickness of the adsorbed phase. Due to the nature of chemisorption, it is much stro nger than physisorption. Three successive steps are commonly proposed to describe adsorption dynamics on porous adsorbents. First, the solute is transported from bulk solution through a liq uid film xt, most of the solute that was transported from the bulk solution diffuses into the pores while a small quantity remains on the external surface (internal diffusion). This is the rate limiting step. Finally, the solute is adsorbed on the interior surface of the pores and capillary spaces of the adsorbent, reaching equilibrium. These steps are influenced by the affinity of the solute for the surface, the solvent for the surface, and the solute for the solvent [66] Adsorption is an equilibrium process. Initially, adsorption proceeds at a rapid rate due to the availability of surface sites for adsorption but, as adsorption sites fill, the rate of adsorption slows while the rate of desorption increases until reaching equilibrium where the rate of adsorption equals the rate of desorption. At a constant temperature, adsorption equilibrium can be represented as an adsorption isotherm. Two common isotherm equations applied to liquid phase adsorption, Freundlich and Langmui r, apply to both chemisorption and physisorption.

PAGE 36

36 Isotherm t heory Adsorption isotherms utilize controlled physisorption and desorption onto a sorbent. An adsorption isotherm is the graphical representation of the relationship between the bulk adsorbate an d the amount adsorbed at a given temperature [67] The International Union of Pure and Applied Chemistry ( IUPAC ) cl assifies adsorption isotherms into six categories as follows (Figure 2 7 ) [35,68] : 1. Type I isotherms, a lso referre d to as Langmuir isotherm s are concave with respect to P/P 0 This isotherm reaches a maximum value of adsorption. The steepness of the slop e of the isotherm from P/P 0 values of zero to 0.05 indicates the narrowness of the micropores. It is generally accepted that Type I isotherms represent microporous solids with a small external surface area such as activated carbon and zeolites. 2. Type II iso therms describe adsorption in the presence of both micropores and open surface. This isotherm contains an inflection point where the curve changes from concave to convex, representing where monolayer coverage ends and multilayer adsorption begins. These is otherms represent solids that are either non porous or macroporous. 3. Type III isotherms are convex and are typical of adsorption at sites with low adsorption potential, such as organic polymeric systems. 4. Type IV isotherms are similar to Type II isotherms bu t includes mesoporosity. Activated carbons will not typically present a plateau in the high relative pressure region. 5. Type V isotherms are characteristic of a low energy, homogeneous, mesoporous solid. 6. Type VI isotherms characterize extremely homogeneous surfaces such as pyrolytic graphite. Measurement is performed using argon or methane rather than nitrogen. Desorption can be slower than adsorption due to a higher activati on energy, forming a hysteresis in which the adsorption and desorption curves of t he isotherms do not follow the same pa th (Figure 2 8 ). Line PQ describes adsorption in microporosity and open surface; smaller pore size results in a steeper PQ line. Line QR indicates reversible

PAGE 37

37 adsorption in the smallest mesopores. Line RS indicates capi llary condensation. Upon lowering the pressure, desorption follows the line SUR. IUPAC has established four categories of hysteresis loops (Figure 2 9 ). When a hysteresis loop occurs within the multilayer range of a gas adsorption isotherm (relative press ure of >0.2), it is usually associated with capillary condensation in mesopores, shown as H1 and H4. The H2 and H3 hysteresis loops are intermediate between these two extremes. The dashed lines represent low pressure hysteresis due to microporosity. Hyster esis shape is often identified with specific pore structures. Type H1 loops are often associated with porous materials consisting of approximately uniform spheres in a regular array and thus a narrow pore size distribution. Type H2 loops do not have a well defined pore size distribution or shape. This hysteresis at one point was attributed to ink bottle pores but this view is now recognized as over simplified. Type H3 loops is associated with slit shaped pores due to plate like particles. Type H4 is also as sociated with slit shaped pores but the Type I isotherm character indicates microporosity [68] Langmuir Isotherm equation The Langmuir equation was the first adsorption isother m equation developed (Equation 2 18 ). This equation relates the amount adsorbed to the equilibrium concentration of the adsorbate in the bulk solution where Y/M is the concentration of adsorbate adsorbed (mg/L) divided by the carbon concentration (mg/L); C is the equilibrium concentration (mg/L), and a and b are constants, determined graphically. The assumptions in this equation are 1) adsorbate is attached to the surface at definite localized sites, 2) each site accepts one adsorbate particle 3) the energy state of the adsorbate is equal at all sites (energetically

PAGE 38

38 homogenous surface with negligible lateral interactions). This equation is idealized and thus its significance in interpreting adsorption data can be limited. (2 18 ) Freundlich Isotherm equation The Freundlich equation relates the solute concentration on the adsorbent surface to the concentration of the solute in the bulk solution where Y/M is the concentration of adsorbate adsorbed (mg/L) divided by the carbon co ncentration (mg/L) C is the equilibrium concentration of the adsorbate in the bulk solution (mg/L) and both k and 1/n are constants. (Equation 2 19 ). A plot of log Y/M versus log C yields a straight line wi th a slope of 1/n and a y inter cept of k, which holds true over a wide range of concentration values including dilute solutions [69] The Freundlich equation is often applied to physisorption and adsorption of solids of limited solubility. Y / M = k C 1/ n (2 19 ) Aqueous phase metal adsorption Aqueous phase adsorption involves interactions between the solute and surface, the solvent and surface, and the so lute and the solvent. Issues that must be considered are competitive adsorption, chemical changes of the solute, and concentration changes of the solute. The solution pH can play a large role in adsorption as the concentration of acidic molecules is functi on of pH and both the dissociated and the non dissociated forms may adsorb. In general, low solubility favors aqueous adsorption. [42] There are several theories regarding adsorption of metal ions. The first theory is electrostatic adsorbate adsorbent interaction (ion exchange). This process is entirely dependent upon the functionality of the carbon, particularly the C(O) complexes. The

PAGE 39

39 second theory is that enhanced adsorption potentials (dispersion forces in the narro west micropores) are strong enough to retain metal ions. The third theory is that of hard and soft acids and bases (HSAB) in response to the amphoteric nature of the carbon surface [35] Metal adsorption can be influenc ed by various characteristics of the adsorbent, matrix, and adsorbate. Adsorbent surface chemistry characteristics that influence adsorption include surface area, pore size distribution, as well as C(O) and other heteroatom functionality. The role C(O) com plexes is determined by a correlation between the amount of ion adsorbed and the amount of participating oxygen functionality. Matrix characteristics that can influence metal ion adsorption include the pH, temperature, and presence of competitively binding ions. Chemical and physical properties of the metal ion adsorbate influence on adsorption; adsorption is affected by ionic radius (access to porosity), solubility (hydrophobic interactions), and pKa (controls dissociation) [35] Mercury adsorption from aqueous solution H g(0) adsorption by activated carbon. While the low solubility of Hg(OH) 2 allowed for removal of Hg via preferential precipitation, Hg(0) does not precipitate and its low solubility and high volatility re sult in more difficult aqueous removal than oxidized species [22,56,70,71] Vapor phase Hg(0) adsorption by activated carbon is known to be affected by var ious mat rix and sorbent characteristics, including surface oxygen functionality [22,27,56,70,71] Gas phase research implicates C(O) c omplexes, reporting that two carbons with similar sulfur, chlorine, bromine, and iodine distribution displayed very differen t sorption capacities for Hg(0) most likely due to differences in surface oxygen functionality [72]

PAGE 40

40 Li et al. [56] proposed that C(O) complexes, particularly the reducible lactone and carbonyl groups, are possible active sites for gas phase Hg(0) adsorption, potentially involving electron tr ansfer from the Hg (0) to th e lactone or carbonyl, followed by subsequent adsorption of Hg(II) through well studied mechanisms. A dsorbed Hg(0) was desorbed as Hg(II), lending support to the oxidation hypothesi s In a theoretical study Liu et al. [73] concluded that lactone and carbonyl favor gas phase Hg(0) adsorption while phenol and carboxyl reduce d Hg(0) capture [73] The role of C(O) complexes in aqueous Hg(0) adsorption is not defined in literatur e. Hg(II) adsorption by activated carbon. Activated carbon is known to have a high affinity for Hg (II) Multiple f actors can influence Hg(II) adsorption, in cluding temper ature surface area and pore volume, and particle size [22,72,74 76] Aqueous Hg(II) can be removed from solution by physisorption, ion exchange, hydrogen bonding, surface precipitation, or reduction/volatiliz ation. C(O) functionality can contribute to Hg removal from solution [75,77] When the pH < pzc, c ationic Hg must overcome electrostatic repulsion to exchange with the H + of a surface oxygen group (Equation 2 20 ) while a nionic Hg is electrostatically attracted to the positive carbon s urface [22,35,75] When the pH > pzc, cationic Hg is electrostatic ally attracted to the d eprotonated C(O) group (Equation 2 2 1 ) while mercury anions are electrostatically repelled by the negative sorbent surface [78] 2 C COOH + Hg 2+ ( C COO) 2 Hg + 2H + (2 20 ) 2 C COO + Hg 2+ ( C COO) 2 Hg (2 21 )

PAGE 41

41 Hydrogen bonding can take place between a n H atom on hydrolyzed Hg and an electronegative surface oxygen. W hen Hg(OH) 2 h as r eached its intrinsic solubility, i t will preferentially precipitate on the carbon surface rather than in solution [79] Activated carbon has been shown at high pH values to remove mercury via reduction and volatilization as Hg (0) [33,80] Phenolic and hydroquinonic surface oxygen groups have been proposed as reduction sites (Equation 2 2 2 ) [36] Confirmed by scanning electron microscopy, HgC l 2 reduction to the sparingly Hg 2 Cl 2 will cause preferential precipitation onto the carbon surface while a complete reduction to Hg ( 0 ) result in Hg volatiliz ation from solution [81] Many researchers do not attempt to dis tinguish the mercury removed via adsorption from the mercury removed via reduction and volatilization. The amount of Hg adsorbed can be determined by the mas s balance equation (Equation 2 23 ). 2( OH) + 2HgCl 2 2(=O) + Hg 2 Cl 2 + 2HCl (2 22 ) [ TOT Hg] = [Hg(II) aq ] + [Hg(0) g ] + [Hg(II) ads ] + [Hg(0) ads ] (2 23 ) Hg adsorption by iron oxides Iron oxides including m agnetite, goethite, and ferrihydrite have been shown to adsor b aqueous Hg(II) [82 84] The ion loading, as with activated carbon, is a function of matrix pH (Equation 2 2 4 and 2 25 ) [63,83] Ternary surface complexes can also form between the surface, Hg 2+ and OH or Cl (Equations 2 2 6 and 2 2 7 ) [85] Hg is likely to chemisorb onto Fe oxides than to phy sisorb [63] Fe OH + Hg 2+ Fe O Hg + + H + ( 2 24 ) ( Fe OH) 2 + Hg 2+ ( Fe O) 2 Hg + 2H + ( 2 25 ) Fe OH + Hg 2+ + H 2 O Fe O Hg OH + 2H + ( 2 26 ) Fe OH+ Hg 2+ + Cl Fe OH Hg Cl + H + ( 2 27 )

PAGE 42

42 The presence of specific ligands can influence the adsorption of mercury onto iron oxides [86] Sulfate has been shown to increase Hg(II) sorption onto iron oxides by reducing the positive surface charge and thereby reducing the electrostatic repulsion that can inhibit adsorption of Hg cations onto the oxide surface [12] Hg reduction by iron oxides. Oxidation of iron oxides can occur with the reduction of an aque ous transition metal [64,87 90] In anoxic conditions, Hg(II) is lost as Hg(0) in the presence of magnetite, shown in E quation 2 2 8 in which n is the charge transfer number and z is the valence state of the transition metal [82,90] Reduction rates decrease with pH. 3[Fe 2+ Fe 2 3+ ]O 4 + m z 2 3+ ]O 3 + Fe 2+ + m z 1 (2 2 8 ) Hg oxidation by iron oxides. In the air pha se, magnetite and maghemite have been shown to oxidize Hg (0) to Hg(II) [91] The water content an d surface area have been shown to impact the Hg(0) oxidation [92,93] I ndirect evidence for Hg (0) oxidation is seen in reduced adsorption in the presence of chloride ions; Elemental mercur y must ionize in order to complex with chloride ions [83,84]

PAGE 43

43 Table 2 1. Stability con stant values for Hg OH and Hg Cl compounds Benjamin Snoeyink Hahne & Kroontje Ligand Complex l og K 1 Ligand Complex log K 1 Ligand Complex log K 1 OH HgOH + 10.6 10.6 OH HgOH + OH HgOH + 11.86 11.86 Hg(OH) 2 11.3 21.9 Hg(OH) 2 Hg(OH) 2 10.27 22.13 Cl HgCl + 6.75 6.75 Cl HgCl + 7.15 7.15 Cl HgCl + 6.74 6.74 HgCl 2 6. 37 13.12 HgCl 2 o 6.9 14.05 HgCl 2 o 6.48 13.22 HgCl 3 1 0.90 14.02 HgCl 3 1 2.0 15.15 HgCl 3 1 0.9 14.07 HgCl 4 2 0.41 14.43 HgCl 4 2 0.7 15.75 HgCl 4 2 1.0 15.07

PAGE 44

44 Table 2 2 Select anthropogenic releases of Hg [4] Mobilization of Hg impurities Coal fired power and heat production Cement production (Hg in lime) Mining and other metallurgic activities Intentional extraction and use Hg mining Chlor alkali production Use of fluorescent lamps Waste treatment Waste incineration Landfills Table 2 3 Reported ranges of chlor alkali wastewater constituents [23,24] Constituent Concentration r ange (mg/L) Total Hg 1.6 7.6 Hg(0) 0.004 0.036 Chloride 460 25,000 Ammonium 0 0.8 Nitrite 0 1.7 Nitrate <5 150 Sulfate 12 650 Ca 27.7 Cd 0.6 Mg 33.3 Na 311.2 Pb 2.9 Dissolved o xygen 6.8 9.1

PAGE 45

45 Table 2 4 Reported ranges of FGD wastewater constituents [19,95] Constituent Concentration r ange (mg/L) Hg 0.01 0.8 Suspended s olids 250 20 ,000 Chloride 1,000 40,000 Ammonium < 10 100 Nitrite < 2 Nitrate 10 20,00 0 Sulfate 1,500 8,000 Sulfite < 20 Sulfide < 20 Ca 750 4,000 Cd < 1 Cr < 5 Cu < 5 Mg 1,100 4,800 Na 670 4,800 Ni < 5 Zn < 10

PAGE 46

46 Figure 2 1. Hydration of Hg 2+ ion in water Figure 2 2. 3 dimensional geometry of Hg 2+ hydration F igure 2 3. Distribution of Hg(II) at different pH values 0.0 0.2 0.4 0.6 0.8 1.0 0 2 4 6 8 10 Fraction pH Hg 2+ HgOH + Hg(OH) 2

PAGE 47

47 Figure 2 4. Mercury Eh pH diagram for Hg O H S Cl syst em Figure 2 5. Distribution of Hg(II) at various chloride concentration s 0 0.2 0.4 0.6 0.8 1 -14 -12 -10 -8 -6 -4 -2 0 Fraction log [Cl] Hg2+ HgCl+ HgCl2 HgCl3HgCl42-

PAGE 48

48 A. B. C. D. F igure 2 6 Hg (II) Speciation at varying pH and chloride concentrations (pCl 7 is 3.5 g/L, pCl 1.85 is 500 mg/L, and pCl 0.15 is 25,000 mg/L )

PAGE 49

49 Figure 2 7. IUPAC gas adsorption isotherm classifications Figure 2 8. Nitrogen adsorption isotherm on mi cro and mesoporous carbon exhibiting a closed hysteresis loop

PAGE 50

50 Figure 2 9. Types of hysteresis loops observed during adsorption

PAGE 51

51 CHAPTER 3 MATERIALS AND METHOD S Che micals and Materials All chemicals used in this work were analytical grade and were applied without further purification. Solutions were prepared using ultrapure Type I water with a resistivity of 18.2 M and a conductivity of 0. 055 S. Hg(II) solutions were prepared by diluting 1000 mg/L stock Hg(NO 3 ) 2 standard solution (Fisher Scientific) in ultrapure water. Prior to preparing Hg(0) solutions, metallic Hg was washed with 0.1M HNO 3 and rinsed five times with ultrapure water to remove oxidized Hg compounds from the surface [98] Hg(0) solutions were prepared by mild heating of elemental Hg under N 2 (g) flow and bubbling the Hg laden N 2 (g) through N 2 (g) purged ultrapur e water for 2 hours to reach an aqueous concentration of 10 /L to 54 /L [99] Commercially available carbons were oven dried at 100 C for a minimum of 24 h prior to use Calgon WPH i s a steam activated powdered carbon made from bituminous coal with a n approximate surface area of 1020 m 2 /g. Norit CASPF is a wood based chemically activated powdered activated carbon with a surface area of about 1200 m 2 /g. Materials Synthesis Iron Impregnation Magnetic powdered activated carbon ( MPAC ) composites were synthesized at room tempe rature by heterogeneous nucleation [11] Fe(II) and Fe(III) salts (ferric chloride (FeCl 3 ) and ferrous ferric oxide (FeO, Fe 2 O 3 )) were dissolved in ultrapure water with mechanical stirring. After carbon addition, rapid alkaline hydrolysis was induced by

PAGE 52

52 adding 5 M NaOH drop wise to the solution to reach pH 10. The hydrolysis products, Fe(OH) + and Fe(OH) 2 + reacted to form ferrihydrite which preferentially precipitated onto the WPH carbon surface but, due to therm odynamic instability, transformed into magnetite (Fe 3 O 4 ) (Equations 3 1 and 3 2) [100] In the presence of atmospheric oxygen, the magnetite is susceptible to oxidation to maghemite [63] 2Fe(OH) 2 + + Fe(OH) + + 3OH 3+ ) 2 (Fe 2+ )(OH ) 8 (3 1) (Fe 3+ ) 2 (Fe 2+ )(OH ) 8 3 O 4 + 4H 2 O (3 2) The amount of activated carbon was adjusted to obtain 1:1, 2:1, and 3:1 C:Fe mass ratios. Samples were rinsed with ultrapure water to remove residual NaOH until a constant water contact pH was achieved and subsequently oven dried at 100 C overnight. Although maghemite is likely the predominant iron sp ecies present on the MPAC surface due to the synthesis technique used, small amounts of non magn etic iron oxides (e.g. hematite or amorphous iron oxides) may occur. Thermal oxidation may convert some of these amorphous iron oxides to magnetic iron oxides s uch as magnetite or maghemite [63] To compare the initial synthesis product to one having undergone thermal oxidation, representative portions of the original MPAC were subjected to oxidation in a programmable muffle furnace (Barnstead Thermolyne 47925 80) under atmospheric air flow The program increased the temperature by 5 C until the desired temperature was reached (250 C, 350 C, and 450 C) held for the desired duration (3 or 6 h), and then gradually cooled.

PAGE 53

53 Nome nclature for the materials is based on carbon to iron ratio, the oxidation temperature and time. For example,1:1 450 6 h repr esents a WPH carbon sample impregnated with a 1:1 mass ratio of Fe prior to a 6h thermal oxidation at 450 C. Surface Oxygen Modif ication Commercially available carbons were modified by wet chemical oxidation at room temperature by exposure to 1M, 5M, and 10M solutions of HNO 3 H 2 SO 4 and NaOH for 6h. Samples were then rinsed with ultrapure water until reaching a constant water conta ct pH and subsequently oven dried at 100 C overnight. As a control, a sample of the virgin WPH carbon was stripped of its surface oxygen groups at 950C under 150 mL/min H 2 (g) flow for 180 min [56,101 103] While temperatures under 400C result in the formation of C(O), temperatures over 400C decompose acidic C(O) groups to CO 2 while basic groups decompose to CO ( E quations 2 17 to 2 20) [22] [101] Anhydrides are removed at 550C, phenols at 630C, lactones at 670C, and 810C for carbonyls and quinones [52] The resulting carbon is basic d ue localized electron pairs at the edges of the graphene layers [101,101,104] Using H 2 rather than N 2 He, or another inert gas flow minimizes O 2 chemisorption after stripping by producing relatively stable edge carbons without unpaired electrons, thus maintaining a hydrophobic carbon surface [50,101,105] This treatment minimally influences porosity [49,52,56,106] The modification process has the potential to form humic substances which may block adsorbent porosity, reducing Hg adsorption. A humic s ubstance removal wash of 0.1 M NaOH followed by a 0.1 M HCl rinse was investigated [107]

PAGE 54

54 Nomenclature for the materials is based on the activated carbon used and both the concentration and the identit y of the modifying reagent. Nitric acid, sulfuric acid, and sodium hydroxide modified carbons are identified as NAC, SAC, and SHAC, respectively. For example, CASPF carbon th at was modified with 5M HNO 3 is represented as CASPF NAC 5M The H 2 (g) stripped carbon is identified as ACH. The feasibility of modifying biochar rather than commercially activated carbon was also investigated (Appendix A) Activated Carbon Characterizat ion Methods Porosity Instrumentation Nitrogen adsorption desorption analyses were performed using a Quantachrome NOVA 2200e. The operating theory, based on ideal conditions, states that the moles of nitrogen transferred from the manifold of a given volume (V m ) at temperature T a into an empty sample cell partly immersed in liquid nitrogen is equal to the moles of nitrogen transferred to the cell cold zone plus the moles transferred to the warm cell zone [108 ] Each sample was outgassed at 110 C under vacuum for 24 h to removed physisorbed substances Then, nitrogen was added and removed in finite volumes at specific pressures with temperature held constant at approximately 77K using a liquid nitrogen bath. The quantity of adsorbed gas plotted against the relative equilibrium pressure results in a hysteresis loop. Surface area The surface area of each sample was calculated by the Brunauer Emmett Teller (BET) equation (Equation 3 3 [109] ) for P/P 0 = 0.1 to 0.3 in which W is the weight of the adsorbed gas at P/P 0 W m is the weight of the adsorbed gas at monolayer coverage,

PAGE 55

55 and C is the BET constant. The BET method is the most widely used procedure for surface area analys is of solids. (3 3 ) The C constant, related to the enthalpy of adsorption of the monolayer, indicates the degree of attraction between the adsorbed gas and the solid is sufficien t to achieve monolayer coverage. A C constant value over 200 indicates micropore filling. The BET method assumes adsorption sites are uniform and randomly occupied, monolayer molecules serve as sites for subsequent layer adsorption, and subsequent layers h ave liquid like properties. When analyzing data using the BET equation, it is important to use the following parameters to reduce the potential for error. The correlation coefficient (R 2 ) should be no less than 0.9975 and the C constant, calculated from th e slope and y intercept, must never be negative. Additionally, the P/P 0 value with the maximum single point BET value should be used as the upper limit for the multipoint BET range. A minimum of three, preferably five, relatively equally spaced data points should be used in the multipoint BET calculation. Finally, data points that curve upward from the straight line at low relative pressure and data points that curve downward from the straight line at high relative pressure should not be used in the multipo int BET calculation [108] Pore v olume Total pore volume is calculated from the amount of vapor adsorbed at the limiting pressure, P/P 0 = 0.99. This assumes that all pore space is filled with adsorbate. I f no macropores are present, the isotherm will remain nearly horizontal over the range of P/P 0 approaching unity. If macropores are present, the isotherm will rise rapidly as the

PAGE 56

56 P/P 0 nears unity. If mesopores are present, the slope should plateau near the limiting pressure, indicating the all pore space is filled. The average pore size is estimated from the pore volume. Pore size Pore volume is distributed over various pore sizes, represented by a pore size distribution. IUPAC classifies pores according t o width [68] Macropores have a pore diameter over 500 (50nm) while micropores fall under 20 (2nm); Mesopores fall in between the two. Various relative pressures correspond t o the seque nce of gas adsorption (Table 3 1 ). P ore size calculations were based upon the Kelvin equation which relates the vapor pressure above a liquid to the pore diameter (Equation 3 4 ) [35] is the surface t ension, is the molar volume of the liquid, R is the molar gas constant (8.314 x 10 7 J/mol K), and r k is the effective radius of curvature The equation is based on the principle that equilibrium vapor pressure over a concave meniscus of a liquid adsorben t is less than the saturation vapor pressure at the same temperature. Therefore, a gas can condense as a liquid inside the porosity of a solid with sufficiently small pore radii filling with liquid at lower equilibrium vapor pressure values, describing cap illary condensation. The pore size distributions over the mesopore region were calculated using the Barrett Joyner Halenda (BJH) equation [110] ( 3 4 ) Point of Zero Charge The surface chemistry of activated carbon is dominated by its amphoteric nature which is dependent u pon heteroatom content, mainly oxygen When immersed in water,

PAGE 57

57 carbon deve lops a surface charge from the dissociation of surface groups or the adsorption of ions from solution. A negative charge can result from dissociation of acidic C(O) while a positive charge may be due to basic C(O) When the pH is lower than the pzc value, water donates more H + than OH groups so the adsorbent surface is positively charged and attracts anions. When the pH is above the pzc value, surface groups will dissociate, leaving the sorbent surface negatively charged, attracting cations (Figure 3 1). I on loading as a function of pH has also been demonstrated for the adsorption of many heavy metals ions by activated carbon [49,111,112] The point of zero charge (pH p zc ) was determined using the abbreviated version (10% by weight). Ultrapure water was purged with N 2 (g) for 20 min before carbon addition for a 24 h contact time, after which the solution pH was obtained in duplicate under N 2 ( g) headspace flow using an Acc umet AB 15 pH meter The manufacturer satted instrument sensitivity is between 1.99 and 19.99 with an accuracy of 0.01 pH units. Total Acidity Titration The Boehm titration technique is a classical equilibrium acid base titration that provides information regarding acid/base features of the carbon surface [41,113] Carbon samples were prepared for total acidity titration using the Boehm titration method by adding 0.5 g carbon and 0.1 g KCl to 25 mL o f 0.05 N NaOH and 0.05 N HCl, respectively, and rotating end over end for 48 h [22,46,47]. The KCl was added to increase the ionic strength of the solution. The titration is performed against a blank with any base consumed due to neutralization of surface functional groups. Blank solutions were prepared using 25 mL 0 .05 N NaOH and 0.05 N HCl each with 0.1g KCl. After the

PAGE 58

58 filters (Fisher Scientific). Filtrate was purged with N 2 (g) for 10 min prior to titration. 0.05N NaOH samples were titrated with 0.1N H 2 SO 4 to pH 4.5 whi le 0.05 N HCl solutions were titrated with 0.05 N NaOH to pH 11 Total acidity was calculated as the difference between the volume of titrant consumed in the sample titrations and the volume of titrant consumed in the appropriate blank titrations with the difference being due to neutralization of surface functional groups. Elemental A nalysis Moisture content of the carbons was determined by the mass difference before and after heati ng at 90C for 16 h. Ash content was determined by the mass difference after heating to 650 C for 16h. Elemental composition (C,H,N) was determined by a Carlo Erba EA 1108 elemental analyzer Assuming negligible presence of other elements, oxygen content was determined by mass difference. X Ray Diffraction X ray diffraction (XRD) can be used to determine purity, crystal size, disorder, and degree of isomorphous substitution. XRD observes the interaction of electromagnetic rays pass through a crystal, each atom in the structure scatters the waves uniform ly in space but in certain directions all the waves combine for enhanced intensity. The direction of this occurrence is related to the distance between atomic planes and the angle that the x rays enter and leave the crystal (Bragg angle). The XRD diagram i s a plot of the observed diffracted intensity vs. Bragg angle [63] X ray diffraction (XRD) patterns of the MPAC were recorded using a Philips APD 3720 X ray unit with Cu K radiation. XRD patterns were analyzed to identify the iron speciation on the MPAC surface. Components were identified using the powder

PAGE 59

59 diffraction identification number according to the International Center for Diffraction Data. Peaks greater than 3 of the baseline noise were used. Vibrating S ample M agnetometry Vibrating Sample Magnetometry (VSM) is used to measure magnetic properties as a function of the external magnetic field strength, temperature, and time. The theory of operation is based upon Farad uniform magnetic field (H), a magnetic moment (m) will be induced in the sample, producing a voltage in stationary sensing coils proportional to the magnetic moment induced (Figure 3 2 ). The data is pr esented as a hysteresis loop that shows the relationship between the induced magnetic flux density (B) and the magnetizing force (H) (Figure 3 3 ). Magnetic characteristics of the MPAC composites were measured using Princeton Measurements Co. MicroMag VSM 3 900. Saturation (value o f B at points a and d ), occurs when almost all magnetic domains are aligned. Therefore, increasing the magnetizing force will not significantly increase the magnetic flux. Retentivity (value of B at point b) indicates the remanence (level of residual magnetism) of the material when the magnetizing force is reduced to zero. This occurs as some magnetic domains remain aligned but others have lost their alignment. Coercivity, H c (value of H at point c) is the amount of reverse magneti c field required to return magnetic flux magnetizing force, describes the ease with which a magnetic flux is established in the material [114] Magnetic A d s orbent R ecovery MPAC, easily dispersed in a queous solution, can be retrieved using a strong magnet such as neodymium, a rare earth magnet. The recovery (%) of MPAC from

PAGE 60

60 aqueous solution and ad sorbent mass balance was determined using the dry mass captured by the magnet, the dry mass retained by a 0 .45 nitrocellulose filter after vacuum filtration, and the dry mass of the initial MPAC dose. The contact time (5 min) and carbon dose (1 g/L) were held constant while the MPAC species varied based on synthesis variables. Preliminary experimentation indicat ed the use of a 5 min contact time because the results did not significantly vary above this contact time while a 1 min contact time produced considerably lower magnetic sorbent recovery from aqueous solution. Adsorbent S tability Iron. As Fe is redox sen sitive with ferrous iron being highly soluble, Fe effluent levels were quantified using a spectrophotometer (Hach DR/4000 Spectrophotometer ) and This method requires 10 mL aqueous sample to which the TPTZ Iron Rea gent Powder Pillow is added, shaken for 30 s and allowed to react for 3 min prior to measurement. Each run was performed with standards including a blank. The manufacturer stated estimated detection limit is 0.022 mg/L total Fe. Mercury. As the modified carb ons have adsorbed toxic Hg, their disposal is potentially regulated under the Resource Conservation and Recovery Act. The Code of Federal Regulations (40 CFR §261.24) identifies Hg as a contaminant that must be tested for using the toxicity characteristic leaching procedure (TCLP ; EPA method 1311) Resulting leaching must have a n Hg level under 0.2mg/L in order to be considered non hazardous. Higher leachate levels necessitate the treatment of the spent adsorbent as a hazardous waste, greatly increasing dis posal costs as it cannot be disposed of in a sanitary landfill.

PAGE 61

61 The appropriate extraction fluid is determined by the water contact pH. Because the water contact pH of the Hg loaded sorbent was under pH 5, the following extraction fluid was prepared : 5.7 mL glacial acetic acid 64.3 mL 1N NaOH to 930 m L of ultrapure wate r. Ten ml of this extraction fluid was applied to 0.5 g Hg loaded carbon and rotated at 30 rpm for 18 h. After the elapsed contact time, the carbon was separated from aqueous solution by vacuum filtration and the pH of the extract was obtained before processing for Hg quantification. Aqueous Mercury Removal Labware P reparation All labware used in adsorption experiments was prepared by soaking for a minimum of 2 h in 20% HNO 3 and subsequen tly rinsing with ultrapure water a minimum of three times before air drying Vessel blanks were performed on each batch of cleaned glassware to ensure labware was free from residual mercury contamination by exposing randomly selected labware items to a 20% HNO 3 solution for 5 min and processing as sample for analysis. Mercury Quantification Methods C old vapor atomic absorption (CVAA) spectrometry is often used to quantify aqueous Hg concentrations due to its ease of use, rapidity, selectivity, and accuracy compared to other technologies [3] CVAA has a detection limit of 0.1 fluorescence spectrometry is used to reach ng/L detection limits. The EPA has developed several standardized methods associated with this technology. T he total mercury is determined for each aqueous sample by reducing all Hg species present to Hg(0) with SnCl 2 before transporting the vapor into the path of radiation fr om a cathode ray tube (Figure 3 4 ). The ground state of elemental me rcury

PAGE 62

62 atoms absorb radiation from the lamp in proportion to the concentration. The reduced signal reaching the detector is recorded. This process is based on the Beer Lambert Law. In this study, t otal aqueous Hg concentrations were measured on a Teledyne Hydra Atomic Absorption Mercury Analyzer using EPA method 245.1 which uses a thermal digestion and SnCl 2 reduction technique. The EPA method has a detection range between 0 and 100 which may be extended based upon sample size, matrix characte ristics, operating conditions, and instrumentation configuration. The manufacturer stated instrumentation detection limit is 0.2 but the operating method detection limit (MDL) was determined to be 0.4 Within 24 h of collection, each sample was acidified to under pH 2 using 0.5 mL HNO 3 and 1mL H 2 SO 4 Standards were prepared with each run. According to EPA method 245.1, each sample was thermally digested prior to analysis using 3mL of 0.32 M KMnO 4 (Fisher Scientific), 1.6 mL of 0.18 M K 2 S 2 O 8 (Fish er Scientific), and 1.2 mL of NaCl hydroxylamine sulfa te solution (2.1M NaCl, 0.73M hydroxylamine sulfate) (Fisher Scientific). Test S tand The batch reactor contained a sealed Teflon mercury carbon contact chamber with 0.8 L/min headspace N 2 (g) flow throu gh an inlet/outlet port to an oxidizing purge trap (Figure 3 5 ) The oxidizing purge trap to capture volatilized Hg was prepared using 0.25 M KMnO 4 (Fisher Scientific) in 10% H 2 SO 4 (Fisher Scientific) solution. All Hg ( 0 ) experiments were performed in a glo ve bag under N 2 (g) flow. The carbons were applied as a slurry at a 1g/L dose to Hg spiked ultrapure water for a specified contact time after

PAGE 63

63 which the adsorbent was s eparated from solution mixed cellulose filter (Fisher Scientific). Hg Mass Balance An integral Hg mass balance verifies the Hg removal performance of an adsorbent. Based on published aqueous Hg(II) mass balances, acceptable mass balance closur e was determined to be within approximately 15% [115,116] This was achieved by quantifying the residual aqueous Hg, adsorbed Hg extracted from MPAC by HF digestion (or sequential chemical extraction where specified), and volatilized Hg captured in the KMnO 4 trap. A total digestion was applied to quantify total adsorbed Hg. This digestion was also applied to virgin carbons to determine trace levels of Hg contamination in the activated carb on from the raw source material. These trace levels of contamination were accounted for in the mass balance calculations. The HF digestion was performed using 400 L aqua regia (3:1, v/v concentrated HCl (J.T. Baker) to concentrated HNO 3 (Fisher Scientific )), 2 mL of concentrated HF (Acros Organics), and 20 mL of saturated H 3 BO 3 (Acros Organics). Batch Studies Identifying a contact time is essential in order to reach adsorption equilibrium during the isotherm assay. Based on the protocol described by Calgo n, a 1 g/L dose of carbon was applied to Hg solution for 0 180 min [69] Isotherm analysis is useful in evaluating the capacity of the carbon for adsorption of specific contaminants. Isotherm analysis was performed by applying varying weights of dried powdered activated carbon to constant volumes of Hg solution for the equilibrium contact time previously identified, after which samples were vacuum filtered

PAGE 64

64 Results were analyzed u sing both the Freundlich and Langmuir isotherms. MPAC adsorption experiments were performed at the pseudo equilibrium contact time of 180 min with a carbon dose of 1g/L and a mercury concentration of 100 g/L Surface modified carbon adsorption experiment s were performed with a contact time of 30 s due to the volatile nature of Hg(0). The carbon dose applied was lowered to 150mg/L as higher doses res ulted in nearly 100% removal for most carbons. Controlled by the solubility of Hg(0), Hg(0) doses ranged bet ween 40 and 60 g/L; Hg(II ) solution concentration was 50 g/L. Investigation of Adsorption Mechanisms Influence of pH and pCl Mercury speciation in the presence of a known chloride concentration at given pH values is well understood. By manipulating these variables, Hg speciation can be controlled and binding mechanisms can be predicted. This concept can be used to investigate the efficiency of the SCE for predicting binding mechanisms based on extraction fraction. Hg speciation at the identified pH and pCl values was p redicted using Visual MINTEQ. The pCl was adjusted using NaCl while pH was adjusted using 0.25M H 2 SO 4 or 0.25 NaOH. Ionic strength calculated using the Debye Huckel equation, was held constant using Na 2 SO 4 The optimal pH and pCl for Hg(II) adsorption by MPAC was found using a fractional factorial approach by manipulating the pH (2, 6, and 10) and pCl (12, 6, and 0). Sequential chemical extraction Sequential chemical extractions (SCEs), first becoming popular in the 1980s, are used to provide information regarding the speciation, bioavailability, and mobility of

PAGE 65

65 metals by applying selective extractants with increasing strength to the same sample aliquot [117,118] The goal is to convert the boun d metal into a soluble form using specific extractants as to elucidate the binding mechanism and speciation. Once extracted, the metals are analyzed by the appropriate analytical technique. If the chemistry of the adsorbate is understood, extractants can b e meticulously applied to elucidating the operating binding mechanisms. When designing an SCE, major factors to consider include the chemical nature of the extractant, efficiency and selectivity of extractants, matrix effects such as re adsorption, order o f extractants, and the nature of the targeted metal [118,119] Problems with sequential extractions include selectivity less than 100%, control of reaction conditions, and inconsistencies betw een extraction protocols [118,120] It is also possible that in removing a fraction of the metal ion, the ion may then redistribute itself among the remaining phases (phase transformation) [121] Several factors have been experimentally determined to have significant affects on the results. Shaking speed should be maintained at 30 rpm [99] The temperature during extrac tions should remain at 20C 2C [99,120] Extraction times should reach 18 4 hours [99,120] Samples should be dried until constant weight a nd the sample slurries should be formulated with a 1:100 solid to extractant ratio [99,117,120,122] When properly designed, an SCE can reach detection levels as [120] In this study, s equential chemical extraction was performed by applying the following extractants with increasing strength to the same sample aliquot : water soluble, ion exchangeable, surface prec ipitated, surface bound, poorly reducible (iron associated), and residual. The water soluble fraction used ultrapure water to target t he

PAGE 66

66 labile non adsorbed Hg within the pores The ion exchangeable fraction used 1M ammonium nitrate to targe t weakly electr ostatically adsorbed Hg. Ammonium nitrate was selected b ecause nitrate will not complex with mercury; therefore, any mercury mobilized will be due to cation exchange with ammonium on the carbon surface. The surface precipitated fraction was targeted using 0.11M a cetic acid A t higher pH values, acetic acid has been shown to have little to no effect on organic carbon or free iron concentrations [122] The surface bound mercury was targeted using 0,1M 2,3 meso dimercaptos uccinic acid ( DMSA ) a chelating agent that sequester s Hg The poorly reducible fraction used 0.128M diothinite, 0.3M citrate, and 1M bicarbonate (DCB) to target the metals associated with the iron oxides by reducing Fe 3+ to the more soluble Fe 2+ form, thu s releasing chemisorbed Hg [118] Due to potential metal impurities in this reagent, a reagent blank was performed to prevent Hg contamination [122] Residual Hg was quantifie d in the final fraction using aqua regia, HF, and H 3 BO 3 as described above for total digestion. Data Analysis All experiments were performed in triplicate and average values reported. All replicate data falls within the 95% confidence interval. Error bars represent the standard error of the mean. The Box Behnken experimental design for response surface methodology was used to identify the optimal MPAC for Hg removal according to the three variables specified. The design required 17 total runs with 12 exper iments and 5 replicates of the center point. The experimental design was analyzed using Design Expert software (version 6.0.5). Visual MINTEQ 2.61, a chemical equilibrium model, was used to calculate metal speciation, complexation reactions, and solubility equilibria.

PAGE 67

67 Linear regression and ANOVA analyses were performed using the statistical software R version 2.14.1

PAGE 68

68 Table 3 1. Surface Area Calculation Methods by P/P 0 range u tilized [108] P/P 0 r ange Me chanism Calculation m ethod <0.1 Micropore filling DFT, HK, SF, DA, DR 0.01 0.1 Sub monolayer formation DR, MP 0.05 0.3 Monolayer formation BET, Langmuir, DR, MP >0.2 Multilayer formation t plot, alpha s, FHH, MP >0.35 Capillary condensation BJH, D H, Fractal FHH, NK 0.1 0.5 Capillary condensation in M 41 S type materials DFT, BJH, DH

PAGE 69

69 Figure 3 1. Common acidic surface oxygen groups on activated carbon with pH above the pHpzc (left to right: carboxyl, phenol, carbonyl) Figur e 3 2 Vibrating Sample Magnetometer Schematic Figure 3 3. Hysteresis loop resulting from VSM ana lysis

PAGE 70

70 Figure 3 4. Cold Vapor Atomic Absorption Spectroscopy Schematic Figure 3 5 Schematic of batch adsorption test stand

PAGE 71

71 CHAPTER 4 CHARACTERIZATION OF MODIFIED ACTIVATED C ARBON Many carbon modifications are discussed in the literature. The application s of the materials p repared in this work are unique as i ron impregnation for Hg adsorption has not been investigate d nor has the influence of surface oxygen functionality on aqueous eleme ntal Hg adsorption been studied In order to best understand the application of these materials to Hg laden wastewaters, knowledge of the material characteristics is necessary The fo llowing discussion addresses the characterization of the carbon adsorbents in terms of porosity, surface charge, crystalline nature, elemental composition magnetic characteristics, and sorbent stability. The objectives were to 1) s ynthesize magnetic carbo ns that are at least 95% recovera ble through magnetic separation 2) i ncrease acidic C( O) on activated carbon surfaces with minimal pore degradation and 3) c haracterize carbons with techniques including nitrogen adsorption desorption, X ray diffraction, p oint of zero charge, and total acidity. MPAC Characterization Porosity Nitrogen a dsorption desorption isotherms for virgin WPH and CASPF carbons are shown in Figure 4 1. The isotherms are Type I, common for microporous substances such as activated carbo n Both carbons display H4 hysteresis loops, indicative of a microporous characteristic with slit shaped pores I sotherms were analyzed to produce BET surface area, average pore diameter, and total pore volume data. The process of iron impregnation was ex pected to reduce the available surface area relative to the virgin activated carbon due to the minimal

PAGE 72

72 surface area of the iron oxides (1.9 m 2 /g). As expected, the 1:1 C:Fe resu lted in a ~ 50% reduction of surface area relative to the raw WPH carbon while the 2:1 and 3:1 C:Fe showed surface areas reduced by the expected 33% and ~ 25%, respectively (Table 4 1). The replicates of each average porosity characteristic reported below ha ve a coefficient of variation (CV) of approximately 7%. With the purpose of converting amorphous iron oxides to ferromagnetic magnetite or mag hemite p ortions of the synthesized MPAC were subjected to thermal oxidation for varying temperatures and durations (250 C, 350 C, and 450 C for 3 h and 6 h). Figure 4 2 demonstrates that thermal oxidation of a 1:1 C:Fe MPAC at 250 C had little effect on p orosity (surface area, pore volume, and pore size) regardless of duration while temperatures of 350 C and 450 C increasingly reduced the surface area and pore volume while increasing the average pore size. This adverse degradation of porosity is likely due to decompositi on of surface oxygen groups and, to a greater extent, gasification of carbon at temperatures over approximately 400 C [22] The 1:1, 2:1, and 3:1 C:Fe MPACs exhibited similar BJH pore size distributions (PSD) to the virgin carbon as calculated from nitrogen adsorption isotherms (Figure 4 3 ) Thermal oxidation of the samples caused pore degradation/collapse, demonstrated by the reduction in cumulative pore volume and slight skewing of the pore volume to h igher pore diameters, seen in the highly oxidized sample ( 3:1 450 3 h). PSD replicates indicated no greater than a 5.5% CV Magnetic Characteristics X r ay d iffraction M aghemite is the most likely iron oxide produced in the synthesis of MPAC but other iron oxides have the potential to precipitate onto the carbon surface. An XRD

PAGE 73

73 analysis was performed to identify the iron oxide species present on the carbon surface. The raw 3:1 C : Fe as well as the oxidized 3:1 C:Fe samples were analyzed (Figure 4 4 ). All sam ples investigated displayed peaks with positions and relative intensities that match well with those for maghemite c (39 1346) and maghemite q (25 1402). The samples exposed to 350 C and 450 C exhibited additional peaks identified as hematite (33 0664), a non magnetic iron oxide All major diffraction peaks were associated with the iron oxides identified. Several specific features of interest are present in these XRD patterns. The raw 3:1 carbon exhibits an amorphous characteristic from roughly 2 15 to 34 and from 2 40 to 50. As the oxidation temperature increased, this amorphous characteristic was reduced and the crystalline structure enhanced, seen in the progressively flattened baseline and the increased sharpness of nearby peaks. It was hypothesized that the thermal oxidation would force amorphous iron oxides to magnetite or maghemite. Although there is no overwhelming evidence of this effect seen in the XRD patterns two unique aspects in the patterns suggest this change may occur. Maghe mite c is known to exhibit small diffraction peaks at 2 32.152 and 44.743 which are present only in the 450 C carbon. It could be argued that the emergence of minor peaks at 2 23.791 and 26.125 in the samples that underwent higher thermal oxidation temperatures is evidence of this change but these peaks could have been present in the original 3:1 sample and only became clear due to the progressively increased crystallinity and thus flattened baseline In several locations, there was clear evidence of the formation of hematite through the thermal oxidation process Hematite formation was expected due to the conversion

PAGE 74

74 of maghemite to hematite in the range from 3 50 C to 750 C depending upon the grain size, degree of oxidation, and defects in the crystal lattice [123] Notice the development of a hematite peak as thermal oxidation temperature increased at 2 24.158, 33.181, 40.890, 49.523, and 64.049. Transformation to hematite may be indicated at approximately 35.5 but interpretation is unclear due to overlap ping peaks of hematite the maghemite k for carbons exposed to 450C for 6 h. Due to the synthesis technique, magnetite (19 0629) may be present on the carbon surface. Distinguishing magnetite from maghemite XRD patterns can be challenging o the location of maghemite c, maghemite q, and hematite peaks. strongest peak, strong magnetite peak that would stand apart from the other iron oxides known to be ed. As magnetite slowly oxidizes over time under atmospheric oxygen exposure, it is possible that a freshly synthesized sample may display a magnetite peak at this location. Vibrating sample m agnetometry The magnetic properties of MPAC were tested b y vibr ating sample magnetometry as shown in T able 4 2 To enable manipulation using conventional magnets, the sorbent must exhibit sufficient saturation magnetization (Ms) of at least 4.5 Am 2 /kg and a remanence (residual magnetization Mr ) high enough to allow f or recapture but not so high as to cause clumping [62] All MPACs tested sh owed sufficiently strong saturation

PAGE 75

75 magnetism to allow for recapture. Thermal oxidation at 250C and 350C slightly increased Ms for the 6h dura tion. The 450C oxidation dramatically increased Ms for the 3h duration but declined for the 6h duration, likely due to the conversion of maghemite and magnetite to non magnetic hematite. Remanence values tended to increase at all thermal oxidation tempera tures with 450C resulting in the highest Mr values of the samples tested No samples exhibited excessi ve clumping upon water contact. Magnetic a dsorbent r ecovery MPAC was retrieved from the aqueous solution via magnetic solid phase extraction. With a coef f icient of variation of only 4.0 %, the C:Fe did not significantly influence the recoverability of the adsorbent (Table 4 3). Sorbent recovery slightly decrease d as the thermal oxidation temperature increased. The relative percent difference between the raw MPAC and the sample exposed to 450 C for 6 h was 15.9% and 27.3% for the 1:1 and 3:1 C:Fe, respectively. Although VSM and XRD data indicated improved magnetic qualities with thermal oxidation, this improvement did not translate to improved sorbent recover y. The 3:1 MPAC meets the objective of being 95% recoverable. The average ad sorbent mass balance closure was 92.3% and ranged from 88.1% to 96.5% with a CV of 9.5%. Adsorbent Stability : Iron Typically, iron is not a concern from a regulatory standpoint an d is commonly a constituent of industrial wastewaters. At unadjusted pH, t he MPAC adsorbent is quite stable and Fe effluent concentrations fell below the detection limit (0.022 mg/L total Fe) for all contact times investigated between 0.5 180 min. Because Fe leaching is sensitive to matrix pH, the stability of the 3:1 MPAC adsorbent at extreme pH values was determined. At pH 2, 1.2 mg Fe leached per gram

PAGE 76

76 MPAC. The leaching did not cause discoloration of the water. At pH 10, the Fe effluent concentrations f ell below the detection limit The leaching of Fe at lower pH values did not impact recoverability, with 98% of the MPAC being recovered. C(O) Modified Carbon Characterization Porosity Nitrogen adsorption desorption isotherms were analyzed to produce BET surface area, average pore diameter, and total pore volume data (Table 4 4) Consistent with literature, the H 2 (g) stripping process did not negatively influence porosity [49,52,56,106] Literature indicated the potential for damage to porosity through the wet chemical oxidation process due to either pore damage or the formation of pore blocking humic substances [107] N itric acid modified samples exhibited progressive porosity damage with increasing concentration. Conversely, sulfuric acid and sodium hydroxide modifications did not result in damage to porosity. The humics removal wash did not significantly influence the adsorbent porosity (CV of only 0.4%) and thus was not applied to carbon samples. The modified carbons and H 2 (g) stripped carbons exhibited similar BJH pore size distributions to the virgin WPH carbon (Figure 4 5). PSD replicates indicated no greater than a 3.2% CV. The t reatment of CASPF carbon impacted porosity similarly to the WPH modification. Surface Oxygen Functionality With a basic pH pzc and a relatively low oxygen content, the total acidity of WPH carbon was expectedly low, at only 85 meq/g [0. 05] NaOH (Table 4 4 ). On the contrary, CASPF carbon displayed an acidic pH pzc and higher oxygen content, resulting in greater total acidity relative to WPH The control carbon, stripped of nearly all C(O), demonstrated a very basic pH pzc and a total acid ity near zero.

PAGE 77

77 In the modification of WPH with nitric acid as the acid concentration increased, the oxygen content and total acidity increased while the pH pzc fell. Relative to 10M HNO 3 treatment, the 10M H 2 SO 4 treatment was less effective at adding surf ace oxygen groups, seen in the reduced oxygen content and the lower total acidity I nterestingly the pH pzc of SAC 10M was slightly lower than NAC 10M for both WPH and CASPF carbons Also note the lack of response in SAC and SHAC carbons to acid con centr ation; t otal acidity, pH pzc and oxygen content remain ed relatively stable The CASPF modified carbons interestingly showed an increase in pH pzc upon modification though the values remained very acidic. Modification with 10 M HNO 3 resulted in a n oxygen c ontent of 21.2% and a high total acidity of 425 meq/g, a 37% increase from raw CASPF carbon Modification with 10 M H 2 SO 4 actually reduced the oxygen content and total acidity relative to the raw CASPF

PAGE 78

78 Table 4 1. Porosity of various MPACs Sample Su rface a rea (m 2 /g) Mean pore s ize () Pore v olume (cm 3 /g) 1:1 579 24.2 0.333 2:1 709 24.3 0.430 3:1 790 23.2 0.457 3:1 450 3h 46 82.4 0.124 Table 4 2. Magnetic characteristics of various MPACs Sample Hc (mT) Mr (Am 2 /kg) Ms (Am 2 /kg) 3:1 2.75 1. 3 10.92 3:1 250 3h 3.17 1.2 10.90 3:1 250 6h 7.10 2.0 11.23 3:1 350 3h 3.65 1.6 10.89 3:1 350 6h 11.96 2.7 11.98 3:1 450 3h 9.28 3.9 19.43 3:1 450 6h 7.29 3.2 14.99 Table 4 3. Magnetic solid phase extraction results for various MPACs Sample Sorbe nt r ecovery (%) 1:1 92.6 2:1 88.1 3:1 95.0 1:1 250 6h 85. 4 1:1 350 6h 76.5 1:1 450 6h 75.6 3:1 250 6h 87.8 3:1 350 6h 81.5 3:1 450 6h 72.2

PAGE 79

79 Table 4 4. Characterization of various C(O) modified carbons Sample Surface area (m 2 /g) Pore s ize () P ore v olume (cm 3 /g) % N % H % O % Ash pH pzc Total acidity (meq/g [0.05] NaOH) Total basicity (meq/g [0.05] HCl) WPH 1020 11.2 0.55 0.0% 0.2% 6.9% 7.4% 8.36 85 136 ACH 1044 15.0 0. 58 0.0% 0.3% 1.9% 8.1% 10.10 11 212 WPH NAC 1M 991 11.2 0.58 0.0% 0.2% 8.4% 6.3% 6.56 94 99 WPH NAC 5M 978 11.0 0.5 4 0.6% 0.6% 15.8% 6.4% 5.02 139 107 WPH NAC 10M 878 11.0 0. 49 0.5% 1.1% 19.6% 6.6% 3.99 231 91 WPH SAC 1M 975 11.1 0. 54 0.4% 0.3% 6.1% 6.3% 3.78 112 48 WPH SAC 5M 989 11.1 0.55 0.3% 0.3% 5.9% 6.4% 3.90 112 33 WP H SAC 10M 975 11.2 0.5 5 0.3% 0.4% 10.5% 5.7% 3.36 118 26 WPH SHAC 1M 994 11.1 0.54 0.4% 0.2% 3.8% 6.8% 7.37 60 105 WPH SHAC 5M 1000 11.1 0.55 0.4% 0.2% 5.1% 6.8% 7.25 64 147 WPH SHAC 10M 1001 11.1 0.54 0.4% 0.2% 6.3% 6.2% 7.13 70 156 CASPF 1201 11.0 0. 91 0.0% 2.3% 19.8% 0.7% 1.93 293 0 CASPF NAC 10M 817 11.7 0.51 2.4% 1.9% 21.2 % 0.9 % 2.83 425 10 CASPF SAC 10M 1269 11.8 0.93 0.1% 2.3% 13.8% 0.4% 2.77 278 2

PAGE 80

80 A. B. Figure 4 1. MPAC nitrogen adsorption desorption isotherms A) WPH B) CASPF 0 200 400 600 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Volume (cm 3 /g) P/P 0 0 200 400 600 800 0.0 0.2 0.4 0.6 0.8 1.0 Volume (cm 3 /g) P/P 0

PAGE 81

81 A. B. C. Figure 4 2. Effect of thermal oxidation on porosity of 1:1 C:Fe. A) surface area, B) pore size, and C) pore volume 0 200 400 600 800 0 2 4 6 Surface Area (m 2 /g) Time (h) 250C 350C 450C 0 40 80 120 160 200 0 2 4 6 Pore Size () Time (h) 250C 350C 450C 0 0.1 0.2 0.3 0.4 0 2 4 6 Pore Volume (cm 3 /g) Time (h) 250C 350C 450C

PAGE 82

82 Figure 4 3. BJH pore size distribution of select MPACs Figure 4 4. Powder XRD patterns of MPAC particles before and after the rmal oxidation 0.00 0.02 0.04 0.06 0.08 0.10 0.12 10 100 Cumulative Pore Volume (cm 3 /g) Pore Diameter () WPH 1:1 C:Fe 2:1 C:Fe 3:1 C:Fe 3:1-450-3h

PAGE 83

83 Figure 4 5. BJH pore size distribution of select C(O) modified carbons 0 0.02 0.04 0.06 0.08 0.1 0 10 20 30 40 50 60 70 80 Cumulative Pore Volume (cm 3 /g) Pore Diameter () WPH NAC-1M ACH WPH

PAGE 84

84 CHAPTER 5 TRACE LEVEL AQUEOUS MERCURY REMOVAL USIN G MODIFIED ACTIVATED CARBON The following discussion investigates the adsorption of H g using commercially availa ble activated carbon modified by several approaches. Previous literature has investigated the application of activated carbon to aqueous Hg adsorption [ 22,72,74,75] However, the literature has not addressed aqueous Hg removal using a magnetic adsorbent The literature is scarce regarding aqueous Hg adsorption using an activated carbon with enhanced surface oxygen functionality [36,75,77] I n fact, little is known regarding aqueous adsorption of Hg(0) or the role of surface oxygen groups in its adsorption. The objective of this study was to d etermine which experiment al conditions yield ed the highest removal of aqueous Hg. MPAC Results Controls Prior to performing Hg adsorption experiments, it was imperative to perform control runs. To verify that the batch reactor was free from residual Hg contamination, an air blank was performed periodically. This was accomplished by running the test stand with only ultrapure water, in the absence of carb on and mercury. Hg levels were quantified in the mercury carbon contact chamber an d volatilization trap. A sorbent blank, determine d via aqua regia and hydrofluoric acid digestion, identified t race levels of Hg in the adsorbent averaging 0.125 g Hg/g virgin WPH Bituminous coal is the raw material used in the production of WPH activated carbon; coal is known to contain trace levels of Hg. These values were considered in the mass balance calculations. A background analysis was performed by run ning Hg spiked ultrapure water through the batch reactor in the absence of carbon. The analysis revealed the following: low levels

PAGE 85

85 of Hg volatilization occurred in the absence of carbon, quantifiable Hg residues (9% total Hg) formed in test stand labware n ecessitating a n HNO 3 rinse to fully quantify the residual Hg, and an average 6% Hg was fugitive (Figure 5 1) The fugitive Hg was likely due to mass and volume measurement errors amplified by the small scale of the experiment. Pseudo Equilibrium Adsorption The amount of adsorption was calculated based on the difference before and after adsorption according to the following equation: = ( 5 1 ) where q e is the equilibrium adsorption capacity of Hg(II) (m g/g), C 0 is the initial concentration of Hg(II) (mg/ L), V is the volume of the Hg(II) solution (mL), and m is the adsorbent dosage (mg). Contact t ime A 1 g/L dose of 3:1 MPAC was applied to 100 solution to study the effect of contact time on the a dsorption of Hg(II) shown in Figure 5 2 The initial adsorption rate was rapid with over 90% of the Hg(II) removed during the first minute of contact. This was followed by a much slower adsorption rate, reaching pseudo equilibrium at 120 min. Before carbon addition, the aqueous solution pH averaged 4. 7 with a percentage change in the pH of 6.5% in the first 30 seconds of contact. Beyond the first 30 s, the pH stabilized to an average of 6.2. B atch t esting of s ynthetic w aters Effect of iron loading Because of its influence on the adsorbent surface ch aracteristics, it was possible for iron loading to impact the removal of Hg from 100

PAGE 86

86 g/L aqueous solution (Figure 5 3 ). 1:1 and 2:1 C:Fe performed similarly with a CV of 2.78% The 3:1 C:Fe exhibited the best Hg removal. As the iron loading influenced porosity (Table 4 1) the effect of surface area changes due to both Fe loading and thermal oxidation is discussed below. Effect of thermal oxidation. Figure 5 4 demonstrates that the thermal oxidation temperatur es investigated in this study minimally influence d the aqueous mercury removal capabilities of MPAC despite the pore damage incurred at o xidation temperatures over 250 C. For each C:Fe, the CV between the raw samples and oxidized samples only varied betwee n 0% and 4.5%. At all oxidation temperatures, the 3:1 MPAC achieved the highest mercur y removal. The 1:1 and 2:1 C:Fe performed similarly for Hg removal, wi th CV values under 2.8% at each temperature. Effect of surface area. This work does not show a stro ng correlation between surface area alone and mercury removal The experimental data best fit a three variable model with an adjusted R 2 of 0.464, identifying surface area, pore volume, and point of zero charge as the variables influencing Hg(II) adsorptio n in the system. An ANOVA test identified the sums of squares for the surface area, pore volume, and pH pzc of 113.1, 2.1, and 341.1, respectively. The pH pzc is the primary variable influencing results. As the R 2 is not close to 1, there are likely other va riables influencing the efficiency of Hg(II) removal; Hg adsorption can be influenced by other sorbent characteristics such as surface oxygen functionality [35,76] Mercury m ass balance. The ave ra ge Hg mass balance closure for experiments was 99.5% with a standard deviation of 8.8%. The mass balance closures ranged from 88.3% to 116.8% but many runs did not fall within the 95% confidence intervals; the

PAGE 87

87 observed distribution fits a random distribu tion curve (Figure 5 5) The challenge in obtaining mass balance closure was likely due to HF extraction inefficiency in quantifying the adsorbed Hg, mechanical loss of C resulting in lower Hg masses extracted in the HF digestion, and volumetric measuremen t errors amplified due to the small scale of the experiment. The mass balance for Hg adsorption onto 3:1 C:Fe MPAC is presented in Figure 5 6 At unadjusted pH, approximately 9 5 % of the Hg was removed from aqueous solution with 2% volatilized and 8 7 % adsor bed while 6 % remained fugitive. Optimization. Box Behnken fractional factorial design was used to identify the optimal MPAC for both Hg removal and MPAC recovery (equally weighted in the experimental design) according t o the following variables: C:Fe and t hermal oxidation temperature and time. The following criteria were used in the numerical optimization: C:Fe within range, minimized oxidation temperature and time, maximized magnetic recovery, and maximized Hg removal. Oxidation parameters were minimized t o reduce the cost of MPAC synthesis. Based on these criteria, the optimal synthesis variables of 3:1 C:Fe with no thermal oxidation would achieve a predicted sorbent recovery of 92.5% (8.3%) and Hg removal of 96.3% (9%). Adsorption Isotherms The effect of the dose of MPAC on Hg(II) adsorption was investigated by varying the MPAC dose from 0.5 to 10 g/L (Figure 5 7 ). The Langmuir equation is derived from the assumption of monolayer adsorption on specific homogenous sites, while the Freundlich model repres ents physical adsorption on heterogeneous surfaces. The good fitting results of both models seen in Figure 5 7A, implied that both chemisorption and physisorption mechanisms took place in the adsorption s ystem The term 1/n was

PAGE 88

88 between 0 and 1, indicating heterogeneity of the MPAC and affinity of the adsorbate for the adsorbent, resulting in favorable adsorption of Hg(II) by the 3 :1 C:Fe MPAC [124] The dimensionless Langmuir constant separation factor, R L given as R L 1/(1+BC o ) where C o is the initial concentration and b is the Langmuir constant. The R L indicates favorable adsorption between 0 and 1 while R L >1 indicates unfavorable adsorption, R L =1 is linear, and R L = 0 indicates irreversible adsorption. The value of R L was found to be 0.002 indicating favorable and nearly irreversible adsorption. Kinetics Studies Three kinetic models have been proposed for Hg(II) adsorption by MPAC: 1) intraparticle diffusion [125] 2) pseudo first order kinetic model [126] and 3) pseudo second order kinetic model [127] The intraparticle diffusion model can be described according to the Weber and Morris equation as: q t = k i d t 1/2 + C (5 2) where k id is the intraparticle diffusion rate constant ( g/ g min 1/2 ), C is the y intercept ( g/g), and q t is the adsorption capacity of Hg(II ) ( g/ g) at time t (min). The plot of q t vs t 1/2 is not linear and does not pass throu gh the origin, therefore intraparticle diffusion is not the sole rate limiting step (Figure 5 8). Multiple rate limiting steps might take place in this system. The pseudo first order rate law was integrated to a linear rate law (Equation 5 3) where k is t he equilibrium rate constant (1/min). The pseudo second order model (Equation 5 4) was expressed where k 2 is the pseudo second order rate constant of adsorption (g/ g min). The applicability of the se model s w as assessed by comparing the R 2 values of the linear plot of log(q e q t ) vs t and (t/q t ) vs t, respectively (Figure 5

PAGE 89

89 8). The data fit the pseudo second order model with an R 2 of 0.9999 indicating that adsorp tion was due to chemisorption [128] The p value for the slope was 7.22 x 10 13 log (q e q t ) = log q e (5 3) (5 4) Adsorbent Stability: Hg T he mobility of Hg once adsorbed to the 3:1 C:Fe was investigated using the standardized TCLP test. The effluent Hg concentrations remained under the regulated limit of 200 g/L until reaching an Hg loading ratio of 1000 g Hg :1 g MPAC where the effluent concentration was found to be double the allowable limit for sanitary landfill disposal (Figure 5 9) C(O) Results Controls As previously described, a n air blank was perfo rmed to verify that the batch reactor was free from residual Hg contamination. Reagent blanks verified all solutions and ultrapure water were free from trace levels of mercury. A sorbent blank identified trace levels of Hg present in the virgin carbons wit h WPH containing 0.125 g Hg/g and CASPF containing 0.071 g Hg/g Through a procedural blank, this residual Hg was not shown to influence aqueous or volatilized Hg levels. Due to the volatile nature of Hg(0), it was important to understand the rate of volatilization in the abs ence of carbon. Figure 5 10 demonstrates that, in the absence of carbon, nearly 50% of the Hg volatilizes after 30 s with o nly 1.2 % of the Hg(0) fugitive. Because of this high rate of volatilization, Hg(0) adsorption experiments were performed at a 30 s contact time rather than at pseudo equilibrium

PAGE 90

90 The HF/H 3 BO 3 total digestion employed when quantifying adsorbed Hg for the MPAC carbons did not produce replicable results with the C(O) modified carbons. For these experiments, mass balance was determined by assuming Hg that did not either volatilize or remain in aqueous solution was adsorbed. Batch T esting of S ynthetic W aters Effect of C(O) on Hg adsorption Due to the multitude of variables that can influence adsorption, the influence of one specific variab le requires regression analysis. A t test with a significance level of = 0.05 revealed oxygen content as a good regression parameter for Hg( II) adsorption (p value = 0.00328 ) but the same does not hold true for Hg(0) adsorption (p value = 0.28850 ) This could be due, in part, to water cluster formation. Acidic C(O) groups tend to adsorb water by hydrogen bonding and dispersion forces followed by clustering of additional water molecules at these adsorption sites [129] These water clusters can block adsorbate access to the activated carbon porosity [130] Studies have shown a drop in adsorption capacity of organic pollutants with an increase in the amount of C(O) groups, evidence of the water adsorption effect [131,132] As Hg(0) is uncharged, it is possible that there are not sufficient attractive forces to overcome the pore blocking effect due to water cluster formation. Although the speciation in the Hg(II) system exists primarily as the uncharged Hg (OH) 2 the carbon particles serve as a nucleation point for the precipitation of solid Hg(OH) 2 which may be able to overcome the pore blocking of the water clusters. As seen in Figure 5 11 the virgin WPH carbon performed fairly well for Hg(II) removal but when applied to Hg(0) adsorption resulted in relatively high levels of volatilization. SAC and SHAC carbons achieved high levels of Hg(0) adsorption with

PAGE 91

91 minimal losses through volatilization. The annealed carbon, with the lowest oxygen content, displa yed the lowest Hg (0) and Hg(II) removal. Effect of porosity on Hg adsorption Hg(II). A t test revealed that surface area alone was poorly correlated to Hg(II) removal with an R 2 value of 0.004 An ANOVA test on the influence of surface area, pore size, and pore volume on Hg(II) removal revealed that pore volume had significantly m ore influence than surface area and pore size, with a sums of squares value of 569.75 compared to 5.31 and 97.47, respectively. The two variable model that best fits the Hg(II) removal data indicates oxygen content and pH pzc as important variables, resulting in an R 2 value of 0.499. An ANOVA test indicated oxygen content to be the primary variable influencing adsorption with sums of squares of 666.82 while the pHpzc sums of squa res was only 62.06. Hg(0). Surface area also poorly correlated to Hg(0) removal with an R 2 value of 0.09 3. A t test analysis of the influence of surface area, pore size, and pore volume on Hg(0) removal resulted in a negative adjusted R 2 making an ANOVA test impractical. The best regression model to fit the Hg(0) data indicates that surface area, pore volume, surface oxygen functionality, and the point of zero charge as important variables, resulting in an R 2 value of 0.5886. The t test identified the pH pzc as a good regression parameter (>95% confidence). An ANOVA test indicated the point of zero charge as the primary variable influencing adsorption with sums of squares of 1041.72. As no model using the measured variables achieved a strong R 2 value, it i s possible that an unquantified variable was influenc ing the results of both Hg(II) and Hg(0) adsorption.

PAGE 92

92 Adsorption Isotherms The effect of the dose of C(O) modified carbons on Hg(II) adsorption was investigated by varying the carbon dose (Figure 5 1 2 ). The good fitting of the experimental data to both models seen in Figure 5 12A, implied that both chemisorption and physisorption mechanisms were occurring in the adsorption system. The Freundlich term 1/n was 0.86, indicating heterogeneity of the carbon s urface and affinity of the adsorbate for the adsorbent, resulting in favorable adsorption of Hg(II) by NAC 1M. The value of R L was found to be 0, indicating irreversible adsorption. Kinetic Studies As with MPAC, three kinetic models were investigated f or Hg(II) adsorption by NAC 1M (Figure 5 1 3 ). The plot of q t vs t 1/2 is fairly linear for ACH, NAC 1M, SAC 1M, and SHAC 1M, with R 2 values of 0.7596, 0.7892, 0.8938, and 0.8322, respectively. The linearity of the experimental data for NAC 1M, SAC 1M, and S HAC 1M indicated that i ntraparticle diffusion may be a rate limiting step in these systems. The ACH carbon demonstrated immediate uptake of Hg(II) at a much higher capacity than the other carbons, likely due to the absence of surface oxygen groups and ther efore an absence of water clusters on the adsorbent surface. Interestingly, the adsorption capacity decreased as time progressed This may be due to competitive adsorption between Hg(II) and H 2 O for the available adsorption sites on the carbon surface wher e equilibrium with water proceeds slower, thus the decrease in adsorption capacity as equilibrium is approached. The applicability of the pseudo first order and pseudo second order models was assessed by comparing the R 2 values of the linear plots. Due to the negative slope of ACH, it could not be assessed for pseudo first order kinetics. NAC 1M, SAC 1M, and

PAGE 93

93 SHAC 1M all fit the pseudo first order model with R 2 values of 0.9188, 0.9822, and 0.8975, respectively. Even so, the data showed a stronger fit with t he pseudo second order kinetic model, with R 2 values higher than 0.99, indicating that chemisorption played a large role in Hg(II) removal. The p values for the slopes were very low, ranging from 1.2 x 10 5 to 6.7 x 10 7 Adsorbent Stability The mobility of Hg(II) once adsorbed to the surface modified carbons was investigated using the standardized TCLP te st (Table 5 1 ). After loading the carbons with 100 g Hg/ g C, the effluent remained under the regulated limit of 200 g/L for all carbons tested.

PAGE 94

94 Tab le 5 1. Hg leaching from various carbons under landfill conditions Sample Effluent ( g Hg) Hg leaching ( g Hg/ g C) ACH 8.4 16.7 WPH 6.9 13.8 CASPF 31.2 62.2 NAC 10M 36.1 72.2 SAC 10M 54.1 107.8 SHAC 10M 11.4 22.7

PAGE 95

95 Figure 5 1. Backgro und Hg (II) mass balance Figure 5 2. Effect of contact time on Hg(II) adsorption (3:1 C:Fe, 1g/L) 82.4% 9.0% 1.2% 6.5% In Solution Residual Volatilized Fugitive 88% 90% 92% 94% 96% 98% 100% 0 50 100 150 200 % Hg Removal Contact Time (min)

PAGE 96

96 Figure 5 3. Effect of iron loading on pseudo equilibrium adsorption of 100 g/L Hg(II) Figure 5 4. Influence of 3h oxidation at 250 C and 450 C on aqueous Hg(II) removal 75 80 85 90 1:1 2:1 3:1 q e ( g Hg/g MPAC) C:Fe 50% 60% 70% 80% 90% 100% 0 100 200 300 400 500 % Hg Removal Temperature ( C) 3:1 C:Fe 2:1 C:Fe 1:1 C:Fe

PAGE 97

97 Figure 5 5. Mass balance distribution Figure 5 6. Hg mass balance for 3:1 C:Fe adsorbent 0% 20% 40% 60% 80% 100% 120% 140% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Mass Balance Closure Run # 87% 2% 5% 6% Adsorbed Volatilized In Solution Fugitive

PAGE 98

98 A. B. C Figure 5 7. Hg(II) adsorption isotherm onto 3:1 MPAC. A) Nonlinearized adsorption isotherm B) Freundlich model and C ) Langmuir model y = 72.458x 53.407 R = 0.9365 y = 1.7861x + 1.2236 R = 0.9906 0 10 20 30 40 50 60 70 80 0.5 1 1.5 2 q e ( g/g) C e (g/L) Freundlich Langmuir log q e = 3.3171 log C e + 1.1545 R = 0.8344 1/n = 0.3015, K F = 14.2725 0 0.5 1 1.5 2 2.5 -0.20 -0.10 0.00 0.10 0.20 0.30 log q e log C e 1/q e = 0.2085 1/C e 0.1175 R = 0.9351 b = 8.51064, a = 0.56355 0 0.05 0.1 0.15 0 0.25 0.5 0.75 1 1.25 1.5 1/q e 1/C e

PAGE 99

99 A. B. C. Figure 5 8. Kinetic models for the adsorption of Hg(II) onto 3:1 MPAC. A) Intraparticle diffusion model, B) Pseudo first order model, and C) Pseudo second order model R = 0.5787 y = 0.6347x + 88.454 84 86 88 90 92 94 96 98 0 2.5 5 7.5 10 12.5 15 q t ( g/g) t 1/2 (min 1/2 ) y = 0.0187x + 0.8525 R = 0.5694 -3 -2 -1 0 1 2 3 0 50 100 150 200 log (q e q t ) t (min) y = 0.0105x + 0.0012 R = 0.9999 0 0.5 1 1.5 2 0 50 100 150 200 t/q t (L/ t (min)

PAGE 100

100 Figure 5 9 Hg leaching from 3:1 C:Fe at various loading rates under landfill conditions Figure 5 10. Background Hg(0) mass balance for a 30 s contact time 0 2 4 6 8 50:1 100:1 200:1 500:1 800:1 1000:1 g Hg leached / g MPAC Loading Ratio ( g Hg/ g MPAC ) Volatilized In Solution Fugitive

PAGE 101

101 A. B Figure 5 11 Hg removal through ads orption and volatilization for various surface modified carbons A) Hg(II) B) Hg(0)

PAGE 102

102 A B. C Figure 5 12. Hg(II) adsorption isotherm onto NAC 1M. A) Nonlinearized adsorption isotherm B) Freundlich model, and C ) Langmuir model y = 35.625x 15.284 R = 0.9996 y = 16.918x + 109.81 R = 0.9652 0 50 100 150 200 250 0.0 2.0 4.0 6.0 8.0 qe Ce Freundlich Langmuir log C e = 1.1612(log q e ) + 1.3937 R = 0.9118 n = 1.1612, k = 1.3937 0 0.5 1 1.5 2 2.5 0.0 0.2 0.4 0.6 0.8 1.0 log q e log C e 1/q e = 0.0062(1/C e ) 0.0037 R = 0.929 a = 16.189, b = 0.9963 0 0.01 0.02 0.03 0.04 0 2 4 6 1/q e 1/C e

PAGE 103

10 3 A. B. C. Figure 5 13. Kinetic models for the adsorption of Hg(II) onto C(O) modified carbons. A) Int raparticle diffusion model, B) pseudo first order model, and C) pseudo second order model 50 60 70 80 90 100 0 5 10 15 q t (ug/g) t 1/2 (min 1/2 ) ACH NAC-1M SAC-1M SHAC-1M -8 -6 -4 -2 0 2 4 0 50 100 150 log(q e q t ) t (min) ACH NAC-1M SAC-1M SHAC-1M 0 0.5 1 1.5 2 0 25 50 75 100 125 150 t/q t (L/mgmin) t (min) ACH NAC-1M SAC-1M SHAC-1M

PAGE 104

104 CHAPTER 6 ADSORPTION MECHANISMS One objective of this study was to determine the influence of Hg speciation on adsorption mechanisms. The Hg speciation in each syste m was determined using Visual MINTEQ. The binding mechanisms were predicted based on this speciation. A sequential chemical extraction was designed with the goal of verifying these binding mechanisms. Prior to applyi ng the SCE to the MPAC and C(O) modified carbon systems, its performance was evaluated by forcing Hg to known speciation through manipulation of pH and pCl and quantifying Hg desorbed in each phase. Proposed Adsorption Mechanisms Mechanisms of Hg (II) Adsorpti on The unadjusted matrix pH ranged b etween 4.4 and 4.7. Using the speciation program Visual MINTEQ 2.61, the mercury speciation in the given matrix conditions was predicted to be 96.5 99% Hg(OH) 2 and 1 3.5% HgOH + MPAC. Upon addition of 3:1 C:Fe MPAC, the pH of the aqueous solution reached a n equilibrium value of 6.2. Under these conditions, the Hg speciation was nearly 100% Hg(OH) 2 which was likely removed from aqueous solution by preferential precipitation on to the MPAC surface once maximum solubility was reached. C(O) modified carbons. Th e pH of the aqueous solution varied between 3.41 and 5.45, depending on the modification of the carbon (Table 6 1) The more basic systems contained Hg primarily as Hg(OH) 2 while the more acidic systems contained Hg in various states of hydrolysis, includi ng Hg 2+ HgOH + and Hg(OH) 2 Hg(OH) 2 was likely removed due to preferential precipitation onto the carbon surface. For the systems with a contact pH below the pzc, the carbon surface was positively charged and

PAGE 105

105 electrostatically repelled the Hg cations The systems with a contact pH above the pzc, such as NAC 10M, SAC carbons, and CASPF modified carbons, the carbon surface was negatively charged and thus Hg cations were electrostatically attracted to the surface. Mechanisms of Hg (0) Adsorption Aqueous Hg(0 ) can undergo physisorption. The matrix pH and pCl do not influence its adsorption. Gas phase research proposed Hg(0) oxidation by carbonyl containing C(O) groups and subsequent adsorption via known Hg(II) adsorption mechanisms while phenolic groups have been shown to decrease Hg(0) adsorption [72,73 ] The lack of correlation of Hg adsorption with C(O) d oes not support this occurrence in the aqueous phase. Ideally, individual surface oxygen groups would be quantified in order to determine their specific relationship, if any, to adsorption. A chemical sequential extraction may provide more insight into the speciation of the elemental Hg, once adsorbed. Influence of pH and pCl on Hg(II) a dsorption The adsorption of Hg by 3:1 MPAC was investiga ted at various pH and pCl values Previously published literature reported a decrease in Hg adsorption onto activate d carbon with an increase in chloride concentration [13 3 ] This study supports those findings. A s the pH increased from pH 2 to pH 10, Hg adsorption decreased for the three pCl values investigated (Figure 6 1) T h e system with the highest chloride concentration showed the most significant decrease in adsorption with increasing pH. The average mass balance closure of these runs was 95% 5%. With an adjusted R 2 of 0.557, a two variable model indicated both pH and pC l are go od regression parameters with p values of 0.0254 and 0.0096, respectively. An ANOVA test showed pH and pCl

PAGE 106

106 have similar influence on the Hg(II) adsorption, with sum of squares of 34.2 and 50.8, respectively. The influence of p H and pCl on Hg volat ilization from the Hg(II) system is presented in Figure 6 2 At pH 2, little volatilization occurred at an y chloride concentration. Regression analysis revealed that pH and pCl do not significantly influence Hg(0) volatilization, with an adjusted R 2 of 0. 02. Sequential Chemical Extraction Protocol Verification In order to ensure the e xtractant selections were sufficiently specific and efficient to predict speciation the pH and pCl was adjusted to control speciation (Table 6 2 ). If properly designed, the distribution of Hg among the extraction fractions can be predicted. Free Hg(II ), although predicted to desorb in the ammonium nitrate fraction due to ion exchange primarily desor bed in the acetic acid, DMSA, and HF residual fractions indicating ion exc hange was not the only primary binding mechanism (Table 6 3 ). Potential causes include: phase transformation that altered the adsorption mechanisms the ammonium nitrate extractant was inefficient at targeting ion exchange, or if the Hg was not present as the predicted species. Surprisingly, a large amount of the Hg remained in solution, unadsorbed. As predicted, the largest portion of uncharged Hg(OH) 2 desorbed in the acetic acid fraction, indicating surface precipitation. Although minimal, detectable lev els were found in other extraction fractions, demonstrating phase transformation or non ideal extractant performance due to a lack of specificity or poor extraction efficiency. Of note, a large

PAGE 107

107 amount of Hg volatilized from this system, indicating an Hg(II ) reduction mechanism that was not expected at the pH and pCl of the system. HgCl 2 was expected to desorb in the surface bound fraction and to volatilize from solution as Hg(0). Although Hg desorbed in the expected fraction, a significant portion also des o rbed in the acetic acid fraction It is possible that HgCl 2 was reduced to Hg(0) and the highly insoluble Hg 2 Cl 2 preferentially precipitating on the carbon surface [81] Hg Cl anions were expected to desorb primarily i n the surface bound phase. With the adsorbent pH pzc of 9.3, the sorbent was positively charged, and should have resulted in an electrostatic attraction between the negatively charged Hg and the positively charged surface enhancing adsorption This was not realized as approximately 14% of the Hg remained in solution at equilibrium. While a large portion of the Hg desorbed in the DMSA extraction fraction, significant desorption also occurred in the acetic acid fraction. A very small percentage of the Hg was predicted to be present as HgCl 2 so reduction to Hg 2 Cl 2 was not expected to largely influence the results. The low rate of volatilization, 1%, further indicates that this reduction does not account for the Hg association with the surface bound fraction. A pplication Although the results were interesting, it was clear that the SCE described could not accurately predict the speciation of Hg that was adsorbed from aqueous solution. It was beyond the scope of this study to pursue a stronger extraction scheme.

PAGE 108

108 Table 6 1. Variation of 30 s Hg (II) DI contact pH with pH pzc of C(O) modified carbons Sample pH pzc Hg DI c ontact pH CASPF 1.93 3.41 WPH 8.36 5. 16 ACH 10.10 5.45 NAC 1M 6.56 4. 84 NAC 5M 5.02 4.82 NAC 10M 3.99 4.73 SAC 1M 3.78 4.91 SAC 5M 3.90 4 .97 SAC 10M 3.36 4.90 SHAC 1M 7.37 5. 23 SHAC 5M 7.25 5.3 5 CASPF SAC10M 2.83 4.51 CASPF NAC10M 2. 77 4.25

PAGE 109

109 Table 6 2 Predicted Hg speciation and SCE extraction fraction for given pH and pCl values pH pCl Hg Speciation Description Proposed e xtractan ts Hg 2+ HgOH + Hg(OH) 2 HgCl + HgCl 2 HgCl 3 HgCl 4 2 1 12 100% Free Hg 2, 5 6 12 100% Precipitated 3 3 4 1% 99 % Uncharged Hg Cl 4 8 0 6 27 % 67 % Hg Cl anions ? Table 6 3 Hg distribution in SCE extraction fractions Hg Speci ation Water Ammoniu m n itrate Acetic a cid DMSA DCB HF Volatilized In s olution Fugitive Free Hg 2+ 5.1% 3.1% 17.0% 27.4% 1.4% 15.2% 2.8% 17.7% 10.3% Hg(OH) 2 2.5% 7.0% 25 .9% 8.4% 0.4% 16.7% 20.8% 3.2% 15 .1% HgCl 2 2.1% 6.7% 23.9% 30.6% 0.9% 21.6% 7.5% 6.6% 0.1% HgCl 3 HgCl 4 2 1.1% 3.9% 21.0% 36.7% 2.5% 18.9% 1.0% 13.8% 1.1%

PAGE 110

110 Figure 6 1. Influence of pH on aqueous Hg(II) adsorption Figure 6 2. Influence of pCl on Hg(II) volatilization 80% 85% 90% 95% 100% 0 2 4 6 8 10 12 Hg Removal (%) pH pCl 12 pCl 6 pCl 0 0% 5% 10% 15% 20% 0 5 10 15 Hg Volatilized (%) pCl pH 2 pH 10 pH 6

PAGE 111

111 CHAPTER 7 CONCLUSIONS AND RECO MMENDATIONS Magnetic Powdered Ac tivated Carbon The magnetic powdered activated carbon synthesized by iron imp regnation and thermal oxidation, was optimized for mercury removal T he 3:1 C:Fe MPAC reached the goal of 95% sorbent recovery w ith only a 25% decrease from the virgin carbon su rface area. The presence of maghemite and amorphous iron oxides was confirmed on the 3:1 C:Fe MPAC. Thermal oxidation succeeded in decreasing the amorphous characteristic of the MPACs but did not provide a significant increase in magnetic recovery or Hg re moval performance. The potential benefits of thermal oxidation are not realized and are outweighed by the damaged porosity and increased cost in production. When exposed to an acidic matrix pH the 3:1 MPAC leached low concentrations of Fe. Iron is not a c oncern from a regulatory standpoint and this leaching did not cause coloration of the water nor did it influence the sorbent recoverability In addition to ideal magnetic recovery, the 3:1 C:Fe MPAC outperformed other MPACs for Hg(II) removal The 3:1 MPA C exhibited the highest adsorption capacity. At a pseudo equilibrium contact time of 120 min with a 100 g/L Hg solution at unadjusted pH, the 3:1 MPAC performed optimally, achieving 91% Hg removal with 2% volatilized, 84% adsorbed, while 4% remained fugit ive. The average Hg mass balance closure for all 17 runs was 99.5% with a standard deviation of 8.8%, verifying the MPAC Hg removal performance. Surface area appears to influence adsorption in this system but, with a correlation of only 0.47 another facto r is also influencing the system The adsorption data fits both the Freundlich and Langmuir models, indicating that Hg adsorption proceeds both as chemisorption and physisorption. As the data strongly fits

PAGE 112

112 the pseudo second order model, chemisorption is c l early involved in this system. Once adsorbed, the Hg is strongly bound to the MPAC surface. Hg leaching does not necessitate special residuals handling until a loading of greater than 800 g Hg/ g MPAC. Matrix pH and pCl are known to influence Hg speciatio n. Both pH and pCl were shown to influence Hg adsorption onto 3:1 C:Fe MPAC. This influence was used investigate the use of a sequential chemical extraction to predict Hg speciation and binding mechanisms. The results clearly showed that the SCE described could not accurately predict the speciation of Hg that was adsorbed from aqueous solution. Surface Oxygen Modified Carbon C ommercially available activated carbon s underwent wet chemical oxidation with HNO 3 H 2 SO 4 and NaOH increasing surface oxygen funct ionality with the goal of increase d Hg(II) and Hg(0) adsorption. Nitric acid modification produced the most surface oxygen groups but resulted in slight damage to porosity Sulfuric acid and sodium hydroxide modification did not damage porosity but were le ss effective than nitric acid at increasing the surface oxygen functionality. The model that best fit Hg(II) adsorption identified oxygen content and pH pzc as important variables, with oxygen content being the primary variable influencing the results. Hg( 0) adsorption data best fit a four variable model, indicating that surface area, pore volume, surface oxygen functionality, and the pH pzc as good regression parameters, with the pH pzc as the primary variable influencing the results. Neither model achieved a strong R 2 value. It is possible that an unquantified variable influenced these results. Due to the uncharged nature of Hg(0), it is possible that water cluster formation due to C(O) groups, limited adsorption. A minimum of surface oxygen groups are

PAGE 113

113 requ ired for the surface to be sufficiently hydrophilic, allowing the surface to be wetted by water and thus useful for water treatment applications. Therefore, a moderate amount of surface oxygen groups are optimal for Hg(0) adsorption from aqueous solution. As no carbons violated TCLP effluent limits, it can be inferred that the Hg is strongly bound to the surface. Hg(II) adsorption onto the C(O) modified carbons fit both the Freundlich and Langmuir models, indicating that both physisorption and chemisorptio n occur. The data fit both pseudo first order and pseudo second order models very well, also supporting the occurrence of both physisorption and chemisorption. In summary, it is possible to tailor activated carbon to allow for magnetic recapture It also possible to enhance aqueous Hg(II) capture through surface oxygen modification, although Hg(0) adsorption is not influenced by these surface groups. Both carbons produced are stable and, under the experimental loading conditions applied, do not require spe cial handling or disposal as a hazardous waste. Th e most effective aqueous Hg treatment method will depend on water chemistry, sorbent surface chemistry and Hg speciation. Contributions to Science Demonstrated that magnetic recovery is possible with low C:Fe without significant changes to surface area, pore size, and pore volume Found that thermal oxidation, although achieving the goal of converting amorphous iron oxides to more crystalline form, did not result in improved sorbent recapture Identified 3 :1 C:Fe without thermal oxidation as the optimal synthesis parameters for trace level aqueous Hg removal Increased understanding of Hg adsorption mechanisms by:

PAGE 114

114 o Suggesting the influence of water cluster formation on aqueous Hg(0) adsorption. o Demonstrated that surface oxygen functionality alone is not strongly correlated to aqueous Hg(0) adsorption. o Identified porosity, adsorbent surface charge, and oxygen content as significant variables in aqueous Hg(0) adsorption. o Demonstrated that, although porosity w as not exerting a large influence on aqueous Hg(II) adsorption, pore volume influenced the results to a greater degree than surface area. Determined that pH and pCl do not significantly influence Hg volatilization from solution. Demonstrated that activate d carbon can be used to adsorb aqueous Hg(0); improved aqueous capture is beneficial by reducing Hg losses to the atmosphere due to volatility. Future Recommendations Combine the magnetic and surface oxygen group modification techniques Apply the modifi ed carbons to real wastewaters Confirm oxidation of Hg(0) using SEM and XRD. Determine the identit y and concentration of surface oxygen functional groups developed with the wet chemical oxidation methods; determi ne any correlation between these groups a nd Hg (II) and removal

PAGE 115

115 APPENDIX A MODIFICATION OF SURFACE OXYGEN FUNCTIONALITY OF BIOCHAR FOR HG ADSORPTION In addition to wood and coal based carbons, recent literature investigates the use of more sustainable biomass carbon sources [75,13 4 ,1 35 ] Biochar, a sustainable and affordable pyroly zed carbon commonly applied to soils to increase fertility and water retention, can exhibit high surface area and may act as a surface sorbent similar to activated carbon. Adsorption of Cu, Ni, Cd, and Pb onto biochar has been correlated with the amount of C(O) groups present, determined by O /C ratio, pHpzc, total acidity, and 1 H NMR analysis [13 6 ] This study utilized the same surface oxygen modification applied to activated carbon. Table A 1 shows the bioch ar characterization results. No biochars investigated demonstrated high surface area. Modification did not significantly alter porosity or surface charge. Table A 2 shows the Hg removal performance. Batch adsorption studies were performed at room temperature with a 150mg/L dose of biochar to 50 g/L Hg DI for a 30 s contact time. No biochars performed as well as the activated carbons previously discussed. The modification did not influence the Hg(II) adsorption efficiency. It is of interest that Hg does have an affinity for biochar, even if this affinity is lower than activated carbon and is not influenced by C(O) groups.

PAGE 116

116 Table A 1. Biochar characterization data Sample Raw 10M H 2 SO 4 Modification pH pzc Surface Area (m 2 /g) Pore Size () Pore Volume (cm 3 /g) O/C pH pzc Surface Area (m 2 /g) Pore Size () Pore Volume (cm 3 /g) Fresh Oak 250 3.9 1 99.7 0.00 0.8 3.7 0 117.2 0.00 Fresh Oak 650 9.7 46 17.8 0.04 0.2 9.4 85 15.1 0.06 Fresh Grass 250 4.4 2 6.7 0.01 0.8 4.5 6 45.3 0.01 Fresh Grass 650 9.7 12 45.1 0.03 0.5 9.6 2 54.0 0.01

PAGE 117

117 Table A 2. Adsorption of aqueous Hg(II) by raw and modified biochar Sample Raw Modified Hg Removal (%) Hg Removal (%) Fresh Oak 250 39.3 35.9 Fresh Oak 650 34.4 40.0 Fresh Grass 250 41.4 38.7 Fresh Grass 650 19.1 22.5

PAGE 118

118 LIST OF REFERENCES [1] U.S. Environmental Protection Agency. Toxics release inventory: Reporting year 2007 public data release summary of key findings. 260 R 09 001, 2009. [2] U.S. Code of Federal Regulations. Water quality criteria for the protection of wildlife. 40CFR132 Table 4, 2010. [3] Drasch, Horvat M, Stoeppler M. Mercury. In: Merian E, Anke M, Ihnat M, Stoeppler M, editors. Elements and their compounds in the en vironment, 2nd ed., Weinheim; Wiley VCH; 2004 p. 931 1006. [4] U.N.E.P Chemicals. Global mercury asse ssment. Geneva, Switzerland, 54790 01, 2002. [5] Burgess J. Ions in solution: Basic principles of chemical interactions. 2nd ed. Coll House, England: Horwood Publishing; 1999. [6] Kim CS, Rytuba JJ, Brown GE. EXAFS study of mercury(II) sorption to F e an d A l (hydr)oxides I. effects of pH. J Colloid Interface Sci 2004; 271(1): 1 15. [7] Benjamin M. Water chemistry. Boston MA : McGraw Hill; 2002. [8] Hahne HC, Kroontje W. Significance of pH and chloride concentration on behavior of heavy metal pollutants: Mercury(II), cadmium(II), zinc(II), and lead(II). J Environ Qual. 1973; 2(4): 444 450. [9] Davis A, Bloom NS, Hee SS. The environmental geochemistry and bioaccessibility of mercury in soils and sediments: A review. Risk Analysis 1997; 17(5): 557 569. [10 ] Merian E editor. Metals and their compounds in the environment. Weinheim NY : VCH; 1999. [11] Schwertmann U, Cornell RM Iron oxides in the laboratory : Preparation and characterization. 2nd ed. Weinheim N Y : Wiley VCH; 2000. [12] Kim CS, Rytuba JJ, Brown GE. EXAFS study of mercury(II) sorption to F e and A l (hydr)oxides: II. effects of chloride and sulfate. J Colloid Interface Sci 2004; 270(1): 9 20. [13] Hahne HCH, Kroontje W. The simultaneous effect of pH and chloride concentrations upon mercu ry(II) as a pollutant. Soil Sci Soc Am J 1973; 37(6): 838 843.

PAGE 119

119 [14] Feick G, Yeaple D, Horne RA. Release of mercury from contaminated freshwater sediments by runoff of road deicing salt. Science 1972; 175(4026): 1142 1143. [15] U.S. E nvironmental P rotection A genc y Mercury study report to congress, volulme V: Health effects of mercury and mercury compounds. EPA 452/R 97 007, 1997. [16] Langford N, F erner R. Toxicity of mercury. J Hum Hypertens 1999; 13(10): 651 656. [17] Harada H. Congenital M inimata disease: In traeuterine methyl mercury poisoning. Teratology 1978(18): 285 288. [18] Bakir F, Damluji S, Amin Zski L. Methyl mercury poisoning in I raq. Science 1973(181): 230 241. [19] U.S. Environmental Protection Agency. Mercury study report to congress, volume II : An inventory of anthropogenic mercury emissions in the U nited S tates. Washington DC, EPA 452/R 97 004, 1997. [20] U.S. Environmental Protection Agency. Roadmap for mercury. Washington DC, EPA HW QPPT 2005 0013, 2006. [21] U.S. Environmental Protection Agency. Compilation of air pollutant emission factors, chap ter 8: i norganic chemical industry. Research Triangle park NC, AP 42 199 5. [22] Bansal R, Goyal M. Activated carbon adsorption. Boca Raton: CRC Press; 2005. [23] Anirudhan TS, Divya L, Ramachand ran M. Mercury(II) removal from aqueous solutions and wastewaters using a novel cation exchanger derived from coconut coir pith and its rec overy. J Hazard Mater 2008; 157(2 3): 620 627. [24] von Canstein H, Li Y, Timmis KN, Deckwer WD, Wagner Dobler I. Re moval of mercury from chloralkali electrolysis wastewater by a mercury resistant pseudomonas putida strain. Appl Environ Microbiol 1999; 65(12): 5279 5284. [25] U.S. E nvironmental P rotection A gency National emission s tandards for h azardous a ir p ollutants : m ercury e missions from m ercury c ell c hlor a lkali p lants; f inal r ule. 40CFR63 2003 [26] Tewalt SJ, Bragg LJ, Finkelman, RB Mercury in U.S. coal abundance, distribution, and modes of occurrence. Washington DC, FS 095 01, 2001.

PAGE 120

120 [27] Chang J, Ghorishi S. Simulation and evaluation of elemental mercury concentration increase in flue gas across a wet scrubber. Environ Sci Technol 2003; 37(24): 5763 5766. [28] U.S. E nvironmental P rotection A gency Steam electric power generating effluent guidelines rulema king supplemental information p ackage #2 for federalism and u nfunded mandates r eform a ct c onsultations, Washington DC, EPA 821 R 09 008, 2011. [29] U.S. Environmental Protection Agency. Capsule report: Aqueous mercury treatment. Washington DC, EPA/625/R 97/004, 1997. [30] Patterson, Luo B, Petropoulou C, Gasca E. Toxicity reduction methodologies. In: Ford D, editor. Toxicity reduction, evaluation, and control, 2nd ed., Lancaster, PA; Technomic Publishing Co.; 1998 p. 109. [31] Findlay DM, McLean RAN. Removal of elemental mercury from wastewaters u sing polysulfides. Environ S ci T echn ol 1981; 15(11): 1388 1390. [32] Moorhouse J. Modern chlor alkali technology, vol. 8. Oxford: Blackwell Science Ltd.; 2001. [33] Huang CP, Blankenship DW. The removal of m ercury(II) from dilute aqueous solution by activated carbon. Water Res 1984; 18(1): 37 46. [34] Humenick MJ, Schnoor JL. Improving mercury (II) removal by activated carbon. J Environ Eng A SCE 1974; 100(NEE6): 1249 1262. [35] Marsh H, Rodriguez Reinoso F. Activated carbon. Amsterdam: Elsevier; 2006. [36] Lopez Gonzalez J, Moreno Castilla C, Guerrero Ruiz A, Rodriguez Reinoso F. Effect of carbon oxygen and carbon sulphur surface complexed on the adsorption of mercuric chloride in aqueous solutions by activ ated carbons. J Chem Tech Biotechnol 1982; 32: 575 579. [37] Krishnan KA, Anirudhan TS. Removal of mercury(II) from aqueous solutions and chlor alkali industry effluent by steam activated and sulphurised activated carbons prepared from bagasse pith: Kinet ics and equilibrium studies. J Hazard Mater 2002; 92(2): 161 183. [38] Manchester S, Wang X, Kulaots I, Gao Y, Hurt RH. High capacity mercury adsorption on freshly ozone treated carbon surfaces. Carbon 2008; 46(3): 518 524. [39] Tong S, Fan M, Mao L, Jia CQ. Sequential extraction study of stability of adsorbed mercury in chemically modified activated carbons. Environ Sc i Technol 2011; 45(17): 7416 7421.

PAGE 121

121 [40] US Environmental Protection Agency, Office of Superfund Remediation and Technology Innovation. Tr eatment technologies for mercury removal from soil, waste, and water. Washington DC, EPA 542 R 07 003, 2007. [41] Boehm HP. Surface oxides on carbon and their analysis: A critical assessment. Carbon 2002; 40(2): 145 149. [42] Yang RT. Adsorbents : Fundame ntals and applications. Hoboken NJ Wiley; 2003. [43] Yin CY, Aroua MK, Daud WMAW. Review of modifications of activated carbon for enhancing contaminant uptakes fr om aqueous solutions. Sep Purif Technol 2007; 52(3): 403 415. [44] Puri BR. Surface complex es on carbons. Chem Phys Carbon 1970; 6: 191 282. [45] Zhao NQ, Wei N, Li JJ, Qiao ZJ, Cui J, He F. Surface properties of chemically modified activated carbons for adsorption rate of cr(VI). Chem. Eng. J. 2005; 115(1 2): 133 138. [46] Saha B, Tai MH, Str eat M. Metal sorption performance of an activated carbon after oxidation and subsequent treatment. Process Saf. Environ. Prot. 2001; 79(B6): 345 351. [47] Biniak S, Szymanski G, Siedlewski J, Swiatkowski A. The characterization of activated carbons with o xygen and nitrogen surface groups. Carbon 1997; 35(12): 1799 1810. [48] Shim JW, Park SJ, Ryu SK. Effect of modification with HNO3 and NaOH on metal adsorption by pitch based activated carbon fibers. Carbon 2001; 39(11): 1635 1642. [49] Biniak S, Pakula M, Szymanski GS, Swiatkowski A. Effect of activated carbon surface oxygen and/or nitrogen containing groups on adsorption of copper(II) ions from aqueous solution. Langmuir 1999; 15(18): 6117 6122. [50] Shen W, Li Z, Liu Y. Surface chemical functional gr oups modification of porous carbon. Recent Patents on Chemical Engineering 2008; 1: 27 40. [51] Santiago M, Stuber F, Fortuny A, Fabregat A, Font J. Modified activated carbons for catalytic wet air oxidation of phenol. Carbon 2005; 43(10): 2134 2145. [52 ] Figueiredo JL, Pereira MFR, Freitas MMA, Orfao JJM. Modification of the surface chemistry of activated carbons. Carbon 1999; 37(9): 1379 1389.

PAGE 122

122 [53] Maroto Valer MM, Dranca I, Lupascu T, Nastas R. Effect of adsorbate polarity on thermodesorption profiles from oxidized and metal impregnated activated carbons. Carbon 2004; 42(12 13): 2655 2659. [54] Rios RRVA, Alves DE, Dalmazio I, Fernando S, Bento V, Donnici CL, et al. Tailoring activated carbon by surface chemical modification with O, S, and N containin g molecules. Mat Res 2001; 6(2): 129 135. [55] Moreno Castilla C, Carrasco Marin F, Maldonado Hodar FJ, Rivera Utrilla J. Effects of non oxidant and oxidant acid treatments on the surface properties of an activated carbon with very low ash content. Carbon 1998; 36(1 2): 145 151. [56] Li YH, Lee CW, Gullett BK. Importance of activated carbon's oxygen surface functional groups on elemental mercury adsorption. Fuel 2003; 82(4): 451 457. [57] Moreno Castilla C, Ferrogarcia MA, Joly JP, Bautistatoledo I, Carr ascomarin F, Riverautrilla J. Activated carbon surface modifications by nitric acid, hydrogen peroxide, and ammonium peroxydisulfate treatments. Langmuir 1995; 11(11): 4386 4392. [58] Salame II, Bandosz TJ. Study of water adsorption on activated carbons w ith different degrees of surface oxidation. J Colloid Interface Sci 1999; 210(2): 367 374. [59] Salame II, Bandosz TJ. Surface chemistry of activated carbons: Combining the results of temperature programmed desorption, boehm, and potentiometric titrations J Colloid Interface Sci 2001; 240(1): 252 258. [60] Oliveira LCA, Rios RVRA, Fabris JD, Garg V, Sapag K, Lago RM. Activated carbon/iron oxide magnetic composites for the adsorption of contaminants in water. Carbon 2002; 40(12): 2177 2183. [61] Oliveira LCA, Petkowicz DI, Smaniotto A, Pergher SBC. Magnetic zeolites: A new adsorbent for removal of metallic contaminants from water. Water Res 2004; 38(17): 3699 3704. [62] Gorria P, Sevilla M, Blanco JA, Fuertes AB. Synthesis of magnetically separable adsor bents through the incorporation of protected nickel nanoparticles in an activated carbon. Carbon 2006; 44(10): 1954 1957. [63] Cornell R, Schwertmann U. The iron oxides. Weinheim: VCH; 1996. [64] Okamoto S. Iron hydroxides as magnetic scavengers. IEEE Tr ans Magn 1974; MA10(3): 923 926.

PAGE 123

123 [65] Masel RI. Principles of adsorption and reaction on solid surfaces. New York: Wiley; 1996. [66] Faust S, Aly O. Adsorption processes for water treatment. Boston: Butterworth Publishers; 1987. [67] Stumm W, Morgan J. Aquatic chemistry : Chemical equilibria and rates in natural waters 3rd ed. ed. New York: Wiley; 1996. [68] International Union of Pure and Applied Chemistry. Manual of Symbols and Terminology, Part 1, Colloid and Surface Chemistry. 2002; Available at: http://www.iupac.org/reports/2001/colloid_2001/manual_of_s_and_t/ Accessed February, 2009. [69] Calgon Carbon Corporation. Laboratory evaluation of granular activ ated carbons for liquid phase applications. Pittsburg PA, IB 1002 0907, 2007; [70] Coolidge AS. The adsorption of mercury vapor by charcoal. J Am Chem Soc 1927; 49: 1949 1952. [71] Shiels DO. The adsorption of mercury vapor by activated charcoal. J Phy s Chem 1929; 33: 1398 1402. [72] Krishnan SV, Gullett BK, Jozewicz W. Sorption of elemental mercury by activated carbons. Environ Sci Technol 1994; 28(8): 1506 1512. [73] Liu J, Cheney MA, Wu F, Li M. Effects of chemical functional groups on elemental me rcury adsorption on carbonaceous surfaces. J Hazard Mater 2011; 186(1): 108 113. [74] Goel J, Kadirvelu K, Rajagopal C, Garg VK. Removal of mercury(II) from aqueous solution by adsorption on carbon aerogel: Response surface methodological approach. Carbon 2005; 43(1): 197 200. [75] Kadirvelu K, Kavipriya M, Karthika C, Vennilamani N, Pattabhi S. Mercury(II) adsorption by activated carbon made from sago waste. Carbon 2004; 42(4): 745 752. [76] Carrott PJM, Carrott MMLR, Nabais JMV. Influence of surface io nization on the adsorption of aqueous mercury chlorocomplexes by activated carbons. Carbon 1998; 36(1 2): 11 17. [77] Sato S, Yoshihara K, Moriyama K, Machida M, Tatsumoto H. Influence of activated carbon surface acidity on adsorption of heavy metal ions and aromatics from aqueous solution. Appl Surf Sci 2007; 253(20): 8554 8559.

PAGE 124

124 [78] Xiao B, Thomas KM. Competitive adsorption of aqueous metal ions on an oxidi zed nanoporous activated carbon Langmuir 2004; 20(11): 4566 4578. [79] SenGupta A. Environmental separation of heavy metals: Engineering processes. Boca Raton : CRC Press; 2002. [80] Radovic LR, Moreno Castilla C, Rivera Utrilla J. Carbon materials as adsorbents in aqueous solutions. Chem Phys Carbon 2001; 27: 227 405. [81] Cox M, El Shafey E, Pichu gin AA, Appleton Q. Removal of mercury(II) from aqueous solution on a carbonaceous sorbent prepared from flax shive. J Chem Technol Biotechnol 2000; 75(6): 427 435. [82] Wiatrowski HA, Das S, Kukkadapu R, Ilton E, Barkay T, Yee N. Reduction of hg(II) to h g(0) by magnetite. Environ Sci Technol 2009; 73(14): 5307 5313. [83] Newton DW, Ellis R, Paulsen GM. Effect of ph and complex formation on mercury (II) adsorption by bentonite. J Environ Qual 1976; 5(3): 251 254. [84] Sarkar D. Preliminary studies on mer cury solubility in the presence of iron oxide phases using static h eadspace analysis. Environ Geo 2003; 10(4): 151 155. [85] Tiffreau C, Lutzenkirchen J, Behra P. Modeling the adsorption of mercury (II) on (hydr)oxides. J Colloid Interf Sci 1995; 172: 82 93. [86] Fauconnier N, Bee A, Roger J, Pons JN. Synthesis of aqueous magnetic liquids by surface complexation of maghemite nanoparticles. J Mol Liq 1999; 83(1 3): 233 242. [87] O'Loughlin EJ, Kelly SD, Kemner KM, Csencsits R, Cook RE. Reduction of A g I, A u III, C u II, and H g II by F e II/Fe III hydroxysulfate green rust. Chemosphere 2003; 53(5): 437 446. [88] Charlet L, Bosbach D, Peretyashko T. Natural attenuation of TCE, as, H g linked to the heterogeneous oxidation of F e(II): An AFM study. Chem Geol 200 2; 190(1 4): 303 319. [89] Scheinost AC, Charlet L. Selenite reduction by mackinawite, magnetite and siderite: XAS characterization of na nosized redox products. Environ Sci Technol 2008; 42(6): 1984 1989. [90] White AF, Peterson ML. Reduction of aqueous transition metal species on the surfaces of fe(II) containing oxides. Geochim Cosmochim Acta 1996; 60(20): 3799 3814.

PAGE 125

125 [91] Senior C, editor. Behavior of mercury in air pollution control devices on coal fired utility boilers. Engineering Foundation Confere nce; 2001; Salt Lake City, UT. [92] Lighty, Silcox and Fry. Fundamentals of mercury oxidation in flue gas. Pittsburg PA, DE FG26 03NT41797, 2004. [93] Dunham GE, DeWall RA, Senior CL. Fixed bed studies of the interactions between mercury and coal combust ion fly ash. Fuel Process Technol 2003; 82(2 3): 197 213. [94] Snoeyink V, Jenkins D. Water chemistry. New York: John Wiley & Sons; 1980. [95] Elliot P, Hartenstein H, editors. Selective Separation of Mercury and Other Heavy Metals During FGD Wastewater Treatment. APC Round Table and Expo; July 8 10, 2007; Chattanooga, TN; 2007. [96] Hahne HCH, Kroontje W. Simultaneous effect of ph and chloride concentrations upon mercur y (ii) as a pollutant. Soil Sci Soc Am J 1973; 37(6): 838 843. [97] Sing K, Everett D, Haul R, Moscou L, Pierotti R, Rouquerol J, et al. Reporting physisorption data for gas solid systems with special reference to the determination of surface area and porosity Pure Appl Chem 1985; 57(4): 603 619. [98] Sanemasa I. Solubility of elemen tal mercury vapor in water. Bull Chem Soc Jpn 1975; 48(6): 1795 1798. [99] Bloom NS, Preus E, Katon J, Hiltner M. Selective extractions to assess the biogeochemically relevant fractionation of inorganic mercu ry in sediments and soils. Anal Chim Acta 2003; 47 9(2): 233 248. [100] Perez OP, Umetsu Y, Sasaki H. Precipitation and densification of magnetic iron compounds from aqueous solutions at room temperature. Hydrometallurgy 1998; 50(3): 223 242. [101] Menendez JA, Phillips J, Xia B, Radovic LR. On the modif ication and characterization of chemical surface properties of activated carbon: In the search of carbons with stable basic properties. Langmuir 1996; 12(18): 4404 4410. [102] Laine NR, Vastola FJ, Walker PL. Importance of active surface ar ea in carbon ox ygen reaction. J Phys Chem 1963; 67(10): 2030 &. [103] Aggarwal D, Goyal M, Bansal RC. Adsorption of chromium by activated carbon from aqueous solution. Carbon 1999; 37(12): 1989 1997.

PAGE 126

126 [104] Menendez JA, Xia B, Phillips J, Radovic LR. On the modification and characterization of chemical surface properties of activated carbon: Microcalorimetric, electrochemical, and thermal desorption probes. Langmuir 1997; 13(13): 3414 3421. [105] Dastgheib SA, Karanfil T. Adsorption of oxygen by heat treated granular a n d fibrous activated carbons. J Colloid Interface Sci 2004; 274(1): 1 8. [106] Li YH, Lee CW, Gullett BK. The effect of activated carbon surface moisture on low temperature mercury adsorption. Carbon 2002; 40(1): 65 72. [107] Chingombe P, Saha B, Wakeman RJ. Surface modification and characterisation of a coal based activated carbon. Carbon 2005; 43(15): 3132 3143. [108] Quantachrome Instruments. Nova e series: High speed surface area and pore size analyzers. 2008; Available at: http://www.quantachrome.com/pdf_brochures/07122.pdf Accessed 08/07, 2009. [109] Brunauer S, Emmett PH, Teller E. Adsorption of ga ses in multimolecular layers. J Am Chem Soc 1938; 60: 309 319. [110] Ba rrett EP, Joyner LG, Halenda PP. The determination of pore volume and area distributions in porous substances .1. computat ions from nitrogen isotherms. J Am Chem Soc 1951; 73(1): 373 380. [111] Cho D, Chon C, Kim Y, Jeon B, Schwartz FW, Lee E, et al. Adso rption of nitrate and cr(VI) by cationic polymer modified granular activated carbon. Chem Eng J 2011; 175: 298 305. [112] Al Rashdi B, Somerfield C, Hilal N. Heavy metals removal using adsorption and nanofiltration techniques. Sep Purif Rev 2011; 40(3): 2 09 259. [113] Laszlo K, Josepovits K, Tombacz E. Analysis of active sites on synthetic carbon surfaces by various methods. Anal Sci 2001; 17: i1741 i1744. [114] Moskowitz B. Hitchhikers guide to magnetism. 1991; Available at: http://www.irm.umn.edu/hg2m/hg2m_index.html Accessed 9/18, 2009. [115] Looney BB, Denham ME, Vangelas KM, Bloom NS. Removal of mercury from low concentration aqueous streams using chemical reductio n and air str ipping. J Environ Eng A SCE 2003; 129(9): 819 825. [116] Balogh SJ, Nollet YH. Mercury mass balance at a wastewater treatment plant employing sludge incineration w ith offgas mercury control. Sci Total Environ 2008; 389(1): 125 131.

PAGE 127

127 [117] Rauret G, Lopez S anchez JF, Sahuquillo A, Barahona E, Lachica M, Ure AM, et al. Application of a modified BCR sequential extraction (three step) procedure for the determination of extractable trace metal contents in a sewage sludge amended soil reference material (CRM 483) complemented by a three year stability study of acetic acid and EDTA extractable metal content. J Environ Monit 2000; 2(3): 228 233. [118] Filgueiras AV, Lavilla I, Bendicho C. Chemical sequential extraction for metal partitioning in environmental solid samples. J Environ Monit 2002; 4(6): 823 857. [119] Hall G, Gauthier G, Pelchat J, Pelchat P, Vaive J. Application of a sequential extraction scheme to ten geological certified reference materials for the determination of 20 elements. J Anal At Spectrom 1996; 11(9): 787 796. [120] Kim CS, Bloom NS, Rytuba JJ, Brown GE. Mercury speciation by X ray absorption fine structure spectroscopy and sequential chemical extractions: A comparison of speciation methods. Environ Sci Technol 2003; 37(22): 5102 5108. [1 21] Fergusson J. The heavy elements: Chemistry, environmental impacts, and health effects. Oxford: Pergamon Press; 1990. [122] Tessier A, Campbell PGC, Bisson M. Sequential extraction procedure for the speciation of particulate trace metals. Anal Chem 197 9; 51(7): 844 851. [123] Gubbins D, Herrero Bervera E, SpringerLink (Online service). Encyclopedia of geomagnetism and paleomagnetism [electronic resource]. Dordrecht: Springer; 2007. [124] Zhang F, Itoh H. Adsorbents made from waste ashes and post consu mer PET and their potential utiliza tion in wastewater treatment. J Hazard Mater 2003; 101(3): 323 337. [125] Annadurai G, Juang R, Lee D. Use of cellulose based wastes for adsorption of dyes from aqueous solutions. J Hazard Mater 2002; 92(3): 263 274. [1 26] Ho Y, McKay G. The sorption of lead(II) ions on peat RID C 8624 2009. Water Res 1999; 33(2): 578 584. [127] Ho Y, McKay G. Pseudo second order model for sorption processes RID C 8624 2009. Process Biochem 1999; 34(5): 451 465. [128] Ho Y. Review of s econd order models for adsorp tion systems RID C 8624 2009. J Hazard Mater 2006; 136(3): 681 689.

PAGE 128

128 [129] Nan Q, LeVan M. Adsorption equilibrium modeling for water on activated carbons. Carbon 2005; 43(11): 2258 2263. [130] Li L, Quinlivan P, Knappe D. Effe cts of activated carbon surface chemistry and pore structure on the adsorption of organic contaminants from aqueous solution. Carbon 2002; 40(8): 2085 2100. [131] Mahajan O, Moreno Castilla C, Walker P. Surface treated activated carbon for removal of phen ol from water. Sep Sci Technol 1980; 15(10): 1733 1752. [132] Lee W, Reucroft P. Vapor adsorption on coal and wood base d chemically activated carbons: surface oxidation states and adsorption of H 2 O. Carbon 1999; 37(1): 7 14. [133 ] Ranganathan K. Adsorp tion of hg(II) ions from aqueous chloride solutions using powdered activated carbons. Carbon 2003; 41(5): 1087 1092. [134 ] Goel J, Kadirvelu K, Rajagopal C. Mercury (II) removal from water by coconut shell based activated carbon: Batch and column studies. Environ Technol 2004; 25(2): 141 153. [ 135 ] Khalkhali RA, Omidvari R. Adsorption of mercuric ion from aqueous solutions using activated carbon. Polish J Environ Studies 2005; 14(2): 185 188. [1 36 ] Uchimiya M, Chang S, Klasson KT. Screening biochars for heavy metal retention in soil: Role of oxygen functional groups. J Hazard Mater 2011; 190(1 3): 432 441.

PAGE 129

129 BIOGRAPHICAL SKETCH Emily Kaye Faulconer is the daughter of Charles and Susan Faulconer born in Galax, VA in 1982. The family relocated to Lynchbu rg, VA in 1984. Following the footst eps of both her mother and her u ncle, Jim Davis, Emily graduated from Shenandoah Valley Academy in 199 degree from Central Virginia Community College in 2002 and continued on to Virginia Commonwealth University, graduati degree in Forensic Science in 2004. After working as a certifying scientist at a forensic drug testing facility and teaching high school sciences for three years, Emily began her pursuit of an Environmental Engineering Sciences graduate degree at the University of Florida in 2008, under the guidance of David Mazyck, Ph.D. She received her Ph.D. in Environmental Engineering Sciences in the spring of 2012.