USF Libraries
USF Digital Collections

Removing arsenic from landfill leachate in batch reactors with kemiron adsorbent, a commercially available iron oxide

MISSING IMAGE

Material Information

Title:
Removing arsenic from landfill leachate in batch reactors with kemiron adsorbent, a commercially available iron oxide
Physical Description:
Book
Language:
English
Creator:
Oti, Douglas
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Adsorption
Langmuir
Freundlich
Diffusion coefficient
Electron potential
Dissertations, Academic -- Civil and Environmental Engineering -- Doctoral -- USF   ( lcsh )
Genre:
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: This research evaluated the effectiveness of a commercially available adsorbent, Kemiron, to remove arsenic from conditions representative of landfill leachate. Kemiron was identified as an iron oxide of 39.8 m2/g surface area, 44 % of which resided in the less than 3 microns pore size range. Batch experiments of As(V) and As(III) were conducted with particle sizes either ≤38 microns and in the range 500 - 600 microns with equilibrium being reached in the smaller particles in ~ 36 hours and estimated at 374 hrs for the larger particles. Ionic strength did not affect the mass loadings of As(V) and As(III) which approached 80 mg/g sorbent and greater than 90 mg/g respectively at pH 7. The effect of Se(IV) and Ni(II) was greater on As(III) than on As(V) sorption with as much as a 40% reduction in As(III) sorption in the presence of a similar amount of Se(IV). Sulfate, calcium and carbonate reduced As(III) sorption whereas calcium enhanced As(V) sorption. As removal tested in synthetic landfill leachate under both young and old landfill conditions indicated that pH, ORP, and Se(IV) as a co-contaminant with 1:1 mg/L concentration to As were the most significant key factors that influence As adsorption. Over 90% of 5 mg/L As(V) as initial concentration was removed at pH 7.2 within an operating range of 197 and 371.6 mV of ORP and 99% removal was alsoachieved at ~ pH 11 under the range of -335.7 and 9.1 mV of ORP where the latter condition would be unlikely in real leachate. Preliminary experiments with real leachate solutions show similar sorption behavior for As(V) though the total amount removed was reduced. Whilst this work shows the potential for sorption technology as a treatment option for heavy metal removal from landfill leachate, further tests are definitely needed to determine the various pre-treatment options needed before real leachate solutions can be treated. Many commercially available sorbents have been developed for contaminated drinking waters and this is the first study that has looked at their application to the more complex leachate matrix.
Thesis:
Dissertation (Ph.D.)--University of South Florida, 2009.
Bibliography:
Includes bibliographical references.
System Details:
Mode of access: World Wide Web.
System Details:
System requirements: World Wide Web browser and PDF reader.
Statement of Responsibility:
by Douglas Oti.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains X pages.
General Note:
Includes vita.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
usfldc doi - E14-SFE0003268
usfldc handle - e14.3268
System ID:
SFS0027584:00001


This item is only available as the following downloads:


Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 22 Ka 4500
controlfield tag 007 cr-bnu---uuuuu
008 s2009 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0003268
035
(OCoLC)
040
FHM
c FHM
049
FHMM
090
XX9999 (Online)
1 100
Oti, Douglas.
0 245
Removing arsenic from landfill leachate in batch reactors with kemiron adsorbent, a commercially available iron oxide
h [electronic resource] /
by Douglas Oti.
260
[Tampa, Fla] :
b University of South Florida,
2009.
500
Title from PDF of title page.
Document formatted into pages; contains X pages.
Includes vita.
502
Dissertation (Ph.D.)--University of South Florida, 2009.
504
Includes bibliographical references.
516
Text (Electronic dissertation) in PDF format.
538
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
3 520
ABSTRACT: This research evaluated the effectiveness of a commercially available adsorbent, Kemiron, to remove arsenic from conditions representative of landfill leachate. Kemiron was identified as an iron oxide of 39.8 m2/g surface area, 44 % of which resided in the less than 3 microns pore size range. Batch experiments of As(V) and As(III) were conducted with particle sizes either ≤38 microns and in the range 500 600 microns with equilibrium being reached in the smaller particles in ~ 36 hours and estimated at 374 hrs for the larger particles. Ionic strength did not affect the mass loadings of As(V) and As(III) which approached 80 mg/g sorbent and greater than 90 mg/g respectively at pH 7. The effect of Se(IV) and Ni(II) was greater on As(III) than on As(V) sorption with as much as a 40% reduction in As(III) sorption in the presence of a similar amount of Se(IV). Sulfate, calcium and carbonate reduced As(III) sorption whereas calcium enhanced As(V) sorption. As removal tested in synthetic landfill leachate under both young and old landfill conditions indicated that pH, ORP, and Se(IV) as a co-contaminant with 1:1 mg/L concentration to As were the most significant key factors that influence As adsorption. Over 90% of 5 mg/L As(V) as initial concentration was removed at pH 7.2 within an operating range of 197 and 371.6 mV of ORP and 99% removal was alsoachieved at ~ pH 11 under the range of -335.7 and 9.1 mV of ORP where the latter condition would be unlikely in real leachate. Preliminary experiments with real leachate solutions show similar sorption behavior for As(V) though the total amount removed was reduced. Whilst this work shows the potential for sorption technology as a treatment option for heavy metal removal from landfill leachate, further tests are definitely needed to determine the various pre-treatment options needed before real leachate solutions can be treated. Many commercially available sorbents have been developed for contaminated drinking waters and this is the first study that has looked at their application to the more complex leachate matrix.
590
Advisor: Maya Trotz, Ph.D.
653
Adsorption
Langmuir
Freundlich
Diffusion coefficient
Electron potential
690
Dissertations, Academic
z USF
x Civil and Environmental Engineering
Doctoral.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.3268



PAGE 1

Removing A rsenic from L andfill L eachate in Batch Reactors with Kemiron Adsorbent a C ommercially A vailable Iron Oxide by Douglas Oti A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Civil and Environmental Engineering College of Engineering University of South Florida Major Professor: Maya Trotz, Ph.D. Daniel Yeh, Ph.D. Mark Stewart, Ph.D. Norma Alcantar, Ph.D Jeffrey Cunningham, Ph.D Date of Approval: Ju ly 13 2009 Keywords: Adsorption, Langmuir, Freundlich, Diffusion coefficient, Electron potential. Copyright 2009, Douglas Oti

PAGE 2

This dissertation is dedicated to my wife Joyce, and to my two daughters; Aba and Effia It is also dedicated to my mum, my dad and to all my siblings.

PAGE 3

Acknowledgments First of all I would like to acknowledge my wife, Joyce, whose deep love got me through the times associated with doing my Ph.D. Her support has been unwavering and priceless. My two daughters, Aba and Effia have always been, and continue to be my source of joy. I als o wish to express my sincere appreciation to Dr. M. Trotz whose capacity as a Principal Advisor has been of great value. I would like to thank my committee members Drs. J effrey Cunningham Daniel Yeh Norma Alcantar, and Mark Stewart whose thoughtful suggestions and encouragement have been a significant motivator for making this possible. Finally, I wish to thank Joniqua Howard, Erlande Omisca, and Ken Thomas and all those who worked in Dr. Trotzs laboratory while I was there for their help and suggesti ons on my work. I again express my appreciation to all my colleagues, especially Roland Okwen, Mr. Jean Cobbold, Mr. Gilbert Koume and to their families. Thanks also to Mr. Ackah for his priceless help throughout this process. This work was supported by Florida Center for Solid and Hazardous Waste Grant No. 1933082.

PAGE 4

i Table of Contents List of Tables iv List of Figures vii Abstract x Chapter 1 Introduction 1 1.1 Motivation 1 1.2 Research Goal and Objectives 6 Chapter 2 Background 9 2.1 Introduction 9 2.2 Landfill L eachate C haracterization 9 2.3 Effect of Waste Composition on Landfill Leachate Characteristics 11 2.4 Effect of Age on Landfill Leachate Characteristics 12 2.5 Effect of pH on Landfill Leachate Characteristics 13 2.6 Effect of O xidation R eduction P otential (ORP) on L andfill L eachate C ha racteristics 14 2.7 Arsenic in Landfill Leachate 14 2.8 Chemistry of Arsenate, Arsenite, Selenite and Other Chemical Constituents 16 2.9 Iron Oxide/Hydroxide Surface Chemistry 20 2.10 Arsenic Adsorption Studies 21 Chapter 3 Theoretical Considerations of Adsorption 23 3.1 Introduction 23 3.2 Adsorption and Adsorption Isotherms 23 3.2.1 Langmuir Model 26 3.2.2 Freundlich Model 29 3.3 Hydroxide Surfaces and Ionic Adsorbate 30 3.4 ORP and Eh 3.5 Batch Kin etics Studies 31 Measurements 30 3.6 Experimental Data Fitting 35 Chapter 4 Materials and Methods 36 4.1 Introduction 36 4.2 Materials 36 4.2.1 Adsorbent 36 4.2.2 Reagents and Stock Solutions 37

PAGE 5

ii 4.2.3 Instrumentation 38 4.3 Methods 39 4.3.1 Batch Adsorption Characterization 40 Chapter 5 Results and Discussion 44 5.1 Introduction 44 5.2 Kemiron Surface Characterization 44 5.2.1 Scanning Electron Microscopy (SEM) 49 5.2.2 X Ray Diffractometry (XRD) 50 5.2.3 Electron Dispersion Spectroscopy (EDS) 51 5.3 Kinetics of Arsenic Adsorption in Binary System 52 5.4 Modeling Rate of Arsenic Adsorption 56 5.4.1 Diffusion Coefficient Estimation 57 5.4.2 Effect of Arsenic Concentration on Diffusion 62 5.5 Batch Equilibrium Sorption of Arsenic 65 5.6 Arsenic Adsorption Isotherms 68 5.7 Effect of Presen ce of Competing Ions and CoContaminants 75 5.8 Impact of Oxidation Reduction Potential (ORP) on As(V) Adsorption 82 5.9 Batch Equilibrium Sorption of Arsenic onto Kemiron in Landfill Leachate 83 5.9.1 Effects of Landfill Age and pH on Ads orption 85 5.9.2 Effect of Se(IV) Present in the Landfill Leachate 87 5.9.3 Effect of Ca2+ 5.9.4 Effect of ORP (E on Arsenic Removal in Landfill Leachate 88 h Landfill Leachate 88 ) on Arsenic Adsorption in Synthetic 5.10 Effect of H ydrogen S ulfide on Arsenic A dsorption 91 5.11 Kinetics of Arsenic in Landfill Leachate 94 5.11.1 As(V) Diffusion Coefficient Estimation in 5.12 Maximum As Removal onto Particle Size in Landfill Leachate 94 Landfill Leachate 96 Chapter 6 Summary, Conclusion, and Recommendation for Future Research 100 6.1 Introduction 100 6.2 Summary 100 6.3 Conclusion 104 6.4 Recommendations for Future Research 104 References 106 Appendices 116 Appendix A Mercury Porosimetry Results 117 Appendix B N 2 Appendix C Non Linear Regression Analysis of Isotherm Data 123 (g) Porosimetry Data for 121 Appendix D Output of Geochemical Impact of HSon Leachate Solution 126

PAGE 6

iii Appendix E Raw Experimen tal Data 131 Appendix F Non Linear Regression of Freundlich Isotherm 141 Appendix G Non Linear Regression of Langmuir Isotherm 142 About the Author End P age

PAGE 7

iv List of Tables Table 1.1 Arsenic and its related health effects 2 Table 2.1 Landfill leachate characteristic parameters (Reitzel et al. 1992, Poland and Harper 1989) 10 Table 2.2 Thermodynamic constants of As(III), As(V), Se(IV), CO3 2 and others 17 T able 2.3 Speciation reaction of iron hydroxide 20 Table 4.1 Synthetic land fill leachate constituents 38 Table 5.1 Properties of Kemiron particles 45 Table 5.2 BET surface areas reported on some iron based adsorbents 51 Table 5.3 G rain sizes and intraparticle diffusion rate constants of As removal 62 Table 5.4 Intraparticle diffusion coefficients of As removal onto other iron oxide 62 Table 5.5 Effect of initial As(V) concentration on mass loadings 63 Table 5.6 Isotherms of As adsorption onto various adsorbents 69 Table 5. 7 Isotherm parameters of As onto 38 m Kemiron particles 75 Table 5. 8 The impact of the various factors on the fractions of As adsorbed 83 Table 5.9 As concentrations in landfill leachate sampled from Polk County North Central landfill on 4/27/06. 84 Ta ble 5.10 Concentrations of some of the contaminants in the leachate 84 Table 5.11 Maximum adsorption densities of As 97 Table 5.12 As loadings at equilibrium/ breakthrough 98 Table A .1 Cumulative pore area and pore size distribution 117

PAGE 8

v Table A .2 Cumulative pore volume and pore size distribution 118 Table B .1 BET surface area input report ( 121 Table B .2 Relative pressure isotherm tabular report ( 121 Table B .3 BET surface area output report ( m grain size) 121 Table B .4 BET isotherm result ( 122 Table B .5 Cumulative pore volume result (500 600 m grain size) 122 Table C .1 Nonlinear regression fit to Langmuir isotherm model at pH 9 123 Table C .2 Nonlinear regression fit to Langmuir isotherm model at pH 8 123 Table C .3 Nonlinear regression fit to Langmuir isotherm model at pH 7 124 Table C .4 Nonlinear regression fit to Freundlich isotherm model at pH 9 124 Table C .5 Nonlinear regression fit to Freundlich isotherm model at pH 7 125 Table C .6 Nonlinear regression fit to Freundlich isotherm model at pH 6 125 Table D .1 Summary of input and output data 126 Table D .2 Species with respective concentrations generated 126 Table D .3 The initial input species with respective concentrations 129 Table D .4 Gases with respective fugacities generated 130 Table E.1 As(V) sorption data on 0.1 g/L Kemiron 131 Table E.2 5 ppm As(III) sorption data on 0.1 g/L Kemiron 132 Table E.3 5 ppm As(III) Isotherm data on 0.1 g/L Kemiron, I = 0.01 N NaNO3 133 Table E.4 5 ppm As(V) Sorption on 0.1 g/L Bayoxide, I = 0.01 N NaNO3 134 Table E.5 5 ppm As(V) Sorption to 0.1 g/L Kemiron in the presence of Se(IV), I = 0.001 N NaNO3 135 Table E.6 5 ppm As(V) S orption to 0.1 g/L Kemiron in the pres ence of Ca2+, I = 0.001 N NaNO3 136

PAGE 9

vi Table E.7 5 ppm As(V) sorption to 0.1 g/L Kemiron in the presence of CO3 2 -, I = 0.001 N NaNO3 137 Table E.8 5 ppm As(V) sorption to 0.1 g/L Kemiron in the presence of S O4 2 I = 0.001 N NaNO3 138 Table E.9 5 ppm As(V) or As(III) sorption to 0.1 g/L Kemiron in the presence of 5 ppm Ni 139 Table E.10 5 ppm As(V) sorption to 0.1 g/L Kemiron ( as a function of pH and ORP in a synthetic landfill leachate solution 140

PAGE 10

vii List of Figures Figure 1.1 Schematic diagram of on site treatment of landfill leachate 4 Figure 2.1 Arsenate speciation diagram with total As(V) of 105 M 16 Figure 2.2 Arsenite speci ation diagram with total As(III) of 105 M 18 Figure 2.3 Se lenite speciation diagram with tot al Se(IV) of 105 M 18 Figure 2.4 Carbonate speciation diagram with total CO3 2 of 105 M 19 Figure 2.5 Ammonia ammonium speciation diagram with total NH4 + of 105 M 19 Figure 3.1 Schematic representation of the adsorption process 25 Figure 4.1 Ultra high pure nitrogen gas sparging setup for a batch system 40 Figure 5.1 Cumulative area mercury porosimetry of 500 600 particle size 47 Figure 5.2 Mercury adsorption isotherm onto 500 600 m particle size of Kemiron 47 Figure 5.3 Nitrogen adsorption isotherm onto 500 600 particle size of Kemiron 48 Figure 5.4 Nitrogen adsorption isotherm onto particle size of Kemiron 48 Figure 5. 5 Scanning Electron Microscope ( SEM ) Microgram of a 500 600 Kemiron particle 49 Figure 5. 6 X Ray diffractogram (XRD) of Kemiron powder ( 50 Figure 5. 7 Energy Dispersive Spectroscopy (EDS) scan of < 38 m Kemiron 52 Figure 5. 8 Rate of uptake of As onto 0.1 g/L Kemiron 54 Figure 5. 9 Fractional mass of As(III) removal onto grain size in a batch system 58

PAGE 11

viii Figure 5.10 Fractional mass of As(V) removal onto grain size in a batch system 58 Figure 5.11 Fractional mass of As(III) removal onto 500 600 grain size in a batch system 59 Figure 5.12 Fractional mass of As(V) removal onto 500 600 grain size in a batch system 59 Figure 5.13 Fractional mass of As removal model in a batch system 60 Figure 5.14 Kinetics of As(V) removal model in a batch system 61 Figure 5.15 Kinetics of As removal model in a batch system 61 Figure 5.16 Effect of initial As(V) concentration on rate of uptake 64 Figure 5.17 Model of Fractional mass of As(V) removal 64 Figure 5.18 Batch equilibration tests of As(V) onto 38 Kemiron grain size 66 Figure 5.19 Batch equilibration tests of As(III) onto 38 Kemiron grain size 66 Figure 5. 20 Batch equilibration tests of both As(V) and As(III) onto 38 m Kemiron 68 Figure 5.21 Effect of pH on As(III) adsorption isotherm in pure system 70 Figure 5.22 Effect of pH on As(V) adsorption isotherm in pure system 71 Figure 5.23 Experimental data and predicted data of As(V) sorption at pH 8 73 Figure 5.24 Experimental data and predict ed data of As(V) sorption at pH 7 74 Figure 5.25 Experimental data and predicted data of As(III) sorption at pH 6 74 Figure 5.26 Effect of Se(IV) or Ni2+ on As( III ) adsorption 76 Figure 5.27 Effect of Se(IV) or Ni2+ on As(V ) adsorption 76 Figure 5.28 Se(IV) sorption as a function of pH 77 Figure 5.29 Effect of CO3 2 or SO4 2 on As(III ) adsorption 79 Figure 5.30 Effect of CO3 2 or SO4 2 on As( V ) adsorption 80

PAGE 12

ix Figure 5. 31 Effect of Ca2+ and NH4 + N on As( III ) adsorption 81 Figure 5.32 Effect of Ca2+ and NH4 + N on As( V ) adsorption 81 Figure 5.33 Adsorption edge of 1 mg/L As(V) on 0.1 g/L Kemiron in a Polk County landfill leachate solution 85 Figure 5.34 Effect of pH or age (acidogenic or methanogenic) of landfill leachate on 5 mg/L As(V) adso rption 86 Figure 5.35 Effect of pH or age of landfill leachate on 5 mg/L As(IIII) adsorption 86 Figure 5.36 E ffect of Se(IV) i n As (V) removal in the synthetic landfill leachate 87 Figure 5.37 Effect of Ca2+ landfill leachate 88 on 5 mg/L As (V) adsorption in synthetic Figure 5.38 The impact of ORP on As removal in synthetic landfill leachate 89 Figure 5.39 Impact of pH and ORP on As removal in synthetic landfill le achate 90 Figure 5.40 Box plot of ORP (mV) as a function of pH 90 Figure 5.41 Eh log { H S} diagram of inorganic arsenic at pH 5 93 Figure 5.42 Eh log {HS} diagram of inorganic arsenic at pH 10 93 Figure 5.43 Eh pH diagram of inorganic arsenic 94 Figure 5.44 Rate of 5 mg/L As(V) removal onto 38 particle size 95 Figure 5.45 Fractional removal model of As onto Kemiron in the synthetic leachate 95 Figure 5.46 Contours of %As sorbed in young synthetic landfill leachate 97

PAGE 13

x Removing Arsenic from Landfill Leachate in Batch Reactors with Kemiron Adsorbent, a Commercially Available Iron Oxide Douglas Oti ABSTRACT This research evaluated the effectiveness of a commercially available adsorbent, Kemiron, to remove arsenic from conditions representative of landfill leachate Kemiron was identified as an iron oxide of 39.8 m2/g surface area, 44 % of which resided in the less than 3 nm pore size range. Batch experiments of As(V) and As(III) were conducted with particle sizes either m and in the range 500 600 m with eq uilibrium being reached in the smaller particles in ~ 36 hours and estimated at 374 hrs for the larger particles. Ionic strength did not affect the mass loadings of As(V) and As(III) which approached 80 mg/g sorbent and greater than 90 mg/g respectively a t pH 7. The effect of Se(IV) and Ni(II) was greater on As(III) than on As(V) sorption with as much as a 40% reduction in As(III) sorption in the presence of a similar amount of Se(IV). Sulfate, calcium and carbonate reduced As(III) sorption whereas calci um enhanced As(V) sorption. As removal tested in synthetic landfill leachate under both young and old landfill conditions indicated that pH, ORP, and Se(IV) as a co contaminant with 1:1 mg/L concentration to As were the most significant key factors that influence As adsorption. Over 90% of 5 mg/L As(V) as initial concentration was removed at pH 7.2 within an operating range of 197 and 371.6 mV of ORP and 99% removal was also

PAGE 14

xi achieved at ~ pH 11 under the range of 335.7 and 9.1 mV of ORP where the latter condition would be unlikely in real leachate. Preliminary experiments with real leachate solutions show similar sorption behavior for As(V) though the total amount removed was reduced. Whilst this work shows the potential for sorption te chnology as a treatment option for heavy metal removal from landfill leachate, further tests are definitely needed to determine the various pre treatment options needed before real leachate solutions can be treated. Many commercially available sorbents ha ve been developed for contaminated drinking waters and this is the first study that has looked at their application to the more complex leachate matrix

PAGE 15

1 Chapter 1 Introduction 1.1 Mo tivation Arsenic (As) contamination in surface and groundwater is a major problem of our time, affecting large populations around the world, including the United States. The outbreaks of As related diseases in Bangladesh and in some parts of West Bengal elevated consciousness of its deleterious health effects like Blackfoot disease (Lamm and Kruse 2005) gastrointestinal disorders, cardiac damage, chronic vascular disorders (Simeonova and Luster 2004) and skin cancer (Rossman et al. 2004) Table 1 .1 summarizes co ncentrations of As detected in various water resources and the health impacts on users in various parts of the world. Other heavy metals also pose health concern s For instance, selenium (Se) is an e ssential nutrient at low levels, but ingestion at concent (Letavayova et al 2006) Arsenic contamination of potable water is due to both natural and man made sources. Landfills are an emerging concern in Florida and other parts of the United States because of the potential leakage of leachate contaminated with arsenic and other contaminants into aquifers (Christensen et al. 2000, Pujari and Deshpande 2005) The leachate from lined landfills is either sent to an external waste water treatment facility, recycl ed back through the landfill, or treated on site. Figure 1.1 depicts a typical lined landfill in which some leachate is recycled through the landfill to assist

PAGE 16

2 Table 1.1: Arsenic and its related health effects Ranges of As concentra tion Place Health effect seen References Unknown cumulative amount 10 less than 20 years and between 20 and 40 years for 100,000 people Southwestern coast of Taiwan Taiwan Blackfoot disease Urinary cancer Tseng (1989) Chiou et al. ( 1995) Unknown cumulative amount Chile Lung cancer and bronchiectasis in young adults Smith et al. (2006) 0.41 mg/L or greater China Induction of oxidative Stress Sugden et al. ( 2004) Unknown cumulative amount Finland Bladder cancer Kurttio et al. ( 1999) Unknown cumulative amount 300 g/L or greater 1 3644 g/L 50 g/L or greater 50 400 g/L for over 20 years 1200 g/L or greater Taiwan and Bangladesh West Bengal, India West Bengal, India Araihazar, Bangladesh West Bengal Bangladesh Diabetes Skin lesions Skin lesions Intellectual impairment of children Bronchiectasis Abdominal pain, vomiting, diarrhea, muscular weakness and cramping. Navas Acien et al. ( 2008) Chowdhury et al. (2000) Rahman et al. ( 2006) Wasserman (2004) Mazumder et al. (2005) Mead ( 2005)

PAGE 17

3 Table 1.1 (continued) Unknown cumulative amount Unknown cumulative amount Unknown cumulative amount 100 g/L or greater New Hampshire (USA) and Sonora (Mexico) Utah, USA West Virginia, USA Nevada and California, USA Decreased DNA Repair in Vitro Increased mortality from hypertensive heart disease, nephritis and nephrosis, and prostate cancer Accelerates atherosclerosis People with diets deficient in protein and other nutrients are more susceptible than others to arsenic caused cancer Andrew et al. ( 2006) Lewis et al. ( 1999 ) Simeonova and Luster (2004) Steinmaus et al. ( 2005) with biodegradation processes.Unlike organic compounds, heavy metals like arsenic do not degrade. An opportunity exists to capture the heavy metals released from a degenerating solid waste into leachate in an onsite treatment step. Such a process would minimize the volume occupied by the heavy metal, making it easier to recycle it or easier to dispose of it in a controlled environment like a more contained, lined landfill cell. Wastewater treatment facilities have limits on the volume of leachate they can process based on the leachat e quality. High concentrations of As content attract surcharges and a presence of high levels of total dissolved solids are sometimes rejected by some treatment facilities. These factors lead to expensive disposal costs for some Florida landfill facilitie s. For example, in 2005, the Polk County Landfill in Lakeland, Florida transported approximately half of its leachate to a treatment facility in Jacksonville at a significant cost to the landfill facility because the local waste water treatment plant did

PAGE 18

4 not have the capacity to accept the high total dissolved solids concentration coupled with the concentrations of toxic metals like arsenic. Recycling of the leachate through the landfill is a potential low cost option for disposal, but the nondegradable nature of heavy metals like arsenic means that the landfill will be a continual source of arsenic, and will always have to be monitored, even after the degradation of toxic organic compounds. Recycling of heavy metals through the landfills may also increase their concentration to levels where microbial activity is significantly reduced due to toxic effects. Figure 1.1: Schematic diagram of onsite treatment of landfill leachate. Collected leachate can be treated onsite via a technology like sorption to mineral oxides in a packed column (Fixed Bed Reactor). Once arsenic is removed, the leachate can either be sent to a wastewater treatment facility for further treatment or recycled through the landfill. Overall, arsenic from the entire site can be collect ed by the packed columns and used appropriately (recycled or disposed of more carefully).

PAGE 19

5 There are onsite treatment methods adopted for leachate treatment so far These methods include precipitation, oxidationsedimentation, coagulationfiltration, reve rse osmosis, and adsorption. However, treatment of As or other heavy metal contamin ants in landfill leachate remains a significant challenge. Adsorption onto mineral oxide sorbents packed into fixed bed reactors is particularly attractive because of the small equipment footprint, efficiency and cost effectiveness This technology is now popular to combat the widespread As contamination of potable water around the world (Wiszniowski et al. 2006), but has not been applied to more complex matrices like landfill leachate T here is extensive background literature on As removal from drinking water sources through sorption to mineral oxides (Bajpai and Chaudhuri 1999, Thirunavukkarasu et al. 2003b, Cincotti et al. 2006, Bang et al. 2005, Zhang and Lindan 2003, Xu et al. 2006, Entezari and Bastami 2006, Agrawal and Sahu 2006) but little is known of the performance of these adsorbents with complex mixtures like landfill leachate. Some investigations on adsorption processes involving other contaminants in landfill leachate have been done For example, NH3 Landfill leachate is mostly characterized by high organic content and high concentrations of inorganic ions. The org anics are measured in terms of C hemical N (Ashrafizadeh et al. 2008, Kargi and Pamukoglu 2003), organic content in the form of B iochemical O xygen D emand (BOD) C hemical O xygen D emand (COD) (Kargi and Pamukoglu 2003, Fan e t al. 2007, Rivas et al. 2003) and T otal O rganic C arbon (TOC) (Fan et al. 2007) among others have been studied in adsorption processes Preliminary results from tests I did with real landfill leachate spiked with As indicated that As in landfill leachate could be removed by mineral oxide surfaces.

PAGE 20

6 Oxygen Demand (COD), or Total Organic C arbon (TOC) (We ber et al. 2002; Wiszniowski et al. 2006, Kuleyin and Ergun 2007, Fan et al. 2006) The concentrations of organics present vary widely with landfill age (Alvarez Vazquez et al. 2004, Cooke et al. 2001) The older landfill leachate generally contains l ower concentration s of organic and inorganic ions than the younger landfill leachate (Alvarez Vazquez et al. 2004, Statom et al. 2004) The pH range has also been found to fall between 5 and 8.5 (Fan et al. 2006) with the lower pH associated with younger landfill leachate and the higher pH related to older leachate. The leachate characteristics differ significantly from the geochemistry of contaminated groundwater drinking water sources, the focus of most arsenic removal technology. 1.2 Research G oal an d O bjectives The goal of this research was to evaluate the effectiveness of a commercially available adsorbent, Kemiron, to remove arsenic from conditions representative of landfill leachate It involve d laboratory batch experiments coupled with modeling and focused on the adsorption capability of Kemiron in a synthetic leachate under various physico chemical conditions. The considered environmental and chemical conditions for the experiments were pH, oxidation reduction potent ial (ORP), ionic strength, C hemical Oxygen D emand (COD), and the presence of co contaminants. Ions like carbonate ( CO3 2 -), sulfate ( SO4 2 -), ammonium ( NH4 + N ) and Ca2+ were evaluated for their impact on arsenic sorption since they represent the most commonly occurring species in landfill leachate or are representative of common types of inorganics in landfill leachate Nickel ( Ni2+) and selenite (Se(IV)) were used as trace co contaminants representing cationic and

PAGE 21

7 anionic type trace metals The specific objectives were: 1) Adsorbent (Kemiron) characterization. This ad sorbent characterization was needed for modeling sorption data and for comparison with published research results of As sorption by other adsorbents The adsorbent character ization included the following experiments and analyses: a) Determination of the BET surface area, particle density, and skeletal porosity of Kemiron. b) Determination of Kemiron composition and mineral identification. 2) As sorption characterization under a range of conditions (pH, ionic strength, presence of other ions (Se(IV), Ni2+, Ca2+, CO3 2 -, SO4 2 -, NH4 + N CH3COO-, C2H5COO-a) Determination of upt ake equilibration time of As(V) and As(III) onto Kemiron ( and in synthetic la ndfill leachate systems at various pH, and at initial solute concentrations. where the last two ions represent COD). This sorption characterization highlight ed the optimum conditions for the As treatment identif ied limitations of the adsorption technology, and provide data needed for modeling. The experiments included: b) Modeling of As(V) and As(III) onto Kemiron in batch systems using isotherm data c) Determination of the effect of co contaminants, represented by Se(IV) and Ni2+Chapter 2 discusses background information on leachate characteristics and geochemistry of the systems under study, Chapter 3 presents the models used to interpret on As removal

PAGE 22

8 experimental data, Chapter 4 summarizes materials and methods used, Chapter 5 presents and discusses experimental results and modeling efforts and Chapter 6 summarizes the major findings of this work and discusses opportunities for further research.

PAGE 23

9 Chapter 2 Background 2.1 Introduction This chapter provides background information on different components of this project. It first reviews the literature on landfill leachate characterization (Sections 2.1 to 2.5) with a special emphasis on some of the factors (type of waste, age of landfill, Oxidation Reduction Potential, pH) that contribute to general leachate composition. It then discusses the relevance of this work by placing it in the context of landfills in the State of Florida, which is by no means the only geographic location where the presence of arsenic in leachat e either is, or will be an issue. Section 2.7 describes arsenic chemistry and provides the thermodynamic constants and construct used to interpret experimental data of the work. Sections 2.8 and 2.9 introduce the concept of the mineral oxide adsorbent, es pecially previous research on arsenic interactions with and removal by mineral oxide adsorbents. 2.2 L andfill L eachate C haracterization Kjeldsen et al. (2002) c haracterized landfill leachate into four major components : 1) inorganic macro components, including cations like magnesium, calcium, iron and anions like bicarbonate, sulfate, chloride and phosphate; 2) heavy metals like arsenic, cadmium, selen ium and many others; 3) dissolved organic matter, usually expressed as

PAGE 24

10 Table 2.1: Landfill leachate characteristic parameters (Reitzel et al. 1992, Pohland and Harper 1989). Parameter Purpose Physical pH Acid base/stabilization phase indicator ORP Oxidation Reduction/stabilization phase indicator Conductivity Ionic strength/activity indicator Temperature Reaction indicator Chemical COD, TOC, TVA Substrate indicator TKN, NH 3 N, PO 4 3 Nutrient indicator /P SO 4 2 /S, NO 3 /NH Stabilization phase indicator 3 TS, Chloride Dilution/mobility indicator Total alkalinity Buffer capacity indicator Alkali/alkaline earth metals Toxicity/environmental effect Heavy metals Toxicity/environmental effect Biological BOD Substrate/biodegradability 5 Total/faecal coliforms Health effect indicators Faecal streptococci Health effect indicator Viruses Health effect indicator Pure/enrichment culture Stabilization phase indicators

PAGE 25

11 biochemical oxygen demand (BOD), chemical oxygen demand (COD) and total organic carbon (TOC); and 4) anthropogenic or ganic contaminants including xenobiotics Whil e Christensen et al. (1994) disregarded the relevance of pathogens in landfill leachate characterization, Mor (2006) and others considered the presence of fecal in dicator bacteria as relevant, but whose number decreased with increasing landfill leachate age. Reitzel et al. (1992), Pohland and Harper (1989) characterized landfill leachate with the parameters shown in T able 2.1. The full meanings of the acronyms in the Table 2.1 can be found in Appendix D. Most authors characterize landfill leachate with fewer than the parameters in Table 2.1 using only BOD, COD, TOC, BOD/COD ratio pH, ammonium nitrogen (NH4 + N), total Kjeldahl nitrogen (TKN) and heavy metals. The results of these parameters depend on the following: 1) constituting waste composition (Kargi and Pamukoglu 2003, Weber et al. 2002), 2) age of the landfill (Alvarez Vazquez et al. 2004) 3) amount of precipitation or moisture content in the landfill (Renou et al. 2008), 4) the presence of active microorganisms (Kargi and Pamukoglu 2003), 5) Oxidation reduction potential (ORP) in the landfill (Bayard et al. 2006), and 6) pH of the landfill (Pokhrel and Viraraghavan 2004). Some of these contributing factors are discussed further in sections 2.2 to 2.5. 2.3 Effect of Waste C omposition on L andfill L eachate C haracteristics Landfill leachate characteristics are reflections of landfill waste composition (Salem et al. 2008, Xiao et al. 2007) Given that landfill leachate is characterized by parameters like BOD, COD, and NH3 N, and the concentrations of organic and inorganic species, the values of these parameters depend on proportional compositions of

PAGE 26

12 the parent waste material (Duggan 2005, Mor, 2006). Despite the variability of the waste compositions and their subsequent variations in the proportions of the characteristic parameters, TOC (particularly cellulosic material) mostly constitutes the largest percentage (Boni et al. 2006). The knowledge of leachate composition of a landfill enabled Kjeldsen e t al. (1998) to characterize the expected chemical composition of the leachate in time and space in the United States Contents and concentrations of organic subs tances in landfill leachate in many cases, determine the nature of pretreatment processes need ed to be undertaken. The wastes in hazardous landfills (generally classified as Class 1 Landfills requiring liners and leachate collection systems) are quite different from that of non hazardous municipal solid waste landfill s Consequently the content of their le achate differ from each other, especially in the levels of toxic elements and compounds that each produce. 2.4 Effect of A ge on L andfill L eachate C haracteristics The age of landfill leachate from municipal, mixed industrial and non hazardous commercial solid waste impacts values of BOD, COD, TOC, and NH3N (Alvarez Vazquez et al. 2004, Kjeldsen et al. 2002) Young landfill leachate (< 1 2 years old) are normally dominate d by organics of lower molecular weights that generally have high values of BOD and COD (Zhang and Selim 2005). As landfill leachate matures organics of higher molecular weight ( i.e. COD) dominates while BOD drops in value. This is due to biodegradation of the lower molecular weight organics. The biodegradation continues until the contaminant constituents become resistant to biodegradation or are simply no longer degraded by microbes under the conditions existing at that time BOD and COD

PAGE 27

13 both degenerat e as leachate ages, h owever, the rate of the degeneration is higher with BOD than with COD. Al Yaqout and Hamoda (2005) showed that the average BOD/COD of a younger landfill leachate in Kuwait was 0.13 while the BOD/COD of older landfill leachate was 0.04 Generally, a BOD/COD of 0.5 of landfill leachate indicates a young age of the leachate while BOD/COD of 0.1 and lower points to older and stable landfill leachate (El Fadel et al. 2002). 2.5 Effect of pH on L andfill L eachate C haracteristics According to Renou et al. (2008), pH of landfi ll leachates usually lies between 5.8 and 8.5. The differences in pH of landfill leachate are related to different levels of biological (aerobic or anaerobic) activities inside the landfills (Poulsen et al. 2002). Salem et al. (2008) r elated the pH changes of landfill leachate to biochemical evolution occurring in the landfill. Poulsen et al. (2002) discovered that pH of a landfill leachate decreased in the first 3 to 5 months, when the high concentration of oxygen was consumed to produce high and increasing concentrations of leached material. In a situation when oxygen is limited, the landfill undergoes acidogenesis to produce high concentrations of organics (BOD and COD), CO2 with increasing Cland NH3/NH4 +, leading to a drop in pH. In an aerobic environment, the pH remained constant to about the 50th month and rises to 8 at about the 150th month. The pH r emains high thereafter and coincides with the growth of a methanogenic microbial population that creat e s high levels of methane and leaching out of heavy metals (Poulsen et al. 2002).

PAGE 28

14 2.6 Effect of O xidation R eduction P otential (ORP) on L andfill L eachate C haracteristics ORP characterizes the strength of an oxidizing agent (electron acceptor) or reducing agent (electron donor). Some examples of oxidizers include chlorine, hydrogen peroxide, ozone, bromine, and hypochlorite and examples of reducers include sodium sulfite, sodium bisulfate, and hydrogen sulfide. ORP measurement in soils and rocks has been used to detect l eakage of landfill leachate in the subsurface and ORP gradients in soils/rocks by leachate migrations have resulted in remobilization of ions in the neighboring rocks (Christensen et al. 2000). Studies have also indicated that As a nd Fe species are sensitive to ORP in their environments. For instance, As(III) and Fe2+ are dominant in reduced environments while As(V) and Fe3+ are prevalent in oxidized environments. Thus high Fe3+/Fe2+ and As(V)/As(III) ratios in a landfill leachate is indicative of higher values of ORP of landfill. In methanogenic, sulfate reducing, and iron reducing landfill leachate plumes, ORP consistently results between 70 and 100 mV while in aerobic plumes ORP yields readings of 200 to 300 mV (Christensen et al. 2000). 2.7 Arsenic in L andfill L eachate Arsenic (As) contamination in landfill leachate is a major concern in Florida and many other parts of the United States. In 2005, an attempt was made to contact 68 Florida landfills (not only active Class 1 landfills) via email and phone to learn about their leachate disposal practices and total arsenic concentrations. Of the 68 landf ills on the list, 26 responses were obtained and of those 26 responses 7 landfills in Florida had

PAGE 29

15 leachate with arsenic concentrations greater than 10 g/L combined with a disposal issue related to arsenic. The seven landfills identified were: Alachua C ounty Lake County Marion County Martin County Orange County Polk County Santa Rosa County These seven landfills paid for offsite leachate disposal and sometimes had an additional surcharge fee because arsenic concentrations were above p ermissible limits. Though leachate contained a list of other heavy metals, arsenic concentrations were closer to or above permissible limits. A r s enic inorganically exists as As(III) and As(V) and organically as methylarsonic acid [ MMA(V) ] dimethylarsinic acid [ DMA(V) ] methylarsonous acid [ MMA(III) ] and dimethylarsinous acid [ DMA(III)] (Sierra Alvarez et al. 2005). As(III) converts to As(V) in oxidizing (oxygenrich) environments and As(V) reduces to As(III) in reducing ( anaerobic) environments where As(V) acts as an electron acceptor (Rittmann and McCarty 2001) The methylated organoarsenical s [ MMA(III)] [ DMA(III) ] [ MMA(V)], and [DMA(V)] occur from biotransformation of the inorganic arsenicals. As(III) is more mobile and more toxic than arsenate As(V) and its toxicity has been linked to the fact that the human skin contains several sulfhydryl groups to which As(III) binds (Maji et al. 2007) Sierra Alvarez et al. 2005 have postulated the following t hree

PAGE 30

16 mechanisms by which As leach es out from solid waste in landfills under aerobic conditions: 1) oxidation of metal sulfides to the more soluble metal sulfates 2) oxidation of metal sulfides to sulfuric acid which results in pH reduction, hence dissolving the metals on contact, and 3 ) co mplexation of metals with humic acid leading to metal mobilization Most metals or metalloids in solid waste landfills leach out to the highest degree at pH 3 and below (Moghaddam and Mulligan, 2008). This pH range, however, is not necessary for arsenic, or other heavy metal mobilization in landfill leachate. 2.8 Chemistry of A rsenate, A rsenite, S elenite and O ther C hemical C onstituents Equilibrium constants are used to predict and understand speciation of chemicals in aqueous environments as a function of solution pH and pe, where pe applies onlt to redox sensitive ions. Table 2.2 provides a list of log K values for some of the chemicals used in this study and Figure s 2.1 2.5 show the dist ribution of various species of relevant to this with a function of pH and/or pe Geochemical Workbench and NIST software were used to calculate the se equilibrium speciation diagrams using the (Gimenez et al. 2007, Jones and Pichler 2007) Figure 2 .1: Arsenate speciation diagram with total As (V) of 105 0 2 4 6 8 10 12 14 0 0.2 0.4 0.6 0.8 1 x 10-5 pHConcentration (g/L)H3AsO4H2AsO4HAsO4-AsO4--M

PAGE 31

17 Table 2.2: Thermodynamic constants for As(III), As(V), Se(IV), CO3 2 and others Assuming Ionic strength = 0, T = 25o Formation Reaction C. p K H 2 AsO 3 + H + = H 3 AsO HAsO3 3 2 + H+ = H2AsO3AsO3 3 + H+ = HAsO3H2 2AsO4 + H+ = H3AsO HAsO4 4 2 + H+ = H2AsO4AsO4 3 + H+ = HAsO4SeO2 3 2 + H+ = HSeO3HSeO3 + H+ = H2SeO HCO3 3 + H+ = H2CO CO3 3 2 + H+ = HCO3CH3COO+ H+ = CH3NH COOH 4 + + OH= NH 9.32 3 12.13 13.41 2.22 7.00 11.54 8.40 2.63 6.35 10.33 4.76 9.25 In the acidic and alkaline regions of Figures 2.1 2.2, the dominant As(V) species is H2AsO4 and HAsO4 2 -, respectively. It is often assumed that the dominant solution species is also the dominant adsorbing species As a result several authors have proposed mechanisms involving one or both of these species for the adsorption of arsenic onto various minerals (Villalobos et al. 2003, Goldberg 2002) Similarly H3AsO3 and H2AsO3 predominate all the other species of As(III) in acidic and alkaline regions. For Se(IV) in Figure 2.3, HSeO3 and SeO3 2 dominate i n acidic and alkaline regions respectively. Whilst As(V) and Se(IV) exist as charged species above pH 2.22 and 2.63

PAGE 32

18 respectively, As(III) remains a fully unspeciated until pH 9.32 and above In a reducing environment, As exists pred ominantly as As(III), while in an oxygen rich environment arsenic exists as As(V). Figures 2.4 and 2.5 show carbonate and ammonium speciation respectively. Figure 2.2: Arsenite speciation diagram with total As (III) of 105 M Figure 2.3: Selenite speciation diagram with total Se(IV) of 105 6 7 8 9 10 11 12 13 14 0 0.2 0.4 0.6 0.8 1 x 10-5 pHConcentration (g/L)H3AsO3 H2AsO3HAsO3-M 2 4 6 8 10 12 14 -22 -20 -18 -16 -14 -12 -10 -8 -6 -4 pHLog of ConcentrationH2SeO3 HSeO3SeO3--

PAGE 33

19 Figure 2.4: Carbonate speciation diagram with total CO3 2 of 105 M Figure 2.5: Ammonia ammonium speciation diagram with total NH4 + of 105 M Like the As(V) species carbonate would have a negative charge in solution and hence could compete with Asfor the surface sites. Ammonium (NH4+ 0 2 4 6 8 10 12 14 -22 -20 -18 -16 -14 -12 -10 -8 -6 -4 pHLog of ConcentrationH2CO3 HCO3CO3-) on the other hand would be positively charged across most of the pH range studied. The speciation diagrams allow us to interpret experimental data and this type of analysis will be invoked in chapter 5. 0 2 4 6 8 10 12 14 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 pHLog of ConcentrationNH4+ NH3(g)

PAGE 34

20 2.9 Iron O xide/ H ydroxide S urface C hemistry Surfaces of hydrated iron oxide are usually assigned +1 and 1 charges. These charges can be neutralized by binding of OHto + to the Again, hydrated as group in solution may be protonated as 2 +, neutral as or deprotonated to form species depending on the pH of the solution. The surface acidity reaction can be written as shown in table 2.3 below. Table 2.3: S peciation reaction of i ron h ydroxide Formation Reaction Log K + = 2 + + H+ 7 = 9.2 From Table 2.3, 2 + dominates below the point of zero charge (pHpzc) while dominates at above the pHpzc. Many iron based adsorbents have their pHpzc between 8 and 9.5 (Naeem et al. 2007; Smith 1998; Sperlich et al. 2005). Villalobos et al (2003) however found that surface area determination method particle siz ing procedure and adsorbent pretreatment methods affected the p Hpzc of goethite Thus a universal pHpzc value for sorbents may introduce errors in surface modeling attempts. For the most part, however, the neutral FeOH dominates iron species in water at pH 8. In adsorption studies, the amount of anions sorbed decreased as a function of pH. This has been explai ned with electrostatic influence. The charges on iron hydroxide surfaces are neutral at pHpzc An increase in pH induces negative charges on the adsorbent surface and this repels As species, a negatively charged adsorbate species.

PAGE 35

21 2.10 A r s enic A dsor ption S tudies Iron based sorbents are used as adsorption media because of their net positive surface charge at low pH and again because the active surfaces form bonds with many anions. Batch equilibrium sorption of As, and Se(IV) onto iron oxide/hydroxide has been extensively studied. As(V) and Se(IV) are most effectively adsorbed at lower pH, while As(III) and Ni2+ adsorption are r ather higher at higher pH level. This is because the pH at which oxides of iron possess zero charge (i.e., pHPZCStudies also show that adsorption of As onto iron oxide/hydroxide is either unaffected by increasing ionic strength or decreases with increasing ionic strengt h (Zhang et al. 2007, Payne and Abel Fattah 2005) Typically, As(III) is more sensitive to changes in ionic strength than As(V). Several spectroscopic investigations regarding the binding of As to various solid surfaces have led to a proposed inner spher e (ligand exchange) type reaction mechanisms for As sorption onto adsorbent surfaces Kanel et al. (2005) described adsorption mechanism of As(III) onto zero valent iron as inner sphere. Again i nner sphere mechanism has been reported on As(III) adsorption onto gibbsite (Oliveira et al. 2006) Duc et al. (2006) have also shown that ionic strength have no effect on the adsorption of selenium onto haematite. Martinez et al. (2006), Catalano et al. (2006) respectively have described the sorption mechanism of Se(IV) onto hematite and magnetite as inner sphere. This supports the inference made by Goldberg (2002) on the ) is generall y between 8.0 and 9.0. At pH below these values the solid surfaces acquire a net positive charge. Consequently, much of the adsorption of As(V) on these surfaces is by electrostatics attraction, while As(III) is sorbed by weak Van der Waal forces. This i s because As(V) speciate into oxyanions at 2.2 and 2.6 respectively but As(III) speciates at 9.3.

PAGE 36

22 description of sorption mechanism of ions onto metal surfaces based on the knowledge of ionic strength responses. Others have demonstrated that anions such as sulfate, phosphate, carbonate, and molybdate may compete with arsenite, and to a lesser degree arsenate, for available surface sorption sites. Consequently, arsenite sorption is signi ficantly hindered in the presence of co adsorbing anions

PAGE 37

23 Chapter 3 Theoretical C onsiderations of Adsorption 3.1 Introduction This chapter will introduce the basic concepts of adsorption and describe two common models used to describe adsorption behavior under equilibrium conditions, the Langmuir and Freundlich isotherms. It then explains the methodology used to calculate and co mpare mass transfer parameters that characterize the rate at which species of interest are removed by the mineral oxide surface. The models described here are used in Chapter 5 to interpret experimental data and provide comparisons with published research on other systems of interest. Finally, Oxidation Reduction Potential (ORP) is explained in a way that relates experimental measurements with theoretical concepts. 3.2 Adsorption and A dsorption I sotherm s Adsorption refers to the accumulation of ions or molecules at the interface between two phases in relation to the concentrations of the ions or the molecules in a bulk solution. Generally, chemical, physical or electrostatic interactions influence the ads orption behavior of inorganic adsorbates onto the surfaces of adsorbents. Chemical interactions include covalent and hydrogen bonding, while electrostatic forces are involved in ion ion and ion dipole interactions. Physical attractions involving relatively weak Van der Waals forces include dipole dipole, dipole induced dipole

PAGE 38

24 interactions. The Van der Waals forces are involved in the sorption of nonionic organic and inorganic molecules from aqueous solutions. Figure 3.1 depicts ads orption processes as they relate to the solidliquid interface. The left hand side of the first scenario, (a), represents a solid (e.g. a mineral oxide) in solution with dissolved ions, where represents sites on the solid surface capable of binding a di ssolved ion. The right hand side of the first scenario depicts the dissolved ion binding to the surface site through the process of adsorption. Scenario (b) depicts the surface sites as hydrated species SH2O and replaces the symbol for dissolved species used in scenario (a) with real species one would expect in a simple solution containing water, sodium nitrate (NaNO3) and some dissolved species A (shown here as an uncharged species). The adsorption of species A to the surface site results in the format ion of SA. Scenario (c) depicts the surface sites with different charges represented as SOH2 +, SOand SOH and adsorbed species as SOA and SO-Na+Isotherms are commonly used to describe equilibrium adsorption behavior To accurately represent the adsorption of an adsorbate over a wide range of conditions, where the difference between the latter two species lies in the strength and type of bond between the adsorbent surface site and the adsorbing species The representations given in Figure 3.1 are only some of the ways in which adsorption processes are conceptualized and are simpler than other models (e.g. Constant Capacitance Model, The Diffuse Layer Model The Triple Layer Model, CD MUSIC Model) which incorporate changes in the electric potential as a function of distance from the surface of the adsorbent and/or require more detailed information on the specific surface site types (Hayes and Leckie 1987; Hiemstra et al. 1989 ; Davis and Kent, 1990)

PAGE 39

25 Figure 3.1: Schematic representation of the adsorption process T he solidwater interface for surface sites, dissolved species and adsorbed species represented as (a) O and, O respectively; (b) SH2O, various species like Na+, H+, A, OH-, and SA respectively; and (c) SOH2 +, SOH, SO-, vari ous species, and SA or SO-Na+ respectively.

PAGE 40

26 the following factors are considered: 1) Characteristics of the adsorbent 2) Characteristics of the solution 3) Characteristics of the adsorbate 4) The interactions of the adsorbent with the solution and the adsorbate The two most frequently used isotherm models for heavy metal adsorption are Langmuir and Freundlich models. 3.2.1 Langmuir M odel The use of the Langmuir isotherm model is based on the following assumptions (Benjamin, 2002): 1) All sites have equal bindi ng energy. 2) The binding sites are uniformly distributed on the adsorbent surface. 3) The affinity of the sites for the adsorbate is independent of the solution condition. 4) There is no effect of the adsorbed species on the adjacent sites. 5) A single value is used to represent the reaction between a given adsorbate and all the surface sites. In a binary system, Benjamin (2002) used a simple equilibrium surface complexation reaction model and a corresponding constant to illustrate the occupation of the sites by an ad sorbate in a system. The equilibrium reaction equation was written as: (3.1) where = aqueous adsorbate is the unoccupied site by the adsorbate is the occupied site by the adsorbate. From equation (3.1), adsorption coefficient,

PAGE 41

27 is given by : (3.2) where the total number of site is given by : (3.3) Manipulation and substitution of equation (3.3) into equation (3.2) leads to (3.4) Equation (3.4) can also be written as (3.5) Equation (3.5) is known as the Langmuir model where = the mass loading of the contaminant per mass of the adsorbent; K L = Kads is the coefficient of the Langmuir model that measures the affinity of the adsorbent for the adsorbate; and = the maximum loading capacity of a given mass of the adsorbent. If 1, then the ads orbent has a very low affinity for the adsorbate and equation (3.5) becomes equal to KL{A(aq)}{qmax (3.6) }. Testing of experimental data fits to the Langmuir Isotherm is usually done by transforming data to fit Equation 3.6. In this work data fitting was actually done using the Gauss Newton to fit Equation 3.5. This was done in MATLAB and Appendix D provides the script for running this analysis. In a ternary system involving two adsorbates A and B along with the adsorbent, Benjamin (2002) derived the competitive Langmuir isotherm model from two simple complexation reactions:

PAGE 42

28 (3.7) (3.8) The ratio of adsorbate B sorbed to adsorbate A sorbed is written as : (3.9) The site balance for this case is: (3.10) By substitution of equations (9) and (10) into equation (7), Langmuir isotherm derived is given by : (3.11) Langmuir sorption model of an adsorption indicates limited adsorption capacity of an adsorbent as per an adsorbate. The limited adsorption capacity occurs as a result of limited number of sorptio n sites the adsorbent possesses as per the adsorbate. Linear and Freundlich models on the other hand indicate unlimited sorption capacity. This may be due to a presence of infinite number of sites on an adsorbent per an adsorbate. Linear and Freundlich models also indicate lower affinity of the adsorbent for the adsorbate. Although there is infinite numbers of sorption site types present in many adsorbent solid surfaces, only fewer types of the sites are usually dominant while most of the sites types ar e insignificant in the amount of adsorbate adsorbed in the removal processes. For instance, Dzombak and Morel ( 1990) ; Papini et al. ( 1999) identified more than one sorption site type in some adsorbents with respect to some adsorbates. Dzombak and Morel (1990) reported of two types of sites on iron oxide adsorbent in arsenic adsorption tests while Papini et al (1999) reported of three site types on a heterogeneous natural

PAGE 43

29 medium (red pozzolan) at a constant ionic strength of 0.1 mol/L for lead (Pb) adsorption. For multisite Langmuir isotherm, (ex. Two sites), the total adsorbate loading qA,tot (3.12) is given by : where 1 Aq = total loadings onto site type 1 and 2 Aq = total loadings onto site type 2. 3.2.2 Freundlich M odel There are other instances when the binding sites cannot be represented as discrete or by some few dominating sites. The surfaces behave as if the sites present are associated with continuous distribution of binding energies. Here Langmuir isotherm f ails but one of the isotherm functions that fit such scenario is Freundlich. Freundlich isotherm model is given by : = (3.13) w here (3.14) and R and T are universal gas constant and absolute temperature respectively. The failure of t he Langmuir isotherm equation is attributable to the various assumptions upon which the model was derived. For instance, Langmuir model assumes that the adsorbent binding surface sites are uniformly distributed, identical and all can be represented with a single value under all conditions. The Langmuir equation again assumes that binding of an adsorbate to any site has no effect on the equilibrium constants for binding of other molecules to the surface. However, these assumptions are not always true.

PAGE 44

30 3.3 Hydroxide S urfaces and I onic A dsorbate In solution of hydroxyl surfaces are us ually represented as 2 +, for one type sorption site species. Depending on pH there is always a relative dominance of one or two site species over the other. Typically surface site species of iron hydroxide groups are written as 2 +, -. During adsorption ionic adsorbates are presumed to bind directly to oxide surface or the Fe ion, displacing the x group. Surface complexes formed by such reactions are relatively strong and are reffered to as innersphere complexes. H+ and OHions are always presumed t o bind to oxide surfaces by innersphere complexes. Some other ions on the other hand are presumed to bind to water molecule which in turn binds to the Fe surfaces. The adsorbates here are not directly linked to the Fe oxide surface but by connection through water molecules Such binding forms weaker complexes with the surface and is known as outer sphere complexes. 3.4 ORP and Eh ORP or redox potential (E M easurements h) studies are of importance in environmental chemistry. This importance stems from the changes in the characteristic properties of the elements that are involved in this adsorption process. ORP changes have effect on the original properties of both the a dsorbent and the adsorbate of a system. For instance, low ORP values change S(VI) as in SO4 2 to is S( II) in H2S. S(VI) is highly soluble, non volatile and relative ly innocuous. However, S( II) in H2S forms insoluble metal precipitation and is also qui te toxic (Benjamin, 2002). ORP is measured in millivolts (mV) or in volts (V) with ORP electrodes containing a silver/silver chloride (Ag/AgCl)

PAGE 45

31 reference The values obtained are then converted into oxidation potential or redox potential Eh voltage values. Eh voltage values are the values that would be obtained if Standard Hydrogen Electrode was used. For the Orion ORP electrode that was used in this work, the ORP (mV) values obtained were converted to Eh (mV) by adding absolute value of 219 mV to the ORP readings values at a temperature of 20C or 220 mV at 25C. Like pH, pe is defined in relation to activity. The relationship between Eh and pe is given by where F is Faradays consta nt given by 96485.309, T is temperature in Kelvin, and R is molar gas constant given by 8.314 J/mol K. 3.5 Batch K inetics S tudies For nonequilibrium adsorption processes in porous solid media the migration of the contaminant species from bulk solution into the solute particle encounters two resistances in series: a resistance due to the external film, and intraparticle resistance. In a fast stirred batch system the thickness of the film surrounding the adsorbent particles is assumed to be thinned out. Consequently, resistance to adsorbate migration across an external boundary layer is considered negligible. In the pores of an ads orbent, the diss olved adsorbate migrates towards the center of adsorbent due to either concentration gradient in the pore water (pore diffusion) on the pore wal ls (surface diffusion). The rate of diffusion in the pores is usually described by Ficks law The intra particle solute concentration under unsteady state with a fixed diffusivity in a spherical adsorbent is given by : r q r rq D t qp p app p22 2 (3.15)

PAGE 46

32 w here pq is solute concentration in the pore water [M/M] appD is apparent diffusion coefficient [L2 r /T] is the rad ial position of the adosrbate in the spherical solid particle [L] and t is time [T] W ith initial condition: 0 ) 0 ( t r qp (3.16) and with boundary conditions: b pC a r t q ) ( ( 3.17) w here = As concentration in the bulk solution [M/L] and M = mass of the adsorbent [M] 0 ) 0 ( r t r qp (3.18) Crank (1975) established t he analytical fractional uptake solution to linear diffusion equation in a limited volume with the same initial and boundary conditions as written above for the equation (3.15). The linear diffusion equation, also known as the Ficks first law of diffusio n is given by : r q D Fluxp app ( 3.19) and Cranks (1975) solution is also given by : 2 2 1 2 2exp 9 9 ) 1 ( 6 1 at q D q M Mn app n n t ( 3.20) where tM and M are the masses of the solute in the adsorbent at time t and at equilibrium time (infinite time) respectively. The parameter nq are non zero roots [ ] of 23 3tann n nq q q and dK a V34 3 [ ] is also expressed in terms of final fractional

PAGE 47

33 uptake of solute by the spherical adsorbent as: 1 10VC M This analytical solution to adsorption model assumed a linear isotherm. The most frequently used mathematical algorithm by which the no nzero roots of 23 3tann n nq q q is estimated is Newton Raphson method. Equation 3.16 states that initially there was no contaminant in the pores of the adsorbent Equation 3.17, also states that at time infinity the solute concentration on the periphery of the spherical adsorbent is in equilibrium with the bulk solution. Finally, Equation 3.18 indicates that the solute concentration gradient in the core of the spherical sorbent is zero Ball and Roberts (1991) experimentally determined fractional uptake from: be bi b bi dC C C C f (3.21) where biC bC and beC are the solute concentrations [M/L] observed in the bulk solution initially, at time t, and at equilibrium respectively. For no instantaneous adsorption and no partioning into to the headspace or otherwise lost from solution, equation ( 3.20) becomes M M C C C C ft be T b T d (3.22) where TC is the total initial concentration of the solute in the bulk solution [M/L] tM and M are the masses of solute loading onto the adsorbent at any given time and at equilibrium respectively Solute diffusion in porous media is hindered by such factors as tortuous pathways, dead end pores, and variable pore diameters.

PAGE 48

34 According to Ball and Roberts (1991), apparent diffusion coefficient relates to effective pore diffusion coefficient by: int* R D De app (3.23) where eD is effective diffusion coefficient [L2 intR /T] is the internal retardation factor [ ] within the internal pores of the adsorbent and is also given by : ] ) ( 1 [int di i pK R (3.24) where i is internal porosity [ ] and diK is linear equilibrium adsorption coefficient [ L/M] Equation 3.12 was derived on the assumption that adsorption is linear in the pores of the adsorbents. Diffusion in the bulk is however assumed to be faster than pore diffusion for two reasons. Bulk diff usion involves simple geometries and a straight path Pore coefficient thus relates to bulk diffusion coefficient as: r b eK D D ( 3.25) where rK is constrictivity factor [ ] ( is tortuosity factor [ ] ( 1). In this research eq uilibrium tests were performed with both 500 600 m and 38 m grain sizes. The purpose for using fine grain size in the equilibrium experimentation was to attain a shorter equilibration time. There have been reported situations when the adsorption capacity of an adsorbent had dropped with the increase in the adsorbent grain size (Giammar et al 2007). One possible way this can occur is when the fine grain sizes of the adsorbents used are far smaller the smallest pore sizes of the bulk such that the fine grain particles become non porous. For instance, nano sized fine grain adsorbent derived from microporous bulk adsorbent as was the case with Giammar et al (2007). Other

PAGE 49

35 possible causes that have been reported include c onstrictivity and dead ends. Such case could impl y that equilibrium experimentation with the fine grain size is not represent ive of that of the coarser grains. Internal constrictivity and dead ends (steric effect) are difficul t to determine separately and so are usually determined as a lump sum (Ball and Roberts, 1991). 3.6 Experimental D ata F itting There are two types of applications adopted in experimental data fitting. These are trend analysis and hypothesis testing. In this work trend analysis was used and this presents a process of using a pattern of data to make a prediction. Least squares regressions are used in a trend analysis to predict imprecise data while interpolations are used to de termine data with high precision (Chapra and Canale 2002). Some of the tools of least squares regression for a best fit are minimization of sum of squares of residual errors and coefficient of determination (r2 (3.26) ) for all the available data. The coefficien t of determination is given by: where ) 2 and ) 2 the total sum of the squares of the residuals betwe en the data points and the mean, the sum of squares of the residual errors, = arithmetic mean of a sample, predicted values. For a perfect fit 0 and r 2Gauss Newton algorithm is also another method that has been employed by many in fitting non linear data. This method which uses minimization of the sum of squares of residual errors was adopted in this work. = 1.

PAGE 50

36 Chapter 4 Material s and M ethods 4.1 Introduction This chapter describes the materials and the methods adopted for this research. The approach to the experiments conducted was grouped into two main categories: Kemiron particle characterization, and batch sorption tests. The batch adsorpt ion tests were subcategorized into rate of sorption and equilibration tests. Both the rate of sorption and the equilibration tests were initially done in binary systems (only arsenic present) and then further tested in more complex systems containing more than one contaminant and in the presence of synthetic landfill leachate. 4.2 Materials 4.2.1 Adsorbent Kemiron is an adsorbent that is manufactured by Kemiron Company, with a local distributor in Florida. The adsorbent particles were ground in a ceramic mortar and sieved through American Society for Testing and Materials (ASTM) stainless steel sieves to obtain two particle sizes referred to as fines ( 38 m) and coarse (500 600 m) fractions. The Kemiron adsorbent was chosen for t his research because it was a newly

PAGE 51

37 developed product with little known about its adsorption performance with arsenic remediation and its availability in Florida. 4.2.2 Reagents and S tock S olutions All the reagents used were of analytical grade and purchased from Fisher scientific. The stock solutions were made by dissolving the given solid reagent in ultrapure water (Barnstead ) with resistance of 18.2 M ohm. The reagents also included sodium hypochlorite, NaOCl, sodium sulfide, Na2S, sodium arsenate heptahydrate, Na2HAsO4.7H2O, sodium nitrate, NaNO3, nitric acid, HNO3, sodium hydroxide, NaOH, s odium acetate, CH3COONa, sodium propionate,C2H5COONa, sodium carbonate, Na2CO3, magnesium chloride, MgCl2, sodium sulfate, Na2SO4, ammonium nitrate, NH4NO3, sodium chloride, NaCl, calcium carbonate, CaCO3, and sodium selenite, Na2SeO3. A stock solution of 150 mg/L Ni (Ni(NO3)2.6H2O) was used as modifier solution in the determination of arseni c using GFAAS analysis. Prior to use, NaNO3 was dried at 80C for 4 hours and stored in a de ssi cator. pH adjustments were done with the nitric acid, HNO3 and NaOH and ORP adjustments were done with NaOCl and Na2S. CO2The s ynthetic landfill leachate was made by combination of various salts partly based on papers and reports from Kjeldsen et al. (2002) Kjeldsen and Christophersen (2001) Table 4.1 shows the resulting species composition and their concentrations. free milliQ water was prepared by boiling ultrapure (Barnstead) water and spargi ng with ultra high pure argon g as (Airgas Incorp.) until cool and maintained under an Argon atmosphere All slurries were also purged with Argon gas for 24 hours prior to spiking with various stock solutions which were freshly prepared for each experiment.

PAGE 52

38 Table 4.1: Synthetic landfill leachate constituents Parameter Concentrations (mg/L) Classification Young phase (acidogenic) Old phase (methanogenic) CH 3 COO 11000 COD 1500 COD Organic species C 2 H 5 COO 11000 COD 1500 COD Na 3270.5 + 4971.9 Mg 470 2+ 180 Ca 1200 2+ 60 NH 4 + 740 N 740 Inorganic species CO 3 2115.5 2 4190 SO 4 500 2 80 Cl 2120 2645.1 NO 3 2544.3 2544.3 Ni 0.17 2+ 0.17 Co contaminant SeO 3 5 2 5 4.2.3 Instrumentation A Hitachi H 7010 coupled to an electron counting sensor (Joel JSM 840) was used for Scanning Electron Microscopy (SEM) and Energy Dispersive Spectroscopy (EDS). SEM provides information on Kemiron particle morphology and EDS provides information on the perce ntage by mass of the elemental constituents of the Kemiron. An X ray diffractometer (XRD) (Philips) was used to identify mineralolgy of Kemiron. A copper target was used for the x ray source with a strongest characteristic radiation (K1 ) at a wavelength o f about 1.54 angstroms. The effective pore size, the total pore volume as

PAGE 53

39 well as the total pore surface area of Kemiron were determined with mercury intrusion porosimetry by Micromeritics in Atlanta, Georgia. BET multipoint surface are a and pore analyses with nitrogen intrusion were also measured. Kemiron was dried at 80C for 18 h and degassed at 80C for 3 h prior to these characterizations A Ro s s semi micro gel filled electrode coupled to an Orion 940 pH meter was used for pH measurements after being calibrated with Fisher Scientific pH buffers, 4.0, 7.0, and 10.0. Oxidation reduction potential of the system was measured with an ORP probe (ORION 9678BNWP) connected to the ORION 940 meter in the relative millivolt mode. Potassium iodide solution (ORI ON) was used as the standard solution. A G raphite F urnace A tomic A bsorbance spectrometer (GFAA), Varian Spectra AA 640 DUO model which was equipped with automated sample injection (GTA 100) was used to measure total arsenic concentrations 4.3 Methods Figure 4.1 illustrates the experimental setup used for batch experiments in this study. Ultra high purity (UHP) nitrogen gas was scrubbed to remove CO2 prior to bubbling into the reactor vessel to remove CO2 and maintain a CO2Prior to the start of the experiments all glassware was washed with Liquinox detergent and then soaked in 1 N sodium hydroxide for more than 1 hour, rinsed with milliQ water and soaked again in 10% nitric acid overnight before finally being rinsed with and left soaking in milliQ water overnight. Cleaning of polycarbonate (PC) free system. An overhead s tirrer was used for experiments with the 500 600 m particle sizes to reduce mechanical alterations on the particle size.

PAGE 54

40 Fig ure 4. 1: Ultra high pure nitrogen gas sparging setup for a batch system containers w as similar to that of glassware except the concentrations of acid and base were both 0.1 N. 4.3.1 Batch A dsorption C haracterization Batch kinetic studies in binary systems we re undertaken on two grains sizes of Kemiron ( 600 The objectives for these were to establish equilibration times and also to estimate diffusion rate constants for As removal in both the binary and in more complex systems involving these two grain sizes. All experiments involving the 500 600 grain size were done with 1000 mL solutions and an overhead stirrer and all of the binary experiments involving the were done in 200 mL solutions. The experiments involving the syntheti c landfill leachate were done in 1000 mL solutions irrespective of the adsorbent grain size. In the binary systems the pH of the solution was first lowered to ~ 5.5 with 0.1 N HNO3 before it was sparged with th e

PAGE 55

41 ultra high pure nitrogen gas overnight The pH was then raised to 7 with 0.1 N NaOH and then 5 mg/L As was spiked into it. The pH was maintained at 7 using both 0.001 N NaOH and 0.1 N HNO3Equilibration tests were conducted on As in both the binary and in the complex systems invol ving only the and the 500 600 grain sizes of Kemiron. In the binary system and with only the fine slurries, the tests were done at ionic strength s of 0.001 N and 0.1 N of NaNO Samples of 2 ml were taken as a function of time for every experiment and also for each of the grain sizes The results from the analyses were modeled with mass transfer equations. The kinetic studies of the landfill leachate were done at pH 9.4 and at ORP of 240 mV. 3. The objectives were to evaluate the impacts of: 1) ionic strengt h on As removal in the binary system, 2) initial As concentration on amount of As adsorbed, and 3) the effect of particle size ( and the 500 600 grain sizes) on the As removed. In each case t he system was made of 0.1 g/L Kemiron in a polycarbonate (PC) batch reactor with CO2 free ultrapu re water. The gas sparging with the pH before and after were the same as described in the kinetic studies. 15 Samples of 8 mL were removed into 10 mL PC tubes at various pH levels. The hea d spaces in the 10 mL PC tubes we re filled with ultra pure nitrogen gas. After 72 hours the sample pH was recorded and t he samples were The filtrate was acidified with concentration HNO3With the synthetic landfill leachate the equilibration tests were done for two different categories on the bas ed on landfill age. Young (acidogenic) landfill leachate to 0.7% and analyzed for total As The equilibra t ion binary batch experiments wer e carried out on initial As concentrations of 5 10, 15, 20, 30, and 40 mg/L.

PAGE 56

42 contained higher concentrations of the constituent ions while old (methanogenic) landfill leachate contained less concentrations of constituent ions. Batch equilibration tests on the synthetic leachate were similar to those of the binary systems in ultrapure water. The ultrapure water used for the leachate preparation was simultaneously boiled and sparged with UHP nitrogen gas to remove dissolved CO2 and O2. pHs in the synthetic leachate were changed with concentrated HNO3. In the competing batch sorption experiments, concentrations of various ions from their respective salt solutions were spiked into the batch slurries along with arsenic The competing ions used were SeO3 3 -, Ni2 +, NH4 +, CO3 2 -, SO4 2 and the objective w as to evaluate the impact of the presence of the ions on As removal. Samples of 8 mL were then taken into the 10 mL PC tubes at various pH levels. The head spaces in the 10 mL PC tubes are filled with ultra pure nitrogen gas. After 72 hrs of equilibration on an end over end rotator, the pH was measured and t he samples filtered through was acidified with concentrat ed HNO3Batch isotherm experiments were also done at room temperature on As with initial concentrations of 5 mg/L, 10 mg/L, 20 mg/L, 30 mg/L, 40 mg/L and 50 mg/L at various pH values (6,7, and 9) for 72 h and with the ionic strength of 0.001 N NaNO to 0.7% and analyzed for As. 3. The objective for this experiment was to ev aluate the impact of initial concentrations of As on the mass density of As adsorbed onto Kemiron. Multiple samples each having a volume of 8 mL we re taken from the slurry into 10 mL PC tubes. The preparations and the filtrations of the samples for analysi s for the isotherm were similar to those for the batch equilibration experiments. Batch equilibration tests in the synthetic landfill leachate were done with two objectives. The first objective was to verify the increased As

PAGE 57

43 adsorption impact as exhibited by the various conditions in binary systems in the landfill leachate system. The second objective was to select the two parameters that had the most impact on As removal in the the binary systems. By applying a 22A set of preliminary experiments were also done using real landfill leachate collected from the North Central Landfill in Polk county, Florida. The leachate was collected in 1L HDPE containers and stored on ice during transport. Once in the laboratory it was filtered through a 0.45 m filter and used for experiments. The leachate was also analyzed for total arsenic concentratio ns. Leachate was digested on an Environmental Express Hotblock set at 105 factorial experimentation method, optimum conditions for maximum As removal in the synthetic landfill leachate were determined. oC (sample temperatures were 95oC). 100 ml of leachate sample was placed in a 250 ml beaker to which 3 ml concentrated HNO3 was added and the mixture boiled down to ~ 5 ml. 3 ml more of concentrated nitric acid were added and the mixture boiled until constant color. 10 ml of concentrated HCl was then added along with ultrapure water and the mixture boiled for 15 minutes after which the cooled mixture was made to mark in a 100 ml volumetric flask. This procedure was repeated using a 100 ml Environmental Express polyprophene container, and then 1 mL of 30% H2O2 plus 2.5 mL of concentrated HNO3 was added and the samples allowed to heat for 2 hours after which they were cooled, diluted and analyzed on the GFAA.

PAGE 58

44 Chapter 5 Results and Discussion 5.1 Introduction This chapter describes and discusses the characterization studies of Kemiron and the results from the adsorption experiments performed. Kemiron characterization w as done with X Ray Diffraction (XRD), Scanning Electron Microscopy (SEM), Electron Dispersion Spectroscopy (EDS), Mercury porisimetry and nitrogen adsorption Brunauer Emmett Teller (BET) surface area determinations. The impacts of various experimental var iables on batch equilibrium, and batch kinetics of the As adsorption onto Kemiron were evaluated. Variables such as As concentration, pH, ionic strength, ORP, and solution composition were tested for their impact on both clean systems and on synthetic land fill leachate at room temperature. The suitable parameters for maximum adsorption were adopted from clean systems and tested in the systems using synthetic landfill leachate. 5.2 Kemiron S urface C haracterization The surface characterization of Kemiron was done on grain sizes. Three commonly used pore size classifications are micropores (pore diameter smaller than 2 nm), mesopores (pore diameter 2 50 nm) and macropores (pore diameter larger than 50 nm) (Sing et al., 1985). Gases like nitrogen (0.15 nm diameter) can access

PAGE 59

45 pore sizes ranging fro m 0.3 to 300 nm whereas molecules of mercury (0.314 nm diameter) access pores ranging from 3 nm to 200 m. Table 5.1 summarizes the Kemiron surface characteristics based on mercury porosimetry and BET nitrogen intrusion performed by Micromeritics, Georgi a. Table 5.1: Properties of Kemiron particles. For t he 500600 m diameter particles and mercury porisimetry analysis. Property Quantitative value Total Pore Volume (ml/g) Bulk Density @ 55 psia (g/ml) Porosity (%) Max Pore Diameter ( ) Min Pore Diameter ( nm ) Median Pore Diameter (n m ) Mean Pore Diameter ( nm ) Total Surface Area (m2BET Surface Area (m /g) 2/g) Skeletal Density (g/ml) 0.42 1.32 55 327 3 7 76 22.1 39.8 2.94 BET nitrogen gas intrusion method. The surface area obtained from mercury porisimetry was 22.1 m2/g for the 500 0.2 m2/g and 39.8 0.2 m2/g for sizes respectively. There was a 44% difference between the surface area obtained for the 500 -

PAGE 60

46 600 m particle size fraction using mercury porisimetry and nitrogen intrusion with the latter method giving the higher of the two surface areas. This can be exp lained in terms of the pore size distribution, where mercury porisimetry is unable to access pores less than 3 nm (minimum pore diameter observed in Table 5.1 was 3 nm using mercury porisimetry and shown in Figure 5.1). N2 adsorption BET surface area analy sis accesses pore sizes down to 0.3 nm and the discrepancy between the two surface areas suggests that roughly 44% of the pores lie in the 0.3 to 3 nm range. It is possible that some of these micropores are inaccessible to ions like H2AsO4 Research on the nature of material pores and sorption isotherm characteristics has been done by Rigby (2005); Gregg and Singh (1982) and the interpretations here are based on these sources. The hysteresis loops presented in both the mercury and the nitrog en gas sorption (lower curve) and desorption (upper curve) are shown in Figures 5.2 and 5.3. The hysteresis characteristic features are associated with capillary condensation occurring in pores. According to the sorption classification of Gregg and Singh (1982), Figures 5.2 and 5.3 conform to Type IV of the adsorption isotherm. However; the linear part of the graph below the hysteresis loop which indicates a stage of monolayer coverage is missing in Figure 5.2 (mercury adsorption graph) but is present in Figure 5.3 (nitrogen adsorption graph). The reason for the missing stage of the graph is unclear but might be suggestive of mercurys inability to access the micropores. Another which has an a verage diameter of around 0.8 nm, however, their contribution to total surface area is significant (Bodek et al., 1988). There was only a 5% difference between the surface areas obtained for the two size fractions studied and this is expected since crushing of particles should not change surface area when the majority of that surface is within the pore structure.

PAGE 61

47 thing that the hyteresis loop indicates is pore structure type present in t he adsorbent particles. In comparison with the classification of Gregg and Singh (1982), Figure 5.2 Figure 5.1: Cumulative area mercury porosimetry of 500 Figure 5.2: Mercury adsorption isotherm onto 500 10-4 10-2 100 102 104 0 5 10 15 20 25 Pore Size Diameter (micron)Cumulative Area Intrusion m 2/g 0 0.1 0.2 0.3 0.4 0.5 0 0.2 0.4 0.6 0.8 1 Relative Pressure P/Po Cummulative vol adsorbed (mL/g

PAGE 62

48 indicates H1 Type pores which are often associated with porous materials consisting of agglomerates of approximately uniform spheres in a fairly regular array, with a nar row Figure 5.3: Nitrogen adsorption isotherm onto 500 Figure 5.4: Nitrogen adsorption isotherm onto 20 40 60 80 100 120 140 160 180 0 0.2 0.4 0.6 0.8 1 Relative Pressure P/Po Cummulative volume adsorbe (mL/g) 0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35Volume adsorbed (mL/g)Relative press (P/Po)

PAGE 63

49 distribution of pore sizes. Figure 5.3 on the other hand indicates a Type H3 loop which does not exhibit any limiting adsorption at high relative pressure (P/Po ). This is associated with aggregates of plate like particles with slitshaped pores. Figure 5.4 shows no hysteresis associated with the e pressure up to 0.33. This suggests the sorption test ended prematurely when compared with Figure 5.3. 5.2.1 Scanning E lectron M icroscopy (SEM) Particle morphology for Kemiron was determined using a Hitachi H 7010 Scanning Electron Microscope (SEM) w ith Joel JSM 840 attached. A representative micrograph of Kemiron is shown in Figure 5.5. The surface of the 500 600 particle appear to consist of aggregates of smaller particles which look rounded and fluffy. This observation is in agreement with the Type H1 pore structure interpretation of Figure 5.2. Figure 5.5: Scanning Electron Microscope (SEM) M icrogram of a 500 600 Kemiron particle.

PAGE 64

50 5.2.2 X R ay D iffractometry (XRD) The X Ray Diffractogram of Kemiron (Figure 5.6) lacked very well defined peaks and had a noisy baseline. This suggested that Kemiron could be amorphous. The closest iron based compound in the database of International Committee Coal Organic Petrology (ICCP) on which high peaks coincided was goethite. The noisy peak characteristics may also depict agglomeration of very fine particles (Alcantar and Pichler, 2007). This interpretation of particle aggregates is supported by the H1 type pore structure classification of Figure 5.2 and the SEM as well. Figure 5.6: X Ray diffractogram (XRD) of Kemiron powder ( goethite for comparison.

PAGE 65

51 Table 5.2 contains some surface area data of goethite and granular ferric hydroxide (GFH) reported by some researchers. The BET N2 adsorption surface area of goe thite presented ranges from 27.5 m2/g to 94 m2 /g and depends on factors like the rate of base addition during the precipitation process (Villalobos et al., 2003). The surface area of the Kemiron also falls within the range seen in the literature for goeth ite and is much lower than what has been reported for the amorphous iron oxides (ferrihydrite in Table 5.2) Table 5.2: BET surface areas reported on some iron based adsorbents. Adsorbent Surface area (m 2 Reference /g) Goethite Goethite micro rod Goethite 25 Goethite 140 GFH Ferrihydrite 27.5 40 3 40.2 47.05 235 8 280 30 Campo et al. (2008) Cwiertny et al. (2009) Kosmulski et al. (2003) Kosmulski et al. (2003) Badruzzaman et al. (2004) Hiemstra and van Remsdjk (2009) 5.2.3 Electron D ispersion S pectroscopy (EDS) Energy Dispersive Spectroscopy (EDS) revealed the major elemental weight percentages in Kemiron as 40.37 % iron, 42.25 % oxygen, 7.92 % carbon and 5.90 % sulfur. Figure 5.7 shows the number of emitted electrons per second of K emiron at a given amount of electron volts generated. The percentages by weight analysis were done by quantitative methods using Atomic number, Absorption and Fluorescence (ZAF) correction.

PAGE 66

52 Figure 5.7: Energy Dispersive Spectroscopy (EDS) scan of < 38 m Kemiron. 5.3 Kinetics of A rsenic A dsorption in B inary S ystem The kinetics of adsorption has received several interpretations in terms of the processes responsible for observed phenomenon. LaBolle and Fogg (2001) explained the process of contaminant transport in a quiescent system by means of molecular diffusion whi ch is enhanced by mechanical dispersion (e.g. stirring ). During contaminant transport into a porous medium from a bulk solution the contaminant migration proceeds through an assumed external boundary between the bulk solution and the solid surface and the n into internal pores toward the center of a porous solid. Mass transfer resistance to the migration of the solute into the pores of the adsorbent result from the external boundary layer surrounding the solid particle and also from diffusional processes i n the internal pores of the adsorbents. 0 100 200 300 400 500 600 700 800 900 0 100 200 300 400 500 600 700 800 900 1000 1100 Voltage (eV)Counts per secondsFeKb FeKa SKa FeLa FeLl O Ka O Ka = Oxygen from K shell (alpha radiation) Fe Ka = Iron from K shell (alpha radiation) Fe Kb = Iron from K shell (beta radiation) Fe La = Iron from L shell (alpha radiation) Fe Ll = Iron from L shell (lambda radiation) S Ka = Sulphur from K shell (alpha radiation)

PAGE 67

53 In the batch experiments involving continuous stirring, external film resistance is minimized and the rate of removal depends on the resistance in the internal pores. Thus in this experiment, the rapid rate step w hich completed within a few minutes might be due to resistance in the macro and meso pore structures (see Figure 5.1). The slower and mass transfer rate limiting step took hours to days to reach equilibrium and might be due to resistances in micropores of the adsorbents (Cunningham et al. 1997). Bulk diffusion is assumed to be faster than pore diffusion and is also assumed to involve simple geometries as well as straight paths while pore diffusion involves complex, tortuous pathways, deadend pores and va riable pore diameters (Ball and Roberts 1991). This means the longer the microscale length, the longer the time for equilibrium. In this research, continuous stirred batch kinetic experiments were done on As with 0.1 g/L Kemiron, at pH 7, and with ionic strength of 0.001 N NaNO3. The particle sizes of Kemiron used for these experiments were 600 m in diameter The objectives for the kinetic studies were to: 1) determine the As adsorption equilibration time, 2) evaluate the impact of th e grain size on the adsorption capacities, and 3) estimate diffusion coefficients. The kinetics of adsorption of As onto Kemiron was expected to be intraparticle diffusion controlled. Also the equilibration time was expected to depend on the diffusional length. The dependence of the equilibration time on diffusional length was based on an assumption that the diffusion coefficient of As migration into the pores of Kemiron is unaffected by the size of the adsorbent. Again, the adsorption capacity of Kemir on was expected to remain unchanged for all grain sizes under the same physico chemical conditions.

PAGE 68

54 Figure 5.8: Rate of uptake of As onto 0.1 g/L Kemiron. Conditions: pH = 7, AsT = 5 mg/L, I = 0.001 N NaNO3, no CO2 and at room temperature. Figure 5. 8 plots the rate of uptake results which clearly indicate that the equilibration time of sorption of both As(V) and As(III) depends on the diffusional length of the Kemiron particles. The experimental time scale of the sorption of both As(V) and As(III) o nto the 500 600 m particle size of the adsorbent was not long enough for equilibrium to be achieved. The assumptions made during the equilibration times were that the diffusion coefficient was constant, the equilibration time was directly proportional to the square of the diameter of the adsorbent, and the sorption capacity remains the same for all grain sizes where that capacity was determined from sorption data of the both As (V) and As(III) sorption onto the sorption onto the 500 600 m particle size, the estimated time for the equilibration was determined mathematically to be 8975 h (374 days). 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 0 50 100 150 200 250 300 350 400 450 time (h) Aqueous As (ug/L) 5 mg/L As(V), 38 um 5 mg/L As(V), 500-600 um 5 mg/L As(III), 500-600 5 mg/L As(III), 38 um

PAGE 69

55 In addition to showing that the re was a higher rate of As removal involving the 38 m particle size than of the 500 600 m grain size, Figure 5.7 also shows that As(V) and As(III) adsorption onto the 500 600 m had two gradients, with the initial one being more steep. This sugges ts an initial fast rate step followed by a slower rate step. The slower rate step as explained by La B olle and Fogg (2001) and Cunningham et al. (1997) could be due to internal pore diffusion. Figure 5.7 shows that Kemiron adsorption capacity appears equa l for the two As species onto adsorption onto the 500 600 m size in the timeframe of the experiment. While there is ~ 80% sorption of 5 mg/L initial concentrations of both As(V) and As(III) onto there was abou t 70% sorption of As(III) and 50% sorption of As(V) onto the 500 600 m particle size by the end of the experiments. The rate of removal of As(V) was slower than As(III) in the 500 600 m grain size. As(III) is known to exist as an undissociated, uncharged molecule at pH 7 while As(V) exists as speciated ions with net negative charge. Wet chemistry tests to determine binding strength usually looks at the effect of ionic strength which was done for this research and the r esults are shown in Figures 5.18 and 5.19. There appears to be no significant effect of ionic strength on the binding strength of either As(V) or As(III) to the Kemiron surface. The strength of the bond between Kemiron and either As(V) or As(III) can also be assessed through the inf luence of a competing ion. Section 5.8 presents results on arsenic sorption in the presence of competing ions which show As(III) to be more sensitive to the presence of such ions (e.g. selenite). This can be interpreted to mean that As(III) forms weaker complexes with the surface than As(V). The faster rate of removal of As(III) compared to As(V) from the 500 600 m grain size could be due to

PAGE 70

56 its weaker surface complexes which are more mobile if a surface diffusion type mechanism (i.e. arsenic diffusio n results from concentration gradients along the particle surface) explains movement into the particle micropores. On the other hand, a pore diffusion model (i.e. arsenic diffusion results from concentration gradients in the pore water) could also explain why the uncharged As(III) species sorbs faster than the charged species whose movement would depend on counterions diffusing out of the micropores. Sections 5.45.8 delve into modeling details of the rate of uptake data. For the subsequent experiments pr esented after section 5.8, an equilibration time of 72 hours was used given that all experiments were conducted on the 38 m grain size. This time was longer than that observed in the previous discussion, but was used to account for any effects that might occur due to the presence of competing ions in the more complex systems. 5.4 Modeling R ate of A rsenic A dsorption The rate of As loading onto the Kemiron grain particles was modeled with Cranks analytical solution to Ficks law of diffusion in a limited volume. The objective for the modeling was to estimate a diffusion coefficient for As removal onto Kemiron in both DI water systems and in synthetic landfill leachate systems. According to Ball (1990) organic solute transport depends among other factors, on the rate of sorption of solute onto an adsorbent and the capacity of the adsorbent for the adsorbate. It was assumed that this relationship also applies to an inorganic solute like As and in an adsorbent like Kemiron.

PAGE 71

57 5.4.1 Diffusion C oeff icient E stimation The diffusion coefficient estimation using Cranks (1975) model required two dimensionless parameters: fractional uptake and dimensionless time, defined by Ball and Roberts (1991) used on the ordinate instead of f d and the for the abscissa instead of time, t The reason was that the ultimate fractional uptake (F) affects fd and consequently affects the diffusion coefficient. equals where C be is the aqueous concentration at equilibrium and Cb is the aqueous concentration at time t of the sorbate ion of interest. The dimensionless was thus used in order to normalize the effect of F. The diffusion rate constants of As transport onto t he various grain sizes of Kemiron were estimated with the assumption that the transfer mechanism was intraparticle diffusion controlled and that the external film resistance was negligible. Cranks (Crank 1975) model adopted for the coefficient estimation was based on a linear isotherm adsorption model and t he results of Cranks model are shown in Figures 5.9 5.15 and in Table 5.3. Cranks (1975) points, a least squares procedure was adopted u sing Gauss Newton method. F or Cranks model one fitting parameter , was used. The fitting computation steps can be obtained upon request. However, the computations of the Gauss Newton method can be found in Appendix D.

PAGE 72

58 Figure 5. 9: Fractional mass of As(III) removal onto grain size in a batch system. Conditions: 5 mg/L As(III)T, 0.1 g/L Kemiron, I = 0.001 N NaNO3 at fixed pH of 7, and Kemiron adsorbent. Figure 5. 10: Fractional mass of As(V) removal onto grai n size in a batch system. Conditions: 5 mg/L As(V)T, 0.1 g/L Kemiron, I = 0.001 N NaNO3 pH = 7, and Kemiron adsorbent. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1E 05 0.0001 0.001 0.01 0.1Kd app/Kd ult Dapp.t/a*a Crank's model data point 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1E 05 0.0001 0.001 0.01 0.1Kd app/Kd ult D app .t/a*a Crank's model data point

PAGE 73

59 Figure 5.11: Fractional mass of As(III) removal onto 500 600 grain size in a batch system. Conditions: 5 mg/L As(III)T, 0.1 g/L Kemiron, I = 0.001 N NaNO3 at fixed pH of 7, and Kemiron adsorbent Figure 5.12: Fractional mass of As(V) removal onto 500 600 grain size in a batch system. Conditions: 5 mg/L As(V)T, 0.1 g/L Kemiron, I = 0.001 N NaNO3 at fixed pH of 7, and Kemiron adsorbent 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0 0.0002 0.0004 0.0006 0.0008 0.001Kd app/Kd ult = Dapp.t/a*a Crank's model data point 0 0.05 0.1 0.15 0.2 0.25 0.3 0 0.0001 0.0002 0.0003 0.0004Kd app/Kd ult D app .t/a*a Crank's model data point

PAGE 74

60 The fractional removal models of As(III) and As(V) onto the less than 38 particles are shown in Figures 5.12. Figure 5.13 and 5.14 compare the results for the model fits of As(V) and As(III) removal onto the less than 38 the 500 grain size. Figure 5.13: Fractional mass of As removal model in a batch system. Conditions: 5 mg/L AsT, 0.1 g/L Kemiron, I = 0.001 N NaNO3 at fixed pH of 7, and Kemiron grain size There have been some reported cases of va rious diffusion coefficients for the same adsorbent but with different grain sizes or with different initial concentrations. According to Ball (1990), situations like these indicate one of the following three possible reasons: 1) the diffusion coefficient may be a function of the particle size; 2) the length scale may not be actually the particle radius; and 3) the kinetic experimental data may not be good. In Table 5.3, the relativ ely low value of Dapp /a2 of As(V) onto the 500 600 was expected because of the high values of the particle radius, a. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1E 05 0.0001 0.001 0.01 0.1Kd app/Kd ult = Dapp*t/a*a As(III) As(V)

PAGE 75

61 Figure 5.14: Kinetics of As(V) removal model in a batch system. Conditions: 5 mg/L As(V)T, 0.1 g/L Kemiron, I = 0.001 N NaNO3 a t fixed pH of 7, and Kemiron grain sizes used are 38 Figure 5.15: Kinetics of As removal model in a batch system. Conditions: 5 mg/L AsT, 0.1 g/L Kemiron, I = 0.001 N NaNO3 at fixed pH of 7, and Kemiron grain sizes used are 38 m and 500 For a constant value of Dapp for particles, the value of Dapp /a2 is expected to be lower for the 500 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 5 10 15 20Kd app/Kd ulttime (d) As(V) diffusion onto 500 600 microns As(V) diffusion onto 38 microns 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 5 10 15 20Kd app/Kd ulttime (d) As(III) diffusion onto 38 microns As(III) diffusion onto 500 600 microns As(V) diffusion onto 500 600 microns

PAGE 76

62 Table 5.4 shows diffusion rate constants of As adsorption reported by others in relation to the grain sizes of the adsorbents. Granular Ferric Hydroxide (GFH) shows three orders of diffusion coefficient s for As(V) adsorption onto the grain size range of 0.6 2.0 mm. Table 5.3: Grain sizes and intraparticle diffusion rate constants of As removal. Conditions: in binary systems, pH 7, I = 0.001 N NaNO3, CO2 Kemiron size fraction (m) absent, room temperature. Solute 2a Dapp (10 8 s 1 ) As(V) 32 500 600 As(III) As(V) 25 0.02 500 600 As(III) 0.07 Table 5.4: Intraparticle diffusion coefficients of As removal onto other iron oxide. Conditions: pH = 7 at room temperature. Adsorbent Grain size Adsorbate appD (10 11 cm 2 Source /s) GFH GFH Iron oxide Iron oxide modified GAC 0.8 1.0 mm 0.6 2.0 mm 0.6 mm As(V) As(V) As(V) As(V) 203.0 324.0 6.4 1.0 90.5 Badruzzaman et al. (2004) Vaughan et al. (2007) Hristovski et al. (2009) Thirunavukkarasu et al. (2003a) 5.4.2 Effect of A rsenic C oncentration on D iffusion This test was limited to As(V) adsorption onto 500 made here was that the trend as exhibited by As(V) would also be exhibited by As(III)

PAGE 77

63 under the same conditions It was expected that the rate of As(V) uptake would increase with increasing initial concentration s of As(V) due to larger concentration gradients onto particle surfaces The rate of As(V) adsorbed was subsequently observed to have higher removal gradien ts with higher initial As concen trations as shown in Figure 5.16. This suggests that the As(V) removal rate depended on the initial concentration. While the slopes leveled off between the 300th and 400th hour after the spiking of the 5 mg/L initial concentration, the times for the leveling off seemed to shift to the right as the initial As concentration increased from the 5 mg/L to the 20 mg/L As (V). Sorption of only the 5 mg/L and 10 mg/L As(V) concentrations were done on the m fraction, henc e limiting the determination of Dapp/a2 to only these two concentrations since equilibrium was not reached over the duration of the experiment for the 500600 m fraction. The results are given in Table 5.5. For the two initial concentration s used Dapp/a2 differed by an order of magnitude. Table 5.5: Effect of initial As(V) concentration on mass loadings. Effects on intraparticle diffusion rate constants in binary systems at pH 7, I = 0.001 N NaNO3 As(V) conc. (mg/L) onto 500 600 Mass of As(V) sorbed (mg/g) 2/ a Dapp (108 /s) 5 25.49 0.02 10 35.50 0.5 15 40.00 Not determined 20 51.00 Not determined

PAGE 78

64 Figure 5.16: Effect of initial As(V) concentration on rate of uptake. Conditions: As(V) at pH 7, I = 0.001N NaNO3 particle size of 500 Figure 5.17: Model of Fractional mass of As(V) removal. Conditions: batch system for 5 m & 10 g/L As(V)T, 0.1 g/L Kemiron, I = 0.001 N NaNO3 at fixed pH of 7, and Kemiron grain size = 500 600 0 5000 10000 15000 20000 0 100 200 300 400 500 600As(V) (ug/L) time (h) 5 mg/L As(V) 10 mg/L As(V) 20 mg/L As(V) 15 mg/L As(V) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.02 0.04 0.06 0.08 0.1Kd app/Kd ult = Dapp.t/a*a 10 mg/L As(V) 15 mg/L As(V) 20 mg/L As(V)

PAGE 79

65 5.5 Batch E quilibrium S orption of A rsenic Batch As adsorption studies were conducted under the following sets of conditions: 1 mg/L, 5 mg/L, and 10 mg/L initial As concentrations; 0.001 N NaNO3 and 0.1 N NaNO3 for ionic strength; over a range of pH of 4 10; 0.1 g/L Kemiron dose with grain size of 38 m for shorter equilibration time, and the absence of CO2 for adsorption competition prevention in the systems unless otherwise noted. The objectives of these ex periments were to evaluate the impact of pH on As adsorption onto Kemiron and to evaluate the impact of background ionic strength on the adsorption. For another set of batch experiments the objective was to determine the impact of the presence of competin g ions on the adsorption of arsenic. Ions like CO3 2 -, SO4 2 -, NH4 +N, and Ca2+, were used to represent the commonest competing bivalent inorganic ions found in landfill leachate. The impact of Ni2+ The equilibrium sorption experiments in the binary systems shown in Figure 5.18 were conducted for 5 mg/L and 10 mg/L total As(V) concentrations while As(III) concentrations of 1 mg/L, 5 mg/L and 10 mg/L were conducted and are shown i n Figure 5.19. The adsorption of As(V) increased as a pH decreased as shown in Figure 5.18 which can be explained in terms of a combination of a positive surface charge as pH decreases and the negatively charged As(V) species The dominating species of As (V) up to pH 2.2 was H and Se(IV) were also evaluated as representatives of trace co contaminants. 3AsO4. Between pH 2.2 and 7.0, H2AsO4 -, species dominated while HAsO4 2 dominated between pH of 7.0 and 12.1.

PAGE 80

66 Figure 5.18: Batch equilibration tests of As(V) onto 38 m Kemiron grain size. Conditions: in CO2 free binary systems and at room temperature. Figure 5.19: Batch equilibration tests of As(III) onto 38 m Kemiron grain size. Conditions: in CO2 free binary systems and at room temperature. 0 10 20 30 40 50 60 70 80 90 100 4 5 6 7 8 9 10 11%As(V) AdsorbedpH 5ppmAs(V), I=0.001N NaNO3 5ppmAs(V), I=0.1N NaNO3 10ppmAs(V), I=0.001N NaNO3 0 10 20 30 40 50 60 70 80 90 100 4 5 6 7 8 9 10 11% As(III) AdsorbedpH 1 mg/L As(III), I = 0.1 N NaNO3 1 mg/L As(III), I = 0.001N NaNO3 5 mg/L As(III), I = 0.1N NaNO3 5 mg/L As(III), I = 0.001N NaNO3 10 mg/L As(III), I = 0.1N NaNO3 10 mg/L As(III), I = 0.001N NaNO3

PAGE 81

67 Figure 5.19 shows the variation of As(III) sorption with pH, the effect of ionic strength and the effect of total As(III) concentration. Like other studies, the sorption is dome shaped and peaks in the pH 8 to 9 region for all total concentrations, though at 1 mg/L As(III) this was least pronounced (Chakraborty et al. 2007). This peak can be explained in terms of both the aqueous speciation of As(III) and the surface charge as a function of pH. Under the solution conditions studied, As(III) remains as an uncharged H3AsO3 species until the first pKa (9.32) after which it becomes the negatively charged H2AsO3 -For all of the total initial As (III) and As(V) concentrations studied the adsorption behavior was also unaffected when the background ionic strength varied from 0.001 N NaNO species. The pzc of iron oxide surfaces occurs around pH 89.5 where they are positively charged below the pzc and negatively charged above (Naeem et al. 2007, Sperlich et al. 2005) Hence, the maximum sorption between As(III) species and the surface occurs in the vicinity where the As(III) is negatively charged and the surface is positively charged, which is expected. 3 to 0.1 N NaNO3, an observation seen by others (Smith and Naidu 2009). Ionic strength has usually been used as a wet chemical diagnosis for whether an ion was bound strongly (usually referred to as an inner sphere complex) or weakly (usually referred to as an outer sphere complex) where unchanged sorption as a function of ionic strength was attributed to inner spher e type sorption mechanisms (McBride 1997; He et al., 1997; Hayes et al., Many iron based adsorbents have their pH point of zero charge (pH 1998). The results shown in Figure 5.19 suggest that As(III) binds in an inner sphere type mechanism to Kemiron. pz c) between 8.0 and 9.5 (Naeem et al. 2007, Sperlich et al. 2005). At the pHPZC, the charges on the

PAGE 82

68 ad sorbent surface are neutral and pH values below and above the pHpzc result in net positive and net negative surface charges respectively. In our equilibr ation experiment, the adsorption edge showed a considerable drop around pH 7.5 for As(V) which is mainly found as negatively charged species in this pH range. Th e sorption edges of both As(V) and As(III) crossed near pH 7.5 shown in Figure 5.20 with highe r As(III) sorption above pH 7.5. In this pH region where the surface changes from a net positive to a net negative/neutral charge, sorption of the uncharged As(III) species is favored over sorption of the dominant negatively charged As(V) species. Fig ure 5.20: Batch equilibration tests of both As(V) and As(III) onto 38 m Kemiron. Conditions: in CO2 free binary systems and at room temperature. 5.6 Arsenic A dsorption I sotherms Adsorption isotherms are usually used to determine the density of surface hydroxyl sites (sites per unit surface area), and to determine the type of adsorption model that best fits the contaminant removal data. In this work, As(V) and As(III) adsorption data were analyzed using Langmuir and Freundlich models and Table 5.6 lists some of 50 55 60 65 70 75 80 85 90 95 100 4 6 8 10% As AdsorbedpH 5 mg/L As(III), I = 0.001N NaNO3 5 mg/L As(V), I = 0.001N NaNO3

PAGE 83

69 the previous research conditions and model fits used. These models were adopted to describe and to compare the relationship between the amount of As loadings on the Kemiron surface and the concentration of arsenic in solution at equilibrium, for constant pH and temperature. Figures 5.21 and 5.22 show As(III) and As(V) sorption isotherms at pH values that range from 6 to 9. The best fits to the experimental data using the F reundlich model for As(III) and Langmuir model for As(V) are also shown in those figures as lines. Table 5.6: Isotherms of As adsorption onto various adsorbents. Adsorbate species Adsorbent Isotherm model Reference As(V) As(III) As(III) As(III)/(V) As(III) As(V) GFH U. cylindricum U. cylindricum Fe(III) Ti(IV) Kemiron Kemiron F L D R L/F F L Abdallah and Gagnon (2009) Sari & Tuzen (2009) Sari & Tuzen (2009) Ghosh et al. (2004) This work This work F Freundlich, L Langmuir, D R DubininRadushkevich. The experimental data in Figure 5.21 shows that As(III) sorption capacity continues to increase under the conditions studied and that for a given pH value, the corresponding amount of arsenic on t he surface increases as a function of pH. The differences in the amount sorbed at a given pH is not great and can be explained by referring to Figure 5.19 which plotted As(III) sorption edges as a function of pH. For the pH range presented in Figure 5.21, As(III) sorption is at its maximum which represents a plateau on the dome shaped sorption curve. It is obvious from the isotherm plots that a Linear model (q =KC) would apply to neither As(III) or As(V) across the full range of sorption densities. The As(V) adsorption isotherms shown in Figure 5.22 begin to

PAGE 84

70 plateau under the experimental conditions, and for a given aqueous equilibrium condition the sorption capacity increases as pH decreases. This is consistent with the results shown in Figure 5.18. C ompared with As(III) for the same pH value, the maximum As(V) sorption capacity observed is lower. For example, the capacity for As(V) at pH 7 is somewhere around 86 mg As/ g sorbent whereas it is greater than 90 mg As/ g sorbent for As(III). At lower pH values As(V) is favored and the capacity for sorption by Kemiron would be greater. It is also possible that a cluster effect causes the lower capacity observed for the case of As(V). While at the pH 7 As(III) exists as an uncharged molecule, As(V) exis ts as a charged ion. It is possible that the binding between the As(V) and the surface during diffusion into the pores forms clusters which hinder the movement of other dissolved ions. Figure 5.21: Effect of pH on As(III) adsorption isotherm in pure system. Freundlich Model fits and experimental data. Conditions: room temperature, I = 0.001N NaNO3 adsorbent grain size of 9 29 49 69 89 109 129 0 5 10 15 20 25 30 35 40 Concentration (mg/l) q(mg/g) As(III) data point at pH 6 As(III) data point at pH 7 As(III) data point at pH 9

PAGE 85

71 Figure 5.22: Effect of pH on As(V) adsorption isotherm in pure system. Langmuir model fits (lines) and experimental d ata shown. Conditions: room temp, I = 0.001N NaNO3 adsorbent grain size of The use of Langmuir and Freundlich models to fit the As(V) and As(III) sorption were based on the assumption that the sorption and the interactions between Kemiron and As followed the same conditions upon which the models were derived. The best fit curves were obtained with Gauss Newton analyses and confirmed by linear least squares methods. This confirmation was done by plotting the predicted data (from the selected m odel) on the y axis and the experimental data on the x axis. A correlation coefficient (r2) was then be derived for a straight line fit through the origin with a slope 1. The empirical constants qmax, KL, KfTable 5.13 summarizes the model fits to experimental data as well as the results from the linearization process used to determine the best model to adopt. Figures 5.23 to and 1/n of the models were determined from the Gauss Newton algorithm. Appendix B indicates the experimental and the predicted data used to evaluate the empirical constants as well as the sum of squares of residual errors. 10 20 30 40 50 60 70 80 90 0 10 20 30 40 50 Concentration of As(V)(mg/L) q(mg/g) data point at pH 9 data point at pH 7 data point at pH 8

PAGE 86

72 5.25 are example plots of the linearilization results. The Langmuir model (r2 = 0.99) fits As(V) adsorption onto Kemiron better than the Freundlich model (r2 < 0.92) with maximum adsorption densities ranging between 68 mg As(V)/g solid at pH 9 and 88 mg As(V)/g solid at pH 7. The Freundlich model (r2 > 0.95) on the other hand fits As(II I) adsorption better than the Langmuir model (r2 < 0.95) with the coefficient Kf increasing between pH values of 6 and 9. The differences between the r2Given the model fits, what can we infer about As(III) and As(V) sorption? The Langmuir model assumes monolayer coverage as well as uniform surface sites whereas the Freundlich model accounts for site heterogeneity. For the pH range studie d in these isotherms, As(III) would exist mainly as an uncharged species. Given its first pK values for the two models were not as significant for As(III) (as great as 0.07) as they were for As(V ) (as great as 0.15). The sum of squares of residual error can also be used to infer best fits to experimental data where a value closest to zero is preferred. The standard error of estimate of the Langmuir models to As(V) sorption were ~2.87, 1.85, and 2 .46 at pH values of 9, 8, and 7 respectively. The standard error of estimate of Freundlich models to As(III) sorption were 3.45, 1.68, and 3.00 at pH values of 6, 7, and 9 respectively. This also suggests good fits between the experimental data and the m odels selected. a value of 9.23, the isotherm at pH 9 would include much higher concentrations of the negatively charged anion H2AsO3 2 -. Although the differences between the Langm uir and Freundlich fits to As(III) isotherm data are not great, geochemistry can be used to explain the better results obtained from the Freundlich model. The surface complex formed between the uncharged As(III) species and the adsorbent could be due to a site type that is different from that involved with the complexation of negatively charged species. Over the pH

PAGE 87

73 range studied, all As(V) species were negatively charged and it is possible that they interact with a uniform site type, hence the better fits obtained with the Langmuir model. Although the adsorption of As(V) and As(III) seems to follow Langmuir and Freundlich models respectively, there is a clear indication that a linear adsorption isotherm model would fit a relatively narrower range of initia l concentrations. This latter point has implications for the modeling previously done with the rate of uptake data where a linear adsorption model was assumed. Given the limited concentration range used, that assumption was appropriate for the purposes of this research. Figure 5.23: Experimental data and predicted data of As(V) sorption at pH 8. Conditions: I = 0.001 N NaNO3 ; Langmuir model used. R = 0.9928 10 20 30 40 50 60 70 80 10 30 50 70 90Predicted As(V) (mg/g)Experimental As(V) (mg/g)

PAGE 88

74 Figure 5.24: Experimental data and predicted data of As(V) sorption at pH 7. Condit ions: I = 0.001 N NaNO3 ; Langmuir model used. Figure 5.25: Experimental data and predicted data of As(III) sorption at pH 6. Conditions: 3 ; Freundlich model used. R = 0.9914 10 20 30 40 50 60 70 80 90 10 30 50 70 90Predicted As(V) (mg/g)Experimental As(V) (mg/g) R = 0.9929 0 20 40 60 80 100 120 0 20 40 60 80 100 120Predicted As(III) (mg/g)Experimental As(III) (mg/g)

PAGE 89

75 Table 5.7: Isotherm parameters of As onto 38 m Kemiron particles. Conditions: binary systems of 0.001 N NaNO3; CO2 absent and at room temperature. Species pH Freundlich parameters Langmuir parameters K 1/n f r q 2 max K (mg/g) r L 2 As(III) 6 29.46 0.3 4 0.9 9 103 0.2 5 0.94 7 35. 81 0.3 1 0.98 115 0.23 0.95 9 42.54 0.2 8 0.9 5 1 23 0. 21 0.92 As(V) 7 44.7 1 0.1 5 0.84 87 0. 34 0.9 9 8 35 0.24 0.86 82 0.31 0.99 9 20 0.33 0.92 68 0.39 0.99 5.7 Effect of P resence of C ompeting I ons and C o C ontaminants Figure 5.26 shows the effect of 5 mg/L (63 M) Se(IV) or 5 mg/L (85 M) Ni (II) on 5 mg/L (65 M) As(III) sorption to Kemiron. On a molar basis, all three concentrations were comparable and both Ni(II) and Se(IV) resulted in reduced As(III) sorption across all pH val ues with Se(IV) having a greater effect and with a lower percentage reduction due to the presence of either ion as the pH increased. For example, at pH 7 As(III) sorption was reduced from close to 80% to 70% in the presence of Ni(II) and to 50% in the pre sence of Se(IV). At pH 8, the As(III) sorption was reduced from approximately 85% to close to 80% in the presence of Ni(II) and 65% in the presence of Se(IV). In solution Se(IV) would form H2SeO3 which dissociates based on its pKa values of 2.63 and 8.4 and hence in the pH range considered in this study, the main form of Se(IV) would be HSeO3 and SeO3 2 -.

PAGE 90

76 Figure 5.26: Effect of Se(IV) or Ni2+ on As(III) adsorption. Conditions: 0.1 g/L Kemiron 3, CO2 excluded. Figure 5.27: Effect of Se(IV) or Ni2+ on As(V) adsorption. Conditions: 0.1 g/L Kemiron 3, CO2 excluded. 20 30 40 50 60 70 80 90 100 4 5 6 7 8 9 10 pH % As(III) adsorbed 5 mg/L As(III) only 5 mg/L Se(IV) present 5 mg/L Ni2+ present 40 50 60 70 80 90 100 4 5 6 7 8 9 10 pH % As(V) adsorbed 5 mg/L Ni2+ present 5 mg/L As(v) only 5 mg/L Se(IV) present

PAGE 91

77 The higher percentage of positively charged surface sites below the pzc, would attract the negatively charged Se(IV) species and th is is confirmed in Figure 5.28 where Se(IV) sorption onto Kemiron increased as pH decreased with the slope of the sorption edge beginning to change around pH 8 and approximately 90% of Se(IV) being sorbed at pH 7. On a molar basis, the moles of sorbate (A s(III) + Se(IV)) used in Figure 5.26 is similar to that of just 10 mg/L As(III) which is approximately 60% at pH 7 from Figure 5.18, and to that of just 10 mg/L Se(IV) which is also approximately 60% at pH 7 from Figure 5.28. Figure 5.28: Se(IV) sorpt ion as a function of pH Conditions: I = 0.001 N NaNO3 and 0.1 N NaNO3, CO2 excluded. Figure 5.27 shows the effect of 5 mg/L (63 M) Se(IV) or 5 mg/L (85 M) Ni (II) 0 10 20 30 40 50 60 70 80 90 100 4 5 6 7 8 9 10 pH % Se(IV) Adsorbed 5ppm Se(IV), I=0.1N NaNO3 5ppm Se(IV), I=0.001N NaNO3 10ppm Se(IV),I=0.1N NaNO3 10ppm Se(IV),I=0.001N NaNO3 15ppm Se(IV), I=0.1N NaNO3 15ppm Se(IV), I=0.001N NaNO3 on 5 mg/L (65 M) As(V ) sorption to Kemiron. In the presence o f 5 mg/L Se(IV), As(V) adsorption dropped by 20% between pH 4.5 and 9 which was less than that observed for

PAGE 92

78 As(III) (40% between pH 5 and 7 and by 20% between pH 7 and 9). Ni(II) did not reduce As(V) sorption across the pH range studied unlike what was observed in Figure 5.26 for As(III). Although Ni(II) sorption to Kemiron by itself was not examined, it is expected to sorb as a typical cation which means its sorption would increase as pH increases (i.e., as the surface becomes increasingly negative it w ould attract more of the positively charged Ni(II) ions). Figure 5.29 shows the effect of either 1000 mg/L (16.7 mM) CO3 2 or 1000 mg/L (10.4 mM) SO4 2 -The effect of sulfate on As(III) sorption decreased as pH increased and can be explained in terms of a competitive sorption mechanism whe re sulfate affinity for the on 5 mg/L As(III) sorption to 0.1 g/L Kemiron. The molar concentration of carbonate used was approximately 1.6 times that of sulfate and may be one reason why the effect of carbonate was greater. Even though the carbonate and sulfate concentrations are more than two orders of magnitude greater than that of As(III), the amount of As(III) reduced is not as pronounced as seen in the case of either Ni(II) or Se(IV) above. Sulfate sorption to mineral oxides typically increases as pH decreases whereas that of carbonate plateaus around pH 6.5 and the amount sorbed of either of the two is reduced as ionic strength increases ( Zhang and Sparks, 1990; He et al., 1997;Villalobos and Leckie, 2000). T he affinity of carbonate and sulfate for adsorption to mineral oxid es is considered low to moderate (Sposito, 1989), however, they have been seen to reduce the sorption of other anions (e.g. selenite) when present in extremely high concentrations (Balistrieri and Chao, 1987; Appelo et al. 2002). In some cases, carbonate enhanced oxyanion sorption to mineral oxides, as was seen in the case of phosphate on goethite (Wijna et al., 2000).

PAGE 93

79 surface decreased as pH increased (He et al., 1997) at the same time that As(III) affinity reached its maximum. Villalobos and Leckie (2000) found that carbonate sorption to goethite peaked close to the first acidity constant f or H2CO3 in the pH 6 region and hence its sorption curve is similar to that of As(III), just that the peak occurs around pH 6 versus between pH 8 and 9 and seen in Figure 5.19. Along with the higher carbonate concentration used when compared with sulfate, this would also explain why carbonate reduces As(III) sorption more than sulfate. Figure 5.29: Effect of CO3 2 or SO4 2 on As(III) adsorption. Conditions: 0.1 g/L Kemiron 3 and 0.1 N NaNO3 CO2 excluded from SO4 2 40 50 60 70 80 90 100 4 6 8 10 12 pH % As(III) adsorbed 5 mg/L As(III) only 1000 mg/L SO42present 1000 mg/L CO32experiments.

PAGE 94

80 Figure 5.30: Effect of CO3 2 or SO4 2 on As(V) adsorption. Conditions: 0.1 g/L Kemiron 3 and 0.1 N NaNO3 CO2 excluded from SO4 2 experiments. The presence of CO3 2 (1 or 1000 mg/L) and SO4 2 -The idea of introducing ions of opposite char ge into systems in order to increase the mass of adsorption has been explored by many researchers and has shown to work many of the times. For instance, Schindler et al. (1990) showed that the presence of anions in solutions might enhance cation adsorption by forming mixed metal/ligand surface complexes whilst Davis and Bhatnagar (1995) showed that humic acids increased Cd adsorption onto the hematite surface. (1 or 1000 mg/L) had very little or no effect on the percentage of 5 mg/L As(V) sorbed (Figure 5.30). These ions are generally found closer to the higher concentration range in landfill leachate and the results from Figures 5.29 and 5.30 suggest that As( V) removal would be favored over As(III), however, Section 5.4 did show the rate of As(III) sorption to be faster than that of As(V) though the differences (from a practical standpoint) may not be significant. 30 40 50 60 70 80 90 100 3 5 7 9 11 pH % As(V) adsorbed 5 mg/L As(V) only 1 mg/L CO32present 1 mg/L SO42present 1000 mg/L CO32present 1000 mg/L SO42present

PAGE 95

81 Figure 5.31: Effect of Ca2+ or NH4 + N on As(III) adsorption. Conditions: 0.1 g/L Kemiron (< 3 and 0.1 N NaNO3 CO2 excluded. Figure 5. 32: Effect of Ca2+ or NH4 + N on As(V) adsorption. Conditions: 0.1 g/L 3 and 0.1 N NaNO3 CO2 excluded. Typical landfill leachate contains Ca2+ and N H4 + in significant concentrations so the impact they have on sorption is important to understand in addition to their potential for enhancing sorption behavior. Figure 5.31 shows the effect of 0.1 mg/L Ca2+ and 300 mg/L NH4 +N on 5 mg/L As(III). The pres ence of 0.1 mg/L Ca2+ 40 50 60 70 80 90 100 3 5 7 9 11 pH % As(III) adsorbed 5 mg/L As(III) only 0.1 mg/L Ca2+ present 300 mg/L NH4+ N did not increase the amount of As(III) sorbed between pH 4 to 7. It actually caused about a 20% drop in the 50 55 60 65 70 75 80 85 90 95 100 3 4 5 6 7 8 9 10 pH % As(V) adsorbed 0.001 mg/L Ca2+ 0.1 mg/L Ca2+ present 5 mg/L As(V) only

PAGE 96

82 amount of As(III) sorbed at pH 7 and above. The presence of 300 mg/L NH4 + Figure 5.32 shows the effect of Ca did not impact the amount of As(III) sorbed from pH 4 to 8 significantly (Figure 5.31). 2+ (0.001 mg/L and 0.1 mg/L) on 5 mg/L As sorption. These results indicated that while 0.001 mg/L of Ca2+ had no impact on the am ount of As(V) sorbed between pH 5.5 and 9, the presence of 0.1 mg/L of Ca2+ increased the sorption of As(V) up to 100% between pH 4 to 7. This enhanced sorption could be due to the formation of a more favorable surface complex involving As(V) and Ca2+ spe cies, or the favorable surface charge achieved by the presence of Ca2+ on the surface. Compared to the As(III) case, As(V) sorption is more favorable when Ca2+ is present and this could be a potential asset given high calcium levels in leachate. 5.8 Impa ct of O xidation R eduction P otential (ORP) on As(V) A dsorption The impact of ORP on As(V) removal was assessed in a binary system at two different ORP values, 295 mV and of 100 mV. Both tests were conducted at pH 7 for a total As(V) concentration of 5 mg /L. The results indicated that 90% of initial 5 mg/L As(V) was adsorbed at 295 mV, while only about 60% of the 5 mg/L As(V) was adsorbed at 100 mV. The amount of As adsorbed (shown in Table 5.8) in the presence of Ca2+, CO3 2 -, COD, NH4 + N, Se(IV), Ni2+ or by the increase or decrease of ORP or pH was used to select the key factors for further testing in the landfill leachate system. A baseline of 12% increase or decrease in the As sorption when there was a change in value of a factor was used for the selection. Ca2+, Ni2+, Se(IV), ORP, and pH were the parameters that qualified for the further test in the synthetic landfill leachate systems.

PAGE 97

83 Table 5.8: The impact of the various factors on the fractions of As adsorbed. Conditions: in binary system at p H 7, I = 0.001 N NaNO3Parameter at room temperature. Low ( ) High ( + ) % As Adsorbed As(V) ( ) As(V) ( + ) As(III) ( ) As(III) ( + ) pH ORP COD Se(IV) Ni Ca2+ CO2+ 3SO2 4NH2 4 + 5 N 150 mV 5 mg/L 0 mg/L 0 mg/L 0.001 mg/L 1 mg/L 1 mg/L 10 + 295 mV 1000 mg/L 5 mg/L 5 mg/L 0.1 mg/L 1000 mg/L 1000 mg/L 300 mg/L 95 60 82 90 90 88 90 90 62 90 72 70 90 100 81 90 55 82 82 80 81 81 80 90 80 50 70 68 71 75 80 + : data obtained under a high condition of the parameters. : data obtained under a low condition of the parameters. 5.9 Batch E quilibrium S orption of A rsenic onto Kemiron in L andfill L eachate Prior to the batch As adsorption experiment using synthetic landfill leachate, an initial batch test was done with natu ral landfill leachate from Polk Countys North Central facility, Florida. The objective was to determine if Kemiron could remove As in the natural landfill leachate. Samples were collected from 3 locations within the leachate system and the concentrations of total arsenic are reported in Table 5.9. Geochemical parameters measured at the North Central Landfill leachate are also listed on Table 5.10 though these were not measured for the Phase 1 leachate used in Figure 5.34.

PAGE 98

84 Table 5.9: As concentrations in landfill leachate sampled from the Polk C ounty North Central landfill on 4/27/06. Sample As in filtrate through 0.45 m Millipore filter (ppb) As in acid digested filtrate (through 0.45 m Millipore filter) (ppb) As in unfi ltered, digested leachate (ppb) Phase 1 29 2 92 14 90 7 Phase 2 76 4 61 3 64 3 Leachate Tank 98 5 126 6 114 6 Table 5.10: Concentrations of some of the contaminants in the leachate. Source: Polk County North Central leachate tank (Data obtained from Polk County Environmental Services Department, Solid Waste Division). 1 M = 74 .9 g/L As. Date As Ni Cr Bicarbonate (mg/L as CaCO 3 pH ) DO (mg/L) 3/14/02 1.60 1.44 0.14 1318 6.92 2.83 3/06/03 0.53 1.23 < 0.02 1873 7.21 6.21 3/26/04 0.95 2.52 0.38 2913 7.51 4.82 The sorption experiment was carried out using the filtered Phase 1 leachate solution without taking any precautions to eliminate biological effects. This solution had an undigested total As concentration of 0.029 mg/L and an acid digested total concentrat ion of 0.092 mg/L. When used to make the 0.1 g/L Kemiron slurry to which 1 mg/L As(V) was added, the result is shown in Figure 5.34. For the given equilibration period between 40 and 50% of the As(V) was sorbed to the Kemiron in the presence of the Polk County Landfill leachate. The shape of the sorption edge was similar to that seen for As(V) in earlier parts of this chapter. Although the percentage sorbed and overall surface coverage was reduced in this leachate solution, the result suggested that the potential is there provided the right conditions or pretreatment steps are undertaken. In the next section,

PAGE 99

85 research using synthetic leachate solutions is presented. These solutions capture some of the major constituents of leachate systems. 5.33: Adsorption edge of 1 mg/L As(V) on 0.1 g/L Kemiron in a Polk County landfill leachate solution. Conditions: Kemiron adsorbent grain size of at room temperature. 5.9.1 Effects of L andfill A ge and pH on A dsorption To test for the impact of age of the synthetic leachate on As(V) adsorption, 5 mg/L As(V) was subjected to the same conditions in both an acidogenic and a methanogenic landfill leachate solution. The procedure here followed the same steps as the equilibration tests and the results are shown in Figures 5.34 and 5.35. Age had no impact on the As(V) removal. However, there was a slight increase in the percentage adsorption of As(III) in the older landfill leachate (acidogenic) compared to the amount sorbed in the methanogenic leachate as shown in Figure 5.35. pH on the other hand continued to have significant influence on As adsorbed in the synthetic landfill leachate solutions as the As(V) and As(III) sorption followed the same trends as were seen in 0 5 10 15 20 25 30 35 40 45 6 6.5 7 7.5 8 8.5% As SorbedpH

PAGE 100

86 Figures 5.18 and 5.19. Compared to the clean systems with just DI water, both As(III) and As(V) sorption decreased in the presence of leachate by roughly 20% across all pH values. Figure 5.34: Effect of pH or age (acidogenic or methanogenic) of landfill leachate on 5 mg/L As(V) adsorpti on. Conditions: 0.1 g/L Kemiron adsorbent, grain size of at room temperature. Figure 5.35: Effect of pH or age of landfill leachate on 5 mg/L As(III) adsorption. Conditions: 0.1 g/L Kemiron adsorbent, grain size of at room temperature 0 10 20 30 40 50 60 70 80 90 100 5 6 7 8 9 10%As(V) adsorbedpH acidogenic methanogenic DI water, I = 0.001 N NaNO3 0 10 20 30 40 50 60 70 80 90 100 5 6 7 8 9 10% As(III) adsorbedpH DI water only, I = 0.001 N NaNO3 Acidogenic Methanogenic

PAGE 101

87 5.9.2 Effect of Se(IV) P resent in the L andfill L eachate To test for the impact of age of the leachate on As(V) adsorption, 5 mg/L As(V) was subjected to the same conditions in both acidogenic and the methanogenic landfill leachate in the presence of the two contaminants. The procedure here followed the same steps as the equilibration tests. Observation made is shown in Figure 5.37. Here, no apparent differences existed in the percentages of As(V) adsorbed in both leachate systems. However, there was about 20% drop in As(V) removal when compared with the As(V) sorbed in the pure system. The trend of As(V) removal in the landfill leachate also conformed to that of the ternary system with 5 mg/L of Se(IV) present. This thus suggests that Se(IV) as the co contaminant may be the main controlling factor in the As(V) removal in the landfill leachate (see Figure 5.34). Figure 5.36: Effect of Se(IV) in As(V) removal in the synthetic landfill leachate. Condition: 0.1 g/L Kemiron, oom temperature. 20 30 40 50 60 70 80 90 100 5 6 7 8 9 10 11 pH %As(V) adsorbed acidogenic leachate methanogenic leachate DI water, I =0.001 N NaNO3 DI water, 5 mg/L Se(IV) present

PAGE 102

88 5.9.3 Effect of Ca2+The impact of Ca on A rsenic R emoval in L andfill L eachate 2+ in the leachate was also evaluated and Figure 5.3 7 shows that Ca2+ had no impact on As(V) in the old landfill leachate after the Ca2+ concentration was inc reased by 1200 mg/L. 5.9.4 Effect of ORP (Eh Figure 5.38 shows a scatter plot of ORP versus percentage As(V) sorbed onto 0.1 g/L Kemiron ( Experiments were conducted in such a way that pH and ORP were varied by the addition of ni tric acid or sodium sulfide respectively. There was no significant trend or ) on Arsenic Adsorption in Synthetic Landfill Leachate Figure 5.37: Effect of Ca2+ on 5 mg/L As(V) adsorption in synthetic landfill leachate. Condition: 0.1 g/L Kemiron, 30 35 40 45 50 55 60 65 70 75 80 6 7 8 9 10 pH % As(V) adsorbed methanogenic with normal Ca2+ Methanogenic with High Ca2+

PAGE 103

89 correlation be tween the ORP of the batch system and the percentage of As(V) sorbed in the acidogenic leachate. Figure 5. 39 plots the percentage As(V) removed as a function of pH and distinguishes data points where ORP was greater than 0 mV and less than 0 mV, crudely r epresenting oxidizing and reducing environments respectively (Christensen et al., 2001). The plot in Figure 5.40 suggests that arsenic sorption decreased up until pH values around 10 and then sharply rose again in the pH 11 range. The steep slope of the sorption curve above pH 10 suggests that there may be other mechanisms like precipitation dominating As(V) removal. The amount of As(V) removed in this pH range was also not affected by whether the ORP values were greater than, or less than 0. This again suggests that another removal mechanism might be important. Figure 5.38: The impact of ORP on As removal in synthetic landfill leachate. Condition: 0.1 g/L Kemiron, controlled, a cidogenic leachate conditions.. 20 30 40 50 60 70 80 90 100 400 300 200 100 0 100 200 300 400% As removedORP (mV)

PAGE 104

90 Figure 5.39: Impact of pH and ORP on As removal in synthetic landfill leachate. Condition: 0.1 g/L Kemiron, initially, acidogenic leachate conditions. Figure 5.40: Box plot of ORP (mV) as a function of pH. Plot shows values that fall within the 25th and 75th percentile (box), the minimum and maximum loading (line) and the median (diamond). The pH plotted represents the average for the given pH range eva luated from 78, 89, 910, and 1011. Condition: 0.1 g/L Kemiron, size, at room temperature, 5 mg/L As(V) initially. 0 10 20 30 40 50 60 70 80 90 100 7 8 9 10 11 12% As removedpH ORP < 0 mV ORP > 0 mV 400 300 200 100 0 100 200 300 400 500 pH 7.7 pH 8.4 pH 9.6 pH 10.4 pH 11.4ORP (mV)Average pH min max 75th percentile 25th percentile median

PAGE 105

91 These high pH ranges were not studied in the simpler batch systems as Kemiron dissolution would be an issue and the likelihood of such high pH values in leachate solutions may not be that common. The experiments were conducted in such a way that addition of chemicals to change ORP were not done under controlled pH conditions. Instead, the systems were allowed to equil ibrate and the final pH recorded after 72 hours along with the amount of arsenic removed. Below pH 10, the amount of arsenic removed increases as pH decreases, with higher percentage removals seen in samples that had ORP values greater than 0 mV. Figures 5.35 and 5.36 show reductions in As(V) and As(III) sorption in the presence of the acidogenic leachate to levels that are much lower than those observed in Figure 5.40, especially for the points with ORP values > 0 mV. Abiotic redox transformations of t he As(V)/As(III) and/or Fe(III)/Fe(II) species could be occurring during these experiments. Dissolution of Kemiron and precipitation of an amorphous iron oxide phase could be one mechanism to enhance total arsenic removal. Figure 5.41 presents a box plot of ORP (mV) as a function of pH for the same set of data shown in Figures 5.39 and 5.40. In general, the majority of ORP values tend to decrease as pH increases. Geochemical modeling is used in Section 5.11 to discuss the aqueous equilibrium speciation expected for arsenic as a function of pH and ORP. 5.10 Effect of H ydrogen S ulfide on A rsenic A dsorption Figures 5.42 5.44 were derived using Geochemist Workbench software and they show arsenic speciation as a function of Eh (in volts ) and sulfide concentration or pH. Eh and ORP are the same and the graphs below use the default plots from Geochemist workbench. Figures 5.42 sand 5.43 show arsenic speciation as a function of sulfide

PAGE 106

92 concentrations for pHs of 5 and 10 respectively when temperature wa s set at 25oC and atmospheric pressure was set at 1.013 bar. Experiments were conducted under total sulfide concentrations of 1 x 105 and 1 x 103 M which would translate into lower values when plotted as HSconcentrations. Given the ORP and pH values measured in experiments, H2AsO4 -, HAsO4 2 -, As(OH)3, As(OH)4 -, and AsS2 could exist based on Figures 5.42 and 5.43 with less likelihood of the sulfide complex because of the total sulfide concentration added. According to the results generated with Geoch emist Workbench software, no precipitates formed when sulfide was included (from sodium sulfide salt) along with the composition of the synthetic leachate solution, 5 mg/L total As, and assuming a total dissolved Fe concentration ~ 103 Benjamin (2002) assumed that adsorption and precipitation of the same target contaminant occur in parallel and the total contaminant removed would be the sum of the amount removed by each process. In the absence of precipitation as is predicted by the simulations, it could be inferred that the total As removed in the synthe tic leachate solution was purely due to adsorption onto the Kemiron particles. M (a very conservat ive estimate based on EDS elemental composition and the fact that experiments were run with 0.1 g/L Kemiron). The data generated with the Geochemical workbench can be in appendix C

PAGE 107

93 Figure 5.41: Eh log { H S} diagram of inorganic arsenic at pH 5 Conditions: at 25C, at a pressure of 1.013 bar Figure 5.42: Eh log {HS} diagram of inorganic arsenic at pH 10. Conditions: at 25C, and at a pressure of 1.013 bar. 0 .5 0 .5 1log a HS-Eh (volts) As(OH)3AsS2 -H2AsO4 -25Cdoti Mon Aug 24 2009Diagram H3AsO4, T = 25 C P = 1.013 bars a [ main ] = 10.176, a [ H2O] = 1 a [ CH3COO-] = 10.762, a [ Na+] = 10.8477, a [ Mg++] = 10.714, a [ Ca++] = 10.524, a [ NH4 +] = 10.585, a [ Cl-] = 10.223, pH = 5 0 .5 0 .5 1log a HS-Eh (volts) As(OH)4 -AsS2 -HAsO4 --25Cdoti Mon Aug 24 2009Diagram H3AsO4, T = 25 C P = 1.013 bars a [ main] = 10.176, a [ H2O] = 1 a [ CH3COO-] = 10.762, a [ Na+] = 10.8477, a [ Mg++] = 10.714, a [ Ca++] = 10.524, a [ NH4 +] = 10.585, a [ Cl-] = 10.223, pH = 11

PAGE 108

94 Figure 5.43: Eh 2OP pH diagram of inorganic arsenic (g) 2HP= 0.21 bar, at 25C, and (g) = 1 bar. 5.11 Kinetic s of A rsenic in L andfill L eachate A batch kinetic study was done for As(V) onto 38 m in the acidogenic synthetic landfill leachate. The objective was to model and estimate diffusion coefficient of As(V) in the landfill leachate. Another objective was to c ompare the coefficients of As(V) in the binary system with that in the landfill leachate to evaluate the impact of the medium on the arsenic removal. 5.11.1 As(V) D iffusion C oefficient E stimation in L andfill L eachate The rate of diffusion of As(V) into Kemiron was evaluated and modeled based on the assumption that As(V) mig ration was intra particle diffusion controlled. We adopted Cranks (1975) fractional uptake solution to Ficks second law of diffusion to model 0 2 4 6 8 10 12 14 .5 0 .5 1pHEh (volts) As(OH)4 -AsO4 ---As(OH)3AsO2OH--H2AsO4 -H3AsO4HAsO4 --25Cdoti Tue Jul 14 2009Diagram As(OH)3, T = 25 C P = 1.013 bars a [ main ] = 10.699, a [ H2O] = 1

PAGE 109

95 Figure 5.44: Rate of 5 mg/L A s(V) removal onto particle size. Conditions: in a synthetic acidogenic landfill leachate at pH 7.5, ORP of 240 mV, and at room temp. Figure 5.45: Fractional removal model of As onto Kemiron in the synthetic leachate. Figure 5.45 shows two diff erent removal rates for As; the initial faster rate which is followed by a slower rate. Most of the removal occurred within the first 12 hours of 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 time (d) Aqueous As(V) (ug/L) DI water at pH 7 Landfill leachate at pH 7.5 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.05 0.1 0.15 0.2 0.25Kd app/Kd ult = Dapp.t/a*a Crank's model data point

PAGE 110

96 reaction. The apparent dif fusion coefficient ( 2a Dapp ) o f the As was estimated to be 8.3 x 107 s1 compared (fit shown in Figure 5.46) with 3.2 x 107s1 of As(V) or 2.5 x 107s1 of As(III) in the binary system. Hence, in the presence of the leachate Dapp is similar to that for the DI water, but there was a significant drop in the sorption capac ity. 5.12 Maximu m As R emoval onto 38 m P article S ize in L andfill L eachate Previous sections showed that Se(IV) reduced As sorption to Kemiron in both clean systems and under synthetic landfill conditions. Experiments were therefore conducted to furt her evaluate the effect of Se(IV) on As removal as a function of pH and ORP for a total As concentration of 5 mg/L (added as As(V) and in the presence of the acidogenic synthetic leachate. The results of the experimental runs are plotted in Figure 5.47. Areas indicated as having 0% arsenic sorbed should be viewed as areas where no data exists. Maximum arsenic removal ( 90%) occurred at pH 8 (ORPs of 200, 0 and 350 mV), and between pH 11and 12 (ORPs of 300 and 0 mV). Loadings of As onto the Kemiron particles ( m) under the optimum ORP and pH values and in the presence of Se(IV) are tabulated in Table 5.11. The loadings measured are comparable to loadings seen for arsenic on other adsorbent surfaces in less comple x systems like surface water or DI water (Table 5.12). It should be noted that particle size varies in the results presented in Table 5.12 and our work using a fine fraction which has been shown to reach equilibrium faster than larger porous particles. Our loadings in the presence of synthetic leachate solutions are comparable to loadings seen in the literature. The Kemiron sorbent costs between $2 to $4 per pound which falls within the range seen for commercially available sorbents ($0.50 to $50 per pound).

PAGE 111

97 Figure 5.46: Contours of %As sorbed in young synthetic landfill leachate. Conditions: at various ORP and pH values and at room temperature. Table 5.11: Maximum adsorption densities of As. Conditions: under optimum pH and ORP conditions at room temperature, Kemiron particle size m, 0.1 g/L Kemiron, 5 mg/L As(V). Arsenic loadings (mg As/g Kemiron) Conditions ORP (mV) pH 47.5 23.8 47.5 47.5 29.0 47.5 47.5 47.5 320 350 200 350 400 0 300 100 7 7.5 8 8 9.0 11 11 12 pHORP (mV) 7 7.5 8 8.5 9 9.5 10 10.5 11 11.5 12 -300 -200 -100 0 100 200 300 400 0 10 20 30 40 50 60 70 80 90 %As(V) Adsorbed

PAGE 112

98 Table 5.12: As loadings at equilibrium/ breakthrough. Condition: DI/surface water as seen on other adsorbents under various pH and at room temperature. Adsorbent As species Initial conc (mg/L) Loading (mg/g) pH Test type reference GFH Zeolite (H24) Zeolite (H90) U. cylindricum Iron coated zeolite Kemiron As(V) As(V) As(V) As(III) As(V) As 0.1 0.11 10 150 10 150 10 400 2.0 5.0 0.99 1.5 35.8 34.8 67.2 0.68 0.53 47.5 in leachate 8.6 6.5 3.2 6.0 4.0 Column Batch Batch Batch Batch Batch Badruzzaman et al. ( 2004) Chutia et al. ( 2009) Chutia et al. ( 2009) Tuzen et al. (2009) Jeon et al. ( 2009) This work Though they offer rapid equilibration times, the less than 38 m particles are not very practical in full scale treatment since their separation from cleaned solutions would pose a challenge. From the standpoint of further developing this research as a vi able treatment technology for landfill leachate, experiments with larger particle sizes will have to be considered since these can be packed into fixed bed reactors thereby eliminating challenges related to separation of the sorbent from the treated soluti ons. The rate of uptake experiments presented in this work, coupled with the modeling of this data, show that the time to reach equilibrium in these larger particles will be longer. This has implications for the envisioned treatment process, however, it i s likely that optimized particle sizes and configurations can assist with reducing mass transfer resistances within the pore structures. Landfill leachate is a very complex water to be treated for arsenic and this work is the first study that we have seen looking at the use of sorption

PAGE 113

99 technology for such applications. Whilst we found that As(V) could be removed by Kemiron in the presence of filtered leachate from a real landfill, the majority of our experiments were done in relatively clean systems desig ned to capture some of the key characteristics of leachate solutions. In thinking of building on the work done here, researchers should think of combined systems that would be most appropriate for treating leachate which may reduce the presence of ions th at can potentially compete with arsenic for sorption sites.

PAGE 114

100 Chapter 6 Summary, C onclusion, and R ecommendation for F uture R esearch 6.1 Introduction This chapter summarizes all of the experimental results obtained according to their chapters. These experimental results provide the first step in a project aimed at removing arsenic from landfill leachate by adsorption onto mineral oxides either packed into fixed bed column reactors or mixed into leachate at the landfill site. 6.2 Summar y The nitrogen BET surface area of the Kemiron (particle sizes m and 500 600 m) was ~40 m2sizes exhibited an inverse relation between the grain sizes and the rate of sorption at pH 7. For the m grain size both As(III) and As(V) sorption reached equilibrium in ~ 36 hours whereas ~ 374 days was required for the larger grain sizes. For the larger grain sizes As(III) reached equilibrium faster likely because of its major uncharged species /g with ~ 44% of the pore sizes in the 500 600 m fraction less than 3 nm. EDS analysis showed that Kemiron was made up of 40 % Fe, 42% O, 8% carbon and 6 % S. XRD analysis indicated that Kemiron was an agglomeration of microparticles and was classified as amorphous though there was some similarity to goethite. The kinetic studies of As(III) and As(V) onto m and 500 600 m grain

PAGE 115

101 (H3AsO3 2a Dapp ). These penetrated pore spaces easier than the negatively charged As(V) species. Given that 44% of surface area was in pore sizes less than 3 nm, the reduction in particle size by a factor of ~15 from an average of 550 m had a major impact on the time to equilibrium. For a standard fixed bed treatment system, column diameters to particle diameters must be greater than 10 and more research should be done on either determining the optimum particle size for packing into columns (faster approach to equilib rium whilst still preventing column clogging) or alternative ways for removing fines if mixed with leachate in a stirred reactor. Alternatively, redesign of sorbent particles should consider access to particle surface area. Cranks model solution to Fick s law of diffusion was used to estimate diffusion coefficients for As adsorption. for As(V) and As(III) were similar for each grain size though As(III) values were always slightly larger than As(V) values. The larger particles had smaller 2a Dapp which is expected given the inclusion of the particle radius, a. 2a Dapp for As(III) and As(V) on the m particles was 25 and 32 x 108 s1 respectively and on the 500600 m it was 0.07 and 0.02 108 s1 r espectively. If average particle radius values were assumed for each grain size (e.g. 19 m and 550 m) the 4 order of magnitude difference between Dapp would not be accounted for. Given that the 38 m particles may contain more smaller sized particles this could reduce the differences seen for DappThe results of the rate of adsorption involving the various concentrations of As (V) onto the 500 600 m grain size indicated that the adsorption capacity was dependent on the initial As(V) concentrations. The pseudo equilibrium graphs of As(V) sorption onto Tortuosity or constrictivity factors could also be used to account for a larger radius needed for the larger particle sizes.

PAGE 116

102 the 500 600 m showed a shift of the equilibration times to the right as the initial As(V) increased. That indicated that the equilibration times got higher as the initial concentration increased. As(III) and As(V) sorption to Kemiron was pH dependent with As(V) sorption increasing as pH decreased and As(III) sorption having a maximum around pH 8. Ionic strength (0.1 N and 0.001 N NaNO3) had no impact on the removal of either As species, suggesting an innersphere type complexation removal mechanism. Table 5.7 summarizes the Freundlich and Langmuir isotherm fits for As(V) and As(III) onto the 2 = 0.99) fit As(V) adsorption onto Kemiron better than the Freundlich model (r2 < 0.92) with maximum adsorption densities ranging between 68 mg As(V)/g solid at pH 9 and 88 mg As(V)/g solid at pH 7, where r2 is the correlation between the experimental and predicted values. The Freundlich model (r2 > 0.95) on the other hand fit As(III) adsorption better than the Langmuir model (r2 < 0.95) with the coefficient KfB oth Ni(II) and Se(IV) resulted in reduced As(III) sorption across all pH values with Se(IV) having a greater effect and with a lower percentage reductio n due to the presence of either ion as the pH increased. In the presence of 5 mg/L Se(IV), As(V) adsorption dropped by 20% between pH 4.5 and 9 which was less than that observed for As(III) (40% between pH 5 and 7 and by 20% between pH 7 and 9). Ni(II) did not reduce As(V) sorption across the pH range studied unlike what was observed for As(III). The presence of either 1000 mg/L CO increasing between pH values of 6 and 9 and adsorption densities as high as ~100 mg As(III)/g sorbent under given experimental conditions. 3 2 -, 1000 mg/L SO4 2 -, or 300 mg/L NH4 + N had

PAGE 117

103 no major impact on As(V) (carbonate caused a reduction of less than 10% acros s the pH range). With the exception of NH4 + N, they all reduced sorption of As(III) across the pH range by up to 20% in some cases. Ca2+ increased As(V) sorption and reduced As(III) sorption. Whilst ions like carbonate and sulfate will be in high conc entrations in leachate solutions, co contaminants like Se(IV) and Ni(II) will compete with As, especially As(III) for sorption sites. There was no difference between acidogenic and methanogenic leachate systems on either As(V) or As(III) sorption and both caused about a 30% reduction in sorption at pH 8 with that number decreasing to ~ 10% at pH 7 for As(V). The effect of calcium on increased As(V) sorption was not observed in the presence of leachate. In the synthetic landfill leachate pH and ORP were identified as the most influential factors for As(V) removal. Subsequently maximum As(V) removal was achieved at optimum values of pHs 8 and 7.5 and between pH 11and 12 under ORPs of between 200 and 400 mV and between ORPs of 300 and 100 mV respectively. Leachate systems usually lie between pH values of 5 and 8 with positive ORP values, suggesting that the potential for this to work is great for older leachates and for younger leachates pH manipulation may have to be considered to increase removal effici ency. High removal amounts were seen in the high pH range, and the use of CO2 to treat such a high pH solution afterwards could be considered. Similarly, oxidation of As(III) to As(V) would also reduce the effect of co contaminants on overall arsenic re moval.

PAGE 118

104 6.3 Conclusion This research showed that Kemiron could be used to remove As(V) and As(III) from solutions with loadings seen as high as ~90 mg As/g Kemiron in relatively clean systems and could also remove arsenic from complex matrices like landfill leachate. Compared to As(V), As(III) sorption was more sensitive to the presence of cocontaminants (Ni(II) and Se(IV)) and high concentrations of ions like CO3 2 -, SO4 2 and NH4 + N showed reduced sorption whereas As(V) sorption was only reduce d in the presence of Se(IV). Synthetic acidogenic and methanogenic leachate solutions reduced sorption of both As(V) and As(III), with a greater impact seen on As(III) above pH 7, but with little difference seen between the two types of leachate on either ion. Using the acidogenic conditions which had higher concentrations of major ions, As(V) sorption could be manipulated by changes in ORP and pH with the most appropriate pH values seen between 5 and 8. Assuming a loading of 45 mg As/g sorbent (this would be 50% of that seen in the clean system which is a conservative estimate given our results thus far) then the amount of Kemiron needed per year would be 66 kg which, at $4/lb ($9/kg), would cost $600 if treating a 0.1 mg/L As leachate solution (assuming a volume of 7,986,529 gallon/yr as was the case of a Florida landfill). Compared to offsite disposal costs of $110/gallon the potential cost savings for an onsite sorption process could be huge provided the equipment and maintenance costs are not great. 6.4 Recommendations for Future Research Future research work could be categorized into two sections: experimental lab and pilot studies, and modeling.

PAGE 119

105 More experiments are needed to test the application in column systems and in the presence of other cocontaminants. Attempts to optimize particle size and particle morphology are also possible bearing in mind the eventual cost of the material. Tests on real landfill leachate should also be conducted, but done in conjunction with other researcher s trying to remove other contaminants (e.g. organics) or materials (e.g filtration or flocculation pretreatment step). Microbial activity was not considered in this research and their effect on the process should be determined. These tests can be scaled up for pilot testing. More mechanistic sorption models would capture surface complexation that responds to pH changes. A linear adsorption model was assumed for finding apparent diffusivities yet experiments show that this model would not apply to As(III ) and As(V) sorption. Hence, future work could couple a mass transfer model with a more mechanistic adsorption model.

PAGE 120

106 References Abdallah, E. A. M. and G. A. Gagnon (2009) Arsenic removal from groundwater through iron oxyhydroxide coated waste products. Can. J. Civ. Eng., 36(5), 881888. Agrawal, A. and K. K. Sahu (2006) Kinetic and isotherm studies of cadmium adsorption on manganese nodule residue. Journal of Hazardous Materials, 137(2), 915924. Alcantar, N. and T. Pichler (2007) Personal communication. Al Yaqout, A. F. and M. F. Hamoda (2005) Prediction of contaminants migration at unlined landfill sites in an arid climate A case study. Water Air Soil Pollut., 162(14), 247264. Alvarez Vazquez, H., B. Jefferson and S. J. Judd (2004) Membrane bioreactors vs conventional biological treatment of landfill leachate: a brief review. J. Chem. Technol. Biotechnol., 79(10), 10431049. Andrew, A. S., J. L. Burgess, M. M. Meza, E. Demidenko, M. G. Waugh, J. W. Hamilton and M. R. Karagas (2006) Arsenic exposure is associated with decreased DNA repair in vitro and in individuals exposed to drinking water arsenic. Environ. Health Perspect., 114(8), 11931198. Appelo, C. A. J., M. J. J. Van der Weiden, C. Tournassat, and L. Charle t (200 2) Surface complexation of ferrous iron and carbonate on ferrihydrite and the mobilization of arsenic. Environmental Science and Technology 36(14), 30963103 Ashrafizadeh, S. N., Z. Khorasani and M. Gorjiara (2008) Ammonia removal from aqueous solutions by Iranian natural zeolite. Sep. Sci. Technol., 43(4), 960978. Badruzzaman, M., P. Westerhoff and D. R. U. Knappe (2004) Intraparticle diffusion and adsorption of arsenate onto granular ferric hydroxide (GFH). Water Research, 38(18), 40024012. Bajpai, S. and M. Chaudhuri (1999) Removal of arsenic from ground water by manganese dioxide coated sand. Journal of Environmental EngineeringA SCE 125(8), 782784. Balistrieri, L.S. and T.T. Chao ( 1987) Seleni um adsorption by goethite. Soil Sci. Soc. Am. J. 1987( 51), 11451151.

PAGE 121

107 Ball, W. P. (1990) Equilibrium sorption and diffusion rate studies with halogenated organic chemical and sandy aquifer material. Doctoral dissertation, Stanford University, Order number 9017828. Ball, W. P. and P. V. Roberts (1991) Long Term s orption of Halogenate d Organic Chemicals by aquifer m aterial. 2. Intraparticle Diffusion. Environ. Sci. Technol., 25(7), 12371249. Bang, S., G. P. Korfiatis and X. G. Meng (2005) Removal of arsenic from water by zerovalen t iron. Journal of Hazardous Materials, 121(13), 6167. Bayard, R., V. Chatain, C. Gachet, A. Troadec and R. Gourdon (2006) Mobilisation of arsenic from a mining soil in batch slurry experiments under biooxidative conditions. Water Research, 40(6), 12401248. Benjamin, M. M. (2002) Water Chemistry. McGraw Hill, New York. Bodek, I., W. J. Lyman, W. F. Reehl, and D. H. Rosenblatt (1988) Environmental Inorganic Chemistry Pergamon Press, NY, USA. Boni, M. R., A. Chiavola and S. Sbaffoni (2006) Pretreated waste landfilling: Relation between leachate characteristics and mechanical behaviour. Waste Manage., 26(10), 11561165. Catalano, J. G., Z. Zhang, P. Fenter and M. J. Bedzyk (2006) Inner sphere adsorption geometry of Se(IV) at the hematite (100) water interface. Journal of Colloid and Interface Science, 297(2), 665671. Chakraborty, S., M. Wolthers, D. Chatterjee and L.Charlet (2007) Adsorption of arsenite and arsenate onto muscovite and biotite mica. Elkin 06. International Electrokinetics Conference, June 2529, N ancy, France, 309(2), 392. Chapra, C. S. and R. P. Canale (1998) Numerical Methods for Engineers McGraw Hill Co.: Singapore, 3rd ed. Chiou, H. Y., Y. M. Hsueh, K. F. Liaw, S. F. Horng, M. H. Chiang, Y. S. Pa, J. S. N. Lin, C. H. Huang and C. J. Chen (1995) Incidence of Internal Cancers a nd Ingested Inorganic Arsenic A 7 Year Follow Up Study In Taiwan. Cancer Res., 55(6), 12961300. C howdhury, U. K., B. K. Biswas, T. R. Chowdhury, G. Samanta, B. K. Mandal, G. C. Basu, C. R. Chanda, D. Lodh, K. C. Saha, S. K. Mukherjee, S. Roy, S. Kabir, Q. Quamruzzaman and D. Chakraborti (2000) Groundwater arsenic contamination in Bangladesh and West B engal, India. Environ. Health Perspect., 108(5), 393 397.

PAGE 122

108 Christensen, T. H., P. L. Bjerg, S. A. Banwart, R. Jakobsen, G. Heron and H. J. Albrechtsen (2000) Characterization of redox conditions in groundwater contaminant plumes. J. Contam. Hydrol., 45(34), 165241. Christensen, T. H., P. Kjeldsen, H. J. Albrechtsen, G. Heron, P. H. Nielsen, P. L. Bjerg and P. E. Holm (1994) Attenuation of Landfill Leachate Pollutants In Aquifers. Crit. Rev. Environ. Sci. Technol., 24(2), 119 202. Chutia, P., S. Kato, T Kojima and S. Satokawa (2009) Arsenic adsorption from aqueous solution on synthetic zeolites J. Hazard. Mater., 162(1), 440447. Cincotti, A., A. Mameli, A. M. Locci, R. Orru and G. Cao (2006) Heavy metals uptake by Sardinian natural zeolites: Experimen t and modeling. Industrial and Engineering Chemistry Research, 45(3), 10741084. Cooke, A. J., R. K. Rowe, B. E. Rittmann, J. VanGulck and S. Millward (2001) Biofilm growth and mineral precipitation in synthetic leachate columns. Journal of Geotechnical and Geoenvironmental Engineering, 127(10), 849856. Crank, J. (1975) The Mathematics of Diffusion. Oxford University Press: Oxford, U.K., 2nd ed. Cunningham, J. A., C. J. Werth, M. Reinhard and P. V. Roberts (1997) Effects of GrainScale Mass Transfer on the Transport of Volatile Organics Through Sediments 1. Model Development. Water Resour. Res., 33(12), 27132726 Davis, A. P. and V. Bhatnagar (1995) Adsorption of Cadmium and Humic Acid onto Hematite. Chemosphere, 30(2), 243256. Davis J.A. and D.B. Kent (1990) Surface complexation modeling in aqueous geochemistry. In: Hochella M.F., Jr., White A.F., eds. Reviews in mineralogy, Vol. 23: Mineral water interface geochemistry. Washington, DC: Mineral ogical Society of America, 177260. Duggan, J. ( 2005) The potential for landfill leachate treatment using willows in the UK A critical review. Resour. Conserv. Recycl., 45(2), 97113. Dzombak, D. A. and F. M. Morel (1990) Surface Complexation Modeling, Hydrous Iron Oxide. Wiley Interscience, New York El Fadel, M., E. BouZeid, W. Chahine and B. Alayli (2002) Temporal variation of leachate quality from pre sorted and baled municipal solid waste with high organic and moisture content. Waste Manage., 22(3), 269282.

PAGE 123

109 Entezari, M. H. and T. R. Bastami (2006) Sonosorption as a new method for the removal of lead ion from aqueous solution. Journal of Hazardous Materials, 137(2), 959964. Fan, H. J., I. W. Chen, M. H. Lee and T. Chiu (2007) Using FeGAC/H2O2 process for landfill leachate treatment. Chemosphere, 67(8), 16471652. Fan, H. J., H. Y. Shu, H. S. Yang and W. C. Chen (2006) Characteristics of landfill leachates in central Taiwan. Science of The Total Environment, 361(13), 25 37. Ghosh, A., M. Mukiibi and W. Ela (2004) TCLP underestimates leaching of arsenic from solid residuals under landfill conditions. Environmental Science and Technology, 38(17), 46774682. Gimenez, J., M. Martinez, J. de Pablo, M. Rovira and L. Duro (2007) Arsenic sorption onto natural hematite, magnetite, and goethi te. Journal of Hazardous Materials, 141(3), 575580. Goldberg, S. (2002) Competitive adsorption of arsenate and arsenite on oxides and clay minerals. Soil Science Society of America Journal, 66(2), 413421. Gregg, S. J. and K. S. W. Singh (1982) Absorpti on, Sorption Area and Porosity Academic Press, London and New York. Hayes, K. F., and Leckie, J. O. (1987) Modeling ionic strength effects on cation adsorption at hydrous oxide/solution interface. Journal of Colloid and Interface Science, 115(2), 56472. Hayes K.F., C. Papelis and J. O. Leckie ( 1998) Modelling ionic strength effects on anion adsorption at hydrous oxide/solution interfaces. J. Colloid Interface Sci., 125:717726. He, L.M., L. W. Zelazny, V. C. Baligar, K. D. Ritchey and D. C. Martens (1997) Ionic strength effects on sulfate and phosphate adsorption on gamma alumina and kaolinite: Triple layer model. Soil Science Society of America Journal, 61( 3), 784793. Hiemstra, T., de wit, J. C. M., and van Riemsdijk, W. H. (1989a) M ultisite proton adsorption modeling at the solid/solution interface of (hydr)oxides: A new approach II. application to various important (hydr)oxides. Journal of Colloid and Interface Science, 133(1), 105116. Jeon, C. S., K. Baek, J. K. Park, Y. K. Oh an d S. D. Lee (2009) Adsorption Characteristics of Iron coated Zeolite. J. Hazard. Mater., 163(23), 804 808.

PAGE 124

110 Jones, G. W. and T. Pichler (2007) Relationship between pyrite stability and arsenic mobility during aquifer storage and recovery in southwest central Florida. Environmental Science and Technology, 41(3), 723730. Kanel, S. R., B. Manning, L. Charlet and H. Choi (2005) Removal of arsenic(III) from groundwater by nanoscale zero valent iron. Environmental Science and Technology, 39(5), 1291 1298. Ka rgi, F. and M. Y. Pamukoglu (2003) Simultaneous adsorption and biological treatment of pre treated landfill leachate by fed batch operation. Process Biochem., 38(10), 14131420. Kjeldsen, P., M. A. Barlaz, A. P. Rooker, A. Baun, A. Ledin and T. H. Christe nsen (2002) Present and longterm composition of MSW landfill leachate: A review. Crit. Rev. Environ. Sci. Technol., 32(4), 297 336. Kjeldsen, P., P. L. Bjerg, K. Rugge, T. H. Christensen and J. K. Pedersen (1998) Characterization of an old municipal landfill (Grindsted, Denmark) as a groundwater pollution source: landfill hydrology and leachate migration. Waste Manage. Res., 16(1), 1422. Kjeldsen, P. and M. Christophersen (2001) Composition of leachate from old landfills in Denmark. Waste Manage. Res., 19(3), 249256. Kuleyin, A. and O. N. Ergun (2007) Studies on removal of ammonium ions from synthetic aqueous solutions and field leachate samples using c linoptilolite. Fresenius Environmental Bulletin, 16(2), 168175. Kurttio, P., E. Pukkala, H. Kahelin, A. Auvinen and J. Pekkanen (1999) Arsenic Concentrations in Well Water and Risk of Bladder and Kidney Cancer in Finland. Environmental Health Perspecti ves 107(9), 710. LaBolle, E. M. and G. E. Fogg (2001) Role of Molecular Diffusion in Contaminant Migration and Recovery in an Alluvial Aquifer System. Transport in Porous Media, 42(12), 155 179. Lamm, S. H. and M. B. Kruse (2005) Arsenic ingestion and bladder cancer mortality What do the dose response relationships suggest about? Human and Ecological Risk Assessment, 11(2), 433450. Letavayova, L., V. Vlckova and J. Brozmanova (2006) Selenium: From cancer prevention to DNA damage. Toxicology, 227(12) 1 14. Lewis, D. R., J. W. Southwick, R. Ouellet Hellstrom, J. Rench and R. L. Calderon (1999) Drinking water arsenic in Utah: A cohort mortality study. Environ. Health Perspect., 107(5), 359365.

PAGE 125

111 Maji, S. K., A. Pal and T. Pal (2007) Arsenic removal from aqueous solutions by adsorption on laterite soil. J. Environ. Sci. Health Part A Toxic/Hazard. Subst. Environ. Eng., 42(4), 453462. Martinez, M., J. Gimenez, J. de Pablo, M. Rovira and L. Duro (2006) Sorption of selenium(IV) and selenium(VI) onto magne tite. Applied Surface Science, 252(10), 37673773. McBride, M. B. (1997) A critique of diffuse double layer models applied to colloid and surface chemistry. Clays and Clay Minerals, 45(4), 598608. Mead, M. N. (2005) Arsenic: In Search of an Antidote to a Global Poison. Environ. Health Perspect., 113(6), A378 A386. Moghaddam, A. H. and C. N. Mulligan (2008) Leaching of heavy metals from chromated copper arsenate (CCA) treated wood after disposal Waste Manage., 28(3), 628637. Mor, S. (2006) Leachate characterization and assessment of groundwater pollution near municipal solid waste landfill site. Environ. Monit. Assess., 118(13), 435456. Naeem, A., P. Westerhoff and S. Mustafa (2007) Vanadium removal by metal (hydr)oxide adsorbents. Water Research, 41(7), 15961602. Navas Acien, A., E. K. Silbergeld and E. Guallar (2008) Arsenic Exposure and Prevalence of Type 2 Diabetes in US Adults. Epidemiology, 19(6), S308S308. Oliveira, A. F., A. C. Q. Ladeira, V. S. T. Ciminelli, T. Heine and H. A. Duarte (2006) Structural model of arsenic(III) adsorbed on gibbsite based on DFT calculations. Journal of Molecular Structure Theochem, 762(13), 1723. Papini, M. P., Kahie, Y. D ., Troia, B., Majone, M. (1999) Adsorption of lead at variable pH onto a natural porous medium: Modeling of batch and column experiments. E nvironmental Science and T echnology 33(24), 44574464 Payne, K. B. and T. M. Abel Fattah (2005) Adsorption of arsen ate and arsenite by iron treated activated carbon and zeolites: Effects of pH, temperature, and ionic strength. Journal of Environmental Science and Health Part A Toxic/Hazardous Substances and Environmental Engineering, 40(4), 723749. Pohland, F. G. and S. R. Harper (1989) Treatment of Landfill Leachate with an Upflow Anaerobic Reactor Combining a Sludge Bed and a Filter Discussion. Water Sci. Technol., 21(12), 15431550. Pokhrel, D. and T. Viraraghavan (2004) Leachate generation and treatment A rev iew. Fresenius Environ. Bull., 13(3B), 223 232.

PAGE 126

112 Poulsen, T. G., P. Moldrup, K. Sorensen and J. A. Hansen (2002) Linking landfill hydrology and leachate chemical composition at a controlled municipal landfill (Kastrup, Denmark) using state space analysis. W aste Manage. Res., 20(5), 445456. Pujari, P. R. and V. Deshpande (2005) Source apportionment of groundwater pollution around Landfill site in Nagpur, India. Environ. Monit. Assess., 111(13), 4354. Rahman, M., M. Vahter, N. Sohel, M. Yunus, M. A. Wahed, P. K. Streatfield, E. C. Ekstrom and L. A. Persson (2006) Arsenic exposure and age and sex specific risk for skin lesions: A populationbased casereferent study in Bangladesh. Environ. Health Perspect., 114(12), 18471852. Reitzel, S., G. Faquhar and E. McBean (1992) Temporal Characterization Of Municipal Solid Waste Leachate. Can. J. Civ. Eng., 19(4), 668679. Renou, S., J. G. Givaudan, S. Poulain, F. Dirassouyan and P. Moulin (2008) Landfill leachate treatment: Review and opportunity. J. Hazard. Mat er., 150(3), 468 493. Rigby, S. P. (2005) Predicting surface diffusivities of molecules from equilibrium adsorption isotherms. Colloid Surf. A Physicochem. Eng. Asp., 262(13), 139149. Rittmann, B. E. and P. L. McCarty (2001) Environmental Biotechnology : Principles and Applications. McGraw Hill, New York. Rossman, T. G., A. N. Uddin and F. J. Burns (2004) Evidence that arsenite acts as a cocarcinogen in skin cancer. Toxicology And Applied Pharmacology, 198(3), 394404. Salem, Z., K. Hamouri, R. Djemaa and K. Allia (2008) Evaluation of landfill leachate pollution and treatment. Desalination, 220(13), 108114. Sari, A. and M. Tuzen (2009) Biosorption of As(III) and As(V) from aqueous solution by macrofungus (Inonotus hispidus) biomass: Equilibrium and kinetic studies. J. Hazard. Mater., 164(23), 13721378. Savarimuthu, X., M. M. Hira Smith, Y. Yuan O. S. von Ehrenstein, S. Das, N. Ghosh, D. N. G. Mazumder and A. H. Smith (2006) Seasonal variation of arsenic concentrations in tubewells in West Bengal, India. Journal of Health Population and Nutrition, 24(3), 277 281. Schindler, S., H. Elias and H. Paulus (1990) Kinetic s a nd Mechanism of Ligand Substitution i n Binuclear Copper(Ii) Complexes As Model Compounds f or Hemocyanin. Z.Naturforsch.(B), 45(5), 607618.

PAGE 127

113 Sierra Alvarez, R., J. A. Field, I. Cortinas, G. Feijoo, M. T. Moreira, M. Kopplin and A. J. Gandolfi (2005) Anaerobic microbial mobilization and biotransformation of arsenate adsorbed onto activated alumina. Water Research, 39(1), 199209. Simeonova, P. P. and M. I. Luster (2004) Arsenic and atherosclerosis. Toxicol. Appl. Pharmacol., 198(3), 444449. Smedley, P L., and D. G. Kinniburgh (2002) A revi ew of the source, behaviour and distribution of arsenic in natural waters. Applied Geochemistry, 17(5), 517568. Smith, A., C. Steinmaus, J. Liaw, G. Marshall, Y. Yuan, C. Ferreccio and O. Von Ehrenstein (2006) Early childhood carcinogenic exposures causing lung cancer in young adults. Epidemiology, 17(6), S316S317. Smith, E. and R. Naidu (2009) Chemistry of inorganic arsenic in soils: kinetics of arsenic adsorptiondesorption. Environ. Geochem. Health, 31, 4959. Smith, E. H. (1998) Surface complexation modeling of metal removal by recycled iron sorbent. Journal of Environmental EngineeringA SCE 124(10), 913920. Sing, K. S. W., D. H. Everett, R. A. W. Haul, L. Moscou, R. A. Pierotti, J. Rouqurol, and T. Siemieniewska ( 1985) Reporting physisorption data for gas/solid systems with special reference to the determination of surface area and porosity. International Union of Pure and Applied Chemistry. Pure and Appl. Chem. 57, 603619. Sperlich, A., A. Werner, A. Genz, G. Amy, E. Worch and M. Jekel (2005) Breakthrough behavior of granular ferric hydroxide (GFH) fixed bed adsorption filters: modeling and experimental approaches. Water Research, 39(6), 1190 1198. Sposito G (1989) The chemistry of soils. Oxford Univ. Press, New York. Statom, R. A., G. D. Thyne and J. E. McCray (2004) Temporal changes in leachate chemistry of a municipal solid waste landfill cell in Florida, USA. Environ. Geol., 45(7), 982991. Steinmaus, C. M., Y. Yuan and A. H. Smith (2005) The temporal stability of arsenic concentrations in well water in W estern Nevada. Environ. Res., 99(2), 164 168. Sugden, K. D., K. M. Rigby and B. D. Martin (2004) Oxidative activation of the human carcinogen chromate by arsenite: A model for synergistic metal activation leading to oxidative DNA damage. Toxicology in Vitro, 18, 741748. Thirunavukkarasu, O. S., T. Viraraghavan and K. S. Subramanian (2003a) Arsenic removal from drinking water using granular ferric hydr oxide. Water SA, 29(2), 161170.

PAGE 128

114 Thirunavukkarasu, O. S., T. Viraraghavan and K. S. Subramanian (2003b) Arsenic removal from drinking water using iron oxide coated sand. Water Air and Soil Pollution, 142(14), 95111. Tseng, W. P. (1989) Blackfoot Disease In Taiwan A 30 Year Follow Up Study. Angiology, 40(6), 547558. Tuzen, M., A. Sari, D. Mendil, O. D. Uluozlu, M. Soylak and M. Dogan (2009) Characterization of biosorption process of As(III) on green algae Ulothrix cylindricum. J. Hazard. Mater., 165(13), 566 572. Vaughan, R. L., B. E. Reed and E. H. Smith (2007) Modeling As(V) removal in iron oxide impregnated activated carbon columns. J. Environ. Eng.ASCE, 133(1), 121124. Villalobos, M and J. O. Leckie ( 2000 ) Carbonate adsorption on goethite under closed and open CO2 conditions. Geochimica et Cosmochimica Acta, 64(22): 37873802. Villalobos, M., M. A. Trotz and J. O. Leckie (2003) Variability in goethite surface site density: evidence from proton and carbonate sorption. Journal of Colloid and Interface Science, 268(2), 273287. Wasserman (2004) Water arsenic exposure and children's intellectual function in Araihazar, Bangladesh (vol 112, pg 329, 2004). Environ. Health Perspect., 112(17), A980A980. Weber, W. J., Y. C. Jang, T. G. Townsend and S. Laux (2002) Leachate from land disposed residential construction waste. Journal of Environmental EngineeringA SCE 128(3), 237245. Wijnja, H and C. P. Schulthess (2000) Interaction of carbonate and organic anion s with sulfate and selenate adsorption on an aluminum oxide. Soil Science Society of America Journal, 64(3), 898908. Wiszniowski, J., D. Robert, J. Surmacz Gorska, K. Miksch and J. V. Weber (2006) Landfill leachate treatment methods: A review. Environmental Chemistry Letters, 4(1), 5161. Xiao, Y., X. M. Bai, Z. Y. Ouyang, H. Zheng and F. F. Xing (2007) The composition, trend and impact of urban solid waste in Beijing. Environ. Monit. Assess., 135(13), 2130. Xu, Y., L. Axe, N. Yee and J. A. Dyer (2006) Bidentate complexation modeling of heavy metal adsorption and competition on goethite. Environmental Science and Technology, 40(7), 2213 2218.

PAGE 129

115 Zhang, C. J. and P. J. D. Lindan (2003) Multilayer water adsorption on rutile TiO2(110): A first pri nciples study. Journal of Chemical Physics, 118(10), 46204630. Zhang, G. S., J. H. Qu, H. J. Liu, R. P. Liu and R. C. Wu (2007) Preparation and evaluation of a novel Fe Mn binary oxide adsorbent for effective arsenite removal. Water Research, 41(9), 19211928. Zhang, H. and H. M. Selim (2005) Kinetics of arsenate adsorptiondesorption in soils. Environmental Science and Technology, 39(16), 61016108. Zhang P. and D. L. Sparks (1990) Kinetics and mechanisms of sulfate adsorption/desorption on goethite using pressure jump relaxation. Soil Sci. Soc. Am. J. 54:12661273 b.

PAGE 130

116 Appendices

PAGE 131

117 Appendix A: Mercury P orosimetry R esults Table A .1: C umulative pore area and pore size distribution C onducted by Micromeritics Instrument Corporation, on 2/27/2006 using mercury intrusion porosimeter. dp ( m) Cumulative pore area (m 2 dp ( m) /g) Cumulative pore area (m 2 /g) 327.6878 227.54985 173.5191125 89.2352375 59.79998125 44.96617813 32.78413125 25.78112344 21.25665156 17.21953594 13.91776875 11.32151797 9.037774219 7.860540625 7.232928906 6.033024219 4.900156641 3.881474219 3.209735352 2.51315957 2.068713281 1.602570605 1.317486621 1.056382031 0.839917383 0.672953271 0.555413916 0.432398877 0.350161328 0.284233057 0.20 59545776 0.226553027 0.183068579 0.150857471 0.139377173 0.129389758 0 0.000391807 0.001992231 0.007810588 0.008456222 0.008889776 0.009327926 0.009644276 0.009885686 0.010087615 0.010356329 0.010593125 0.010916039 0.011092873 0.011251267 0.011566636 0.011741296 0.011960343 0.012347241 0.012966263 0.01343941 0.014325585 0.015491906 0.017722609 0.019865477 0.024533082 0.031453006 0.041619271 0.055746056 0.077651411 0.092751451 0.116956413 0.169815227 0.238609686 0.271915466 0.309654266 0.082366656 0.077124146 0.072485175 0.068383612 0.067093756 0.063579523 0.060412372 0.055784235 0.051826349 0.048360913 0.045349207 0.042679626 0.040343109 0.038285596 0.036297885 0.034231061 0.033016724 0.031560455 0.030240915 0.029024652 0.027905322 0.026871759 0.025912503 0.02417337 0.022654832 0.021322115 0.020129604 0.019490553 0.018878058 0.018030894 0.009816467 0.009632477 0.009430498 0.009144685 0.008918386 0.008702028 0.613288164 0.68414408 0.735545158 0.79736352 0.819862127 0.873644352 0.936363518 1.026214719 1.116279244 1.214369059 1.288767934 1.383346796 1.459046245 1.538657069 1.643633485 1.739577532 1.801502585 1.878663898 1.976126671 2.076627731 2.136651039 2.226849556 2.332393408 2.496361017 2.63280344 2.84517765 3.033255816 3.147248745 3.262808561 3.454088926 7.680464745 7.867991924 8.088423729 8.420746803 8.653759956 8.925909042

PAGE 132

118 Appendix A (continued) Table A .1 (continued). dp ( m) Cumulative pore area (m 2 dp ( m) /g) Cumulative pore area (m 2 /g) 0.120856372 0.017255145 0.016468413 0.015751926 0.015093037 0.014374319 0.013828618 0.013268764 0.012940515 0.012631678 0.012408337 0.012073802 0.011721764 0.011463208 0.011179498 0.010876588 0.010655597 0.010440336 0.010231044 0.010004567 0.349265516 3.611133575 3.831953287 4.065076351 4.287619591 4.519478798 4.726214409 5.01078701 5.163272858 5.361508369 5.524181843 5.759036064 5.903332233 6.146080494 6.391561508 6.620385647 6.842644215 6.976864815 7.225616455 7.451769352 0.008537014 0.008359469 0.008207018 0.007989118 0.007799093 0.007618278 0.007507674 0.007339589 0.007222354 0.007108595 0.006985001 0.006839588 0.006712788 0.006602528 0.006507509 0.006403679 0.00623812 0.006132047 0.006029965 0.00594064 8.925909042 9.178256989 9.337397575 9.601500511 9.928987503 10.29830074 10.44942856 10.6950779 11.24076939 11.42803192 11.67472553 11.89489937 12.23061752 12.53014565 12.70552349 13.03821087 13.3563509 13.50846195 13.89335632 13.97526073 Table A .2: Cumulative pore volume and pore size distribution. Conducted by Micromeritics Instrument Corporation, on 2/27/2006 using mercury intrusion porosimeter. dp ( m) C umulative pore volume (m L /g) dp ( m) C umulative pore volume (m L /g) 327.6878 227.54985 173.5191125 89.2352375 59.79998125 44.96617813 32.78413125 25.78112344 21.25665156 17.21953594 3.85505E 30 0.027193252 0.107428282 0.298528105 0.310555875 0.316233605 0.32049188 0.322807789 0.324227214 0.325198412 0.003015605 0.003119041 0.003230148 0.003318718 0.003412499 0.00365387 0.003767985 0.003889517 0.004018863 0.004109827 0.418612689 0.416973919 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722

PAGE 133

119 Appendix A (continued) Table A .2 (continued). dp Cumulative pore volume (mL/g) dp Cumulative pore volume (mL/g) 0.015093037 0.014374319 0.013828618 0.013268764 0.012940515 0.012631678 0.012408337 0.012073802 0.011721764 0.011463208 0.011179498 0.010876588 0.010655597 0.010440336 0.010231044 0.010004567 0.009816467 0.009632477 0.009430498 0.009144685 0.008918386 0.008702028 0.008537014 0.008359469 0.008207018 0.007989118 0.007799093 0.007618278 0.007507674 0.007339589 0.007222354 0.007108595 0.006985001 0.006839588 0.006712788 0.006602528 0.006507509 0.385733902 0.386587948 0.387316763 0.38828066 0.388780236 0.389413893 0.389923066 0.390641779 0.391070992 0.391774505 0.392469287 0.393100172 0.393698394 0.394052327 0.394695073 0.395267129 0.39583376 0.396289647 0.396814913 0.397586524 0.398112655 0.398712069 0.399255842 0.399591953 0.400138855 0.400801867 0.401530713 0.401821971 0.40228644 0.402707458 0.403287828 0.403623283 0.40405789 0.404438376 0.405007094 0.405505627 0.405793041 0.017968045 0.018821135 0.019417583 0.02006114 0.021246402 0.022551984 0.024061369 0.025809195 0.026721527 0.027748093 0.028854703 0.030049753 0.031345477 0.032796786 0.03398844 0.036058444 0.038059836 0.04012941 0.042477521 0.045199429 0.04812095 0.051586279 0.055623175 0.060252893 0.063355811 0.066916986 0.068212219 0.072270886 0.076850433 0.082242633 0.088143085 0.095059583 0.106266272 0.112904944 0.120423975 0.129109424 0.139088293 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415266722 0.415136099 0.414662212 0.41402784 0.413348764 0.412734091 0.412186325 0.411470085 0.410816699 0.410134822

PAGE 134

120 Appendix A (continued) Table A.2 (continued). 13.91776875 11.32151797 9.037774219 7.860540625 7.232928906 6.033024219 4.900156641 3.881474219 0.113243518 0.106497888 0.095366675 0.088414441 0.082366656 0.077124146 0.072485175 0.068383612 0.067093756 0.063579523 0.060412372 0.055784235 0.051826349 0.048360913 0.045349207 0.042679626 0.326244295 11.32151797 9.037774219 7.860540625 7.232928906 6.033024219 4.900156641 3.881474219 0.113243518 0.106497888 0.095366675 0.088414441 0.082366656 0.077124146 0.072485175 0.068383612 0.067093756 0.063579523 0.060412372 0.055784235 0.051826349 0.048360913 0.045349207 0.042679626 0.004170942 11.32151797 9.037774219 7.860540625 7.232928906 6.033024219 4.900156641 3.881474219 0.113243518 0.106497888 0.095366675 0.088414441 0.082366656 0.077124146 0.072485175 0.068383612 0.067093756 0.063579523 0.060412372 0.055784235 0.051826349 0.048360913 0.045349207 0.042679626 0.415266722 11.32151797 9.037774219 7.860540625 7.232928906 6.033024219 4.900156641 3.881474219 0.113243518 0.106497888 0.095366675 0.088414441 0.082366656 0.077124146 0.072485175 0.068383612 0.067093756 0.063579523 0.060412372 0.055784235 0.051826349 0.048360913 0.045349207 0.042679626

PAGE 135

121 Appendix B: N2 TriStar 3000 V6.07 A Unit 1 Port 1 Serial #: 2059 Started: 9/13/2007 4:33:24PM. Analysis Adsorptive: N2 Completed: 9/13/2007 7:11:10PM. Analysis Bath Temp.: 77.300 K Report Time: 9/14/2007 10:19:13AM. Sample Mass: 1.6713 g Warm Free Space: 6.5257 cm Measured. Cold Free Sp ace: 20.4456 cm Measured Equilibration Interval: 10 s Low. Pressure Dose: None Sample Density: 1.000 g/cm. Automatic Degas: Yes Stage Soak Temperature (C) Ramp Rate (C/min) Soak Time (min) 1 80 10 180 (g) P orosimetry D ata for G rain S ize Table B .1: BET surface area input report ( Table B .2: Relative pressure i sotherm t abular r eport ( Relative Pressure (P/Po) Absolute Pressure (mmHg) Quantity Adsorbed (cm/g STP) Elapsed Time (h:min) Saturation Pressure (mmHg) 01:04 737.62726 0.048377386 35.68448 7.5086 01:24 0.073976302 54.56694 8.1308 01:32 0.102309658 75.46639 8.7114 01:39 0.123118993 90.81593 9.1012 01:45 0.147660233 108.91821 9.5345 01:51 0.172691625 127.38205 9.9580 01:58 0.198041172 146.08057 10.3714 02:03 0.223502487 164.86153 10.7776 02:09 0.249290584 183.88353 11.1791 02:15 0.275267857 203.04507 11.5779 02:21 0.301460081 222.36517 11.9751 02:26 Table B .3: BET surface area output report ( BET Surface Area: 37.5978 0.1598 m 2 /g Slope: 0.114518 0.000483 g/cm 3 STP Y Intercept: 0.001266 0.000092 g/cm3 STP C: 91.479128 Qm: 8.6368 cm3/g STP Correlation coefficient: 0.9999 Molecular cross sectional area: 0.1620 nm 2

PAGE 136

122 Appendix B (continued) Table B .4: BET isotherm result ( Relative Pressure (P/Po) Quantity Adsorbed (cm/g STP) 1/[Q(Po/P 1] 0.048377386 7.5086 0.006770 0.073976302 8.1308 0.009825 0.102309658 8.7114 0.013083 0.123118993 9.1012 0.015427 0.147660233 9.5345 0.018170 0.172691625 9.9580 0.020962 0.198041172 10.3714 0.023810 0.223502487 10.7776 0.026707 0.249290584 11.1791 0.029705 0.275267857 11.5779 0.032806 0.301460081 11.9751 0.036038 Table B .5: Cumulative pore volume result ( 500 600 m grain size) Relative Pressure (P/Po) Quantity Adsorbed (cm/g STP) 0.0566 39.84724 0.0981 44.87312 0.1477 49.89687 0.1969 54.37599 0.2467 58.75347 0.29 63.08098 0.3999 71.83011 0.4959 79.75732 0.6 88.20927 0.6976 97.22081 0.7969 109.2003 0.8952 126.6693 0.9931 155.1257 0.9 141.6986 0.8001 127.7297 0.70299 110.2745 0.59716 96.34435 0.491606 85.88597 0.395091 73.95985 0.292459 64.61581

PAGE 137

123 Appendix C : Non L inear R egression A nalysis of I sotherm D ata Table C .1: Nonlinear regression fit to Langmuir isotherm model at pH 9 Conditions: As(V) species; 38 m Kemiron particle size in binary systems of 0.001 N NaNO3; CO2 Aqueous As(V)(mg/L) absent, and at room temperature. K L = 0. 39 q max = 6 7 .9 5 q(mg/g) Predicted (q mg/g) R esidual error 0.800 18.00 18.6574515 1.6939 5.30505 46.2556 47.9036208 0.2681 14.46464 54.18855 58.1544328 3.6347 25.01477 59.7002 61.3619097 2.0021 31.86075 64.21011 62.3754093 1.2638 43.2572 67.5292 63.3832014 3.3370 Table C .2: Nonlinear regression fit to Langmuir isotherm model at pH 8 Condition: As(V) species; 38 m Kemiron particle size in binary systems of 0.001 N NaNO3; CO2 Aqueous As(V)(mg/L) absent, and at room temperature. K L = 0.3 1 q max = 8 1 77 q(mg/g) Predicted (q mg/g) R esidual error 0.702 15 14.84385 0. 5789 4.86068 49.6552657 48.94122 0.8215 9.693 58 60.66914 3.1047 13.39593 65.9746503 65.00001 0.2801 22.3973375 74 70.27777 2.6687 42.10535 76.6098981 74.49094 0.7443 53.181067 75.5904437 75.56313 1.4340 62.23 77.5062344 76.16918 0.1742

PAGE 138

124 Appendix C (continued) Table C .3: Nonlinear regression fit to Langmuir isotherm model at pH 7. Condition: As(V) species; 38 m Kemiron particle size in binary systems of 0.001 N NaNO3; CO2 Aqueous As(V)(mg/L) absent, and at room temperature. K L = 0. 34 q max = 87 25 q(mg/g) Predicted (q mg/g) R esidual error 0.823456 20 19.215473 1.0650 4.969 53.1 54.734978 1.5032 12.603485 72.40837 70.390978 1.8021 24.44568 74.7 77.368467 3.0952 30.2038 82.2 78.956456 2.7638 41.91872 80.934201 80.927202 0.5411 Table C .4: Nonlinear regression fit to Freundlich isotherm model at pH 9. Condition: As(III) species; 38 m Kemiron particle size in binary systems of 0.001 N NaNO3; CO2 Aqueous As(III)(mg/L) absent, and at room temperature. K f = 42.54 1/n = 0.2 8 q(mg/g) Predicted (q mg/g) R esidual error 0.009200967 9.751959 10.82010158 1.7253 0.60808 40.79814 36.04042262 3.7819 3.641226667 61.97635 60.24887908 0.9425 11.59478 81.80263 84.01648729 2.5532 19.987175 94.12435 98.23369582 4.0937 30.461075 109.7598 110.8659051 0.7297 39.268075 123.0015 119.2514106 4.3870

PAGE 139

125 Appendix C (continued) Table C .5: Nonlinear regression fit to Freundlich isotherm model at pH 7. of As(III) Condition: 38 m Kemiron particle size in binary systems of 0.001 N NaNO3; CO2 Aqueous As(III)(mg/L) absent, and at room temperature. K f = 35.67, 1/n = 0.35 q(mg/g) Predicted (q mg/g) R esidual error 0.023381 9.612393455 8.892546334 1.6022 0.997515 37.18053878 32.02433913 1.3971 4.17858 56.73898635 52.22156098 1.0211 11.49344 74 73.76582615 2.1786 29.95287 103.4 102.2966378 0.9700 38.1549 110.185385 111.1075207 0.2022 Table C .6: Nonlinear regression fit to Freundlich isotherm model at pH 6. Condition: As(III) species; 38 m Kemiron particle size in binary systems of 0.001 N NaNO3; CO2 Aqueous As(III)(mg/L) absent, and at room temperature. K f = 40.10, 1/n = 0.29 q(mg/g) Predicted (q mg/g) R esidual error 0.0574205 9.277358 10.15114251 1.8812 1.3029017 34.34369 30.81896051 2.1142 4.631712 52.3225 48.3904839 2.7322 13.204807 66.13327 70.24409421 4.6577 21.22822 83.26322 83.16737564 0.0816 30.493129 92.84054 94.60276874 1.2326 39.291775 105.0341 103.5305812 2.4992

PAGE 140

126 Appendix D : Output of G eochemical I mpact of HSTemperature = 25C Pressure = 1.013 bars pH = 5.0 log f on L eachate S olution Generated with Geochemical Workbench. Table D .1: Summary of input and output data O2 = 61.752 Eh = 0.0200 volts pe = 0.3381 Ionic strength = 4036200.159965 Activity of water = 0.997976 Solvent mass = 1.000000 kg Solution mass = 193877.440114 kg Solution density = 1.014 g/cm3 Chlorinity = 0.059797 molal Dissolved solids = 999995 mg/kg sol'n Rock mass = 0.000000 kg Carbonate alkalinity = 0.00 mg/kg as CaCO 3 No minerals in system. Table D .2: Species with respective concentrations generated Aqueous species Molality mg/kg sol'n act. coef. log act. -------------------------------------------------------------------------------------------SO 4 -2.018*106 9.998*105 0.1111 5.3506 HSO4 3.090*102 1.547*102 0.7139 2.3436 CO2 (aq) 1.088*101 2.469*102 1.0000 0.9635 Cl5.980*102 1.093*102 0.6267 1.4263 NH4SO4 4.097*102 2.411*102 0.7139 1.5339 NaSO4 3.987*102 2.448*102 0.7139 1.5457 CaSO4 2.994*102 2.102*102 1.0000 1.5237 MgSO4 1.934*102 1.201*102 1.0000 1.7136 HCO3 6.229*103 1.960*103 0.7502 2.3304 FeSO4 9.356*104 7.331*104 1.0000 3.0289 As(OH)3 6.671*105 4.333*105 1.0000 4.1758 HSe3.761*105 1.551*105 0.7139 4.5710 N2 (aq) 3.188*10 5 4.606*106 1.0000 4.4965 H+ 1.050*105 5.458*108 0.9524 5.0000 NiSO4 2.896*106 2.311*106 1.0000 5.5383 H2SO4 2.180*106 1.103*106 1.0000 5.6616 H2Se 1.770*106 7.391*107 1.0000 5.7521 H2S(aq) 1.814*107 3.188*108 1.0000 6.7414 CO3 -1.563*107 4.838*108 0.1354 7.6743 Fe(SO4)2 9.527*108 1.218*107 0.7139 7.1674 Na+ 3.600*108 4.268*109 0.7139 7.5901

PAGE 141

127 Appendix D (continued) Table D .2 (continued). NH 4 + 2.613*10 8 2.431*10 9 0.5738 7.8241 As(OH) 4 5.457*109 4.024*109 0.7139 8.4094 Ca++ 3.037*109 6.278*1010 0.2106 9.1942 HS3.024*109 5.158*1010 0.6731 8.6914 Mg++ 1.664*109 2.086*1010 0.3071 9.2917 OH1.529*109 1.341*1010 0.6731 8.9877 AsS2 1.456*109 1.044*109 0.7139 8.9833 CaCl+ 1.682*1010 6.554*1011 0.7139 9.9205 NaHCO3 1.616*1010 7.003*1011 1.0000 9.7915 Fe++ 1.252*1010 3.605*1011 0.2106 10.5792 H2AsO4 8.932*1010 6.493*1011 0.7139 10.1954 CaHCO3 + 6.270*1011 3.269*1011 0.7945 10.3026 HAsS2 5.227*1011 3.775*1011 1.0000 10.2818 MgCl+ 3.791*1011 1.168*1011 0.7139 10.5676 MgHCO3 + 3.423*1011 1.506*1011 0.7139 10.6120 NaCl 2.422*1011 7.302*1012 1.0000 10.6157 HAsO4 -1.002*1011 7.232*1012 0.1111 11.9535 FeHCO3 + 3.438*1012 2.073*1012 0.7139 11.6100 FeCl+ 3.316*1012 1.562*1012 0.7139 11.6257 NH3 7.915*1013 6.953*1014 1.0000 12.1015 Ni++ 4.751*1013 1.439*1013 0.2106 12.9998 HCl 2.977*1013 5.598*1014 1.0000 12.5263 H3AsO4 1.135*1013 8.308*1014 1.0000 12.9451 FeCl2 4.656*1014 3.044*1014 1.0000 13.3320 AsO2OH-3.416*1014 2.184*1014 0.1111 14.4208 Se-2.755*1014 1.122*10 14 0.1111 14.5143 FeSO4 + 2.262*1014 1.772*1014 0.7139 13.7919 CaCO3 2.225*1014 1.149*1014 1.0000 13.6526 MgCO3 8.940*1015 3.888*1015 1.0000 14.0487 H2(aq) 8.441*1015 8.777*1017 1.9293 13.7882 FeCO3 2.672*1015 1.597*1015 1.0000 14.5732 NaCO3 2.434*1015 1.042*1015 0.7139 14.7601 S4 -1.059*1015 7.002*1016 0.1111 15.9296 FeOH+ 2.439*1016 9.167*1017 0.7139 15.7590 S5 -1.694*1016 1.401*1016 0.1111 16.7253 MgOH+ 1.156*1016 2.463*1017 0.7139 16.0833 AsO4 --5.646*1017 4.046*1017 0.0050 18.5493 SeO3 -4.986*1017 3.265*1017 0.0004 19.7071 S 2 -2.735*1017 9.046*1018 0.1111 17.5174 CaOH+ 1.830*1017 5.390*1018 0.7139 16.8838 NaOH 1.659*1017 3.423*1018 1.0000 16.7801 S-1.586*1017 2.623*1018 0.1603 17.5946 S 6 -1.306*1017 1.296*1017 0.1111 17.8384

PAGE 142

128 Appendix D (continued) Table D .2 (continued). S 3 -9.001*10 18 4.466*10 18 0.1111 18.0000 HSeO 3 5.465*1018 3.607*1018 0.7139 17.4088 NiOH+ 3.018*1018 1.179*1018 0.7139 17.6666 FeHSO4 ++ 3.934*1018 3.103*1019 0.1603 19.2002 Fe(OH)2 + 1.667*1019 7.727*1020 0.7139 18.9244 FeOH++ 2.244*1020 8.431*1021 0.1603 20.4440 H2SeO3 1.506*1020 1.002*1020 1.0000 19.8221 Fe(OH)3 5.317*1021 2.931*1021 1.0000 20.2744 FeCO3 + 8.650*1021 5.169*1022 0.7139 21.2093 Mg2CO3 ++ 9.931*1023 5.564*1023 0.1603 22.7979 Fe(OH)2 9.928*1023 4.602*1023 1.0000 22.0031 Fe+++ 8.332*1023 2.400*1023 0.0669 23.2537 FeCl++ 3.934*1023 1.853*1023 0.1603 23.2001 CH4 (aq) 3.782*10 23 3.129*1024 1.9293 22.1369 Ni(OH)2 2.650*1023 1.268*1023 1.0000 22.5767 FeCl2 + 1.480*1024 9.675*1025 0.7139 23.9761 Fe(OH)4 1.751*1025 1.119*1025 0.7139 24.9030 Mg2OH+++ 2.281*1026 7.718*1027 0.0493 26.9494 FeCl3 3.954*1027 3.308*1027 1.0000 26.4030 CH3COO1.042*1027 3.172*1028 0.7502 27.1072 HCH3COO 4.458*1028 1.381*1028 1.0000 27.3509 Ni(OH)3 1.518*1029 8.590*1030 0.7139 28.9652 FeCl4 2.497*1030 2.546*1030 0.7139 29.7489 Fe(OH)3 2.188*1030 1.206*1030 0.7139 29.8063 Ni2OH+++ 4.064*1031 2.818*1031 0.0493 31.6984 Ni(NH3 )2 ++ 5.302*1032 2.537*1032 0.1603 32.0705 NaCH3COO 1.319*1035 5.581*1036 1.0000 34.8797 MgCH3COO+ 1.042*1035 4.479*1036 0.7139 35.1285 FeCH3COO+ 6.588*1036 3.904*1036 0.7139 35.3276 Ni(OH)4 -9.124*1037 5.965*1037 0.1111 36.9942 SeO4 -7.892*1038 5.819*1038 0.1111 38.0572 CaCH3COO+ 4.607*1038 2.355*1038 0.7139 37.4830 Fe2(OH)2 ++++ 2.309*1038 1.735*1038 0.0151 39.4587 FeHSeO3 4.371*1040 4.144*1040 1.0000 39.3594 HSeO4 9.967*1042 7.401*1042 0.7139 41.1478 AsH3 (aq) 8.308*1042 3.340*1042 1.0000 41.0805 FeCH3COO++ 1.713*1046 1.015*1046 0.1603 46.5611 NiSeO4 4.147*1049 4.313*1049 1.0000 48.3823 Fe3(OH)4 5+ 7.115*1054 8.645*1054 0.0012 56.0640 Mg4(OH)4 ++++ 1.004*1055 8.560*1056 0.0151 56.8202 NO2 1.053*1057 2.498*1058 0.6267 57.1806 Ni4(OH)4 ++++ 2.202*1059 3.440*1059 0.0151 60.4792 HNO 2 1.097*1059 2.659*1060 1.0000 58.9599

PAGE 143

129 Appendix D (continued) Table D .2 (continued). O 2 (aq) 1.157*10 65 1.910*10 66 1.9293 64.6512 Fe(CH 3COO)2 + 1.893*1070 1.699*1070 0.7139 69.8691 NO3 1.023*1074 3.272*1075 0.6267 74.1930 Ni(NH3)6 ++ 1.369*1076 1.136*1076 0.1603 76.6585 FeNO2 ++ 3.254*1077 1.709*1077 0.1603 77.2826 CaNO3 + 4.578*1083 2.411*1083 0.7139 82.4857 NiNO3 + 1.904*1087 1.186*1087 0.7139 86.8667 Fe(CH3COO)3 1.055*1095 1.268*1095 1.0000 94.9768 FeNO3 ++ 2.228*1096 1.355*1096 0.1603 96.4470 (O phth)-2.345*10106 1.985*10106 0.1111 106.5842 H(O phth)9.318*10107 7.936*10107 0.7139 106.1770 H2(O phth) 5.924*10109 5.076*10109 1.0000 108.2274 Na(O phth)4.699*10114 4.535*10114 0.7139 113.4743 Ca(O phth) 4.377*10114 4.610*10114 1.0000 113.3588 ClO4 5.403*10147 2.771*10147 0.6731 146.4393 Ni(NO 3)2 7.354*10163 6.930*10163 1.0000 162.1335 Table D .3: The initial input species with respective concentrations. In fluid Sorbed Kd Original basis total moles moles mg/kg moles mg/kg L/kg ------------------------------------------------------------------------------------------------As(OH) 4 6.67*105 6.67*105 4.92*105 Ca++ 0.0299 0.0299 0.00619 Cl0.0598 0.0598 0.0109 Fe++ 0.000936 0.000936 0.000270 H+ 309.0 309.0 1.60 H2O 55.7 55.7 5.17 HCO3 0.115 0.115 0.0362 Mg++ 0.0193 0.0193 0.00242 NO3 0.0410 0.0410 0.0131 Na+ 0.0399 0.0399 0.00473 Ni++ 2.90*106 2.90*106 8.77*107 O2(aq) 0.197 0.197 0.0325 SO4 -2.02*106 2.02*106 1.00*106 SeO3 -3.94*105 3.94*105 2.58*10 5

PAGE 144

130 Appendix D (continued) Table D .4: Gases with respective fugacities generated. Gases fugacity log fug. ---------------------------------------------------------------------------------------------CO 2(g) 3.082 0.489 N2(g) 0.04878 1.312 Steam 0.03125 1.505 H2S(g) 1.940*106 5.712 H2(g) 2.108*1011 10.676 S2(g) 1.958*1016 15.708 CH4(g) 4.823*1020 19.317 O2(g) 1.769*1062 61.752

PAGE 145

131 Appe n dix E: Raw Experimental Data Table E. 1: As(V) sorption data on 0.1 g/L Kemiron. 5 ppm As(V), 0.001 N 5 ppm As(V), 0.1 N 5 ppm2 As(V), 0.1 N Final pH % sorbe d Final pH % sorbe d Final pH % sorbe d 9.70 57.83 9.34 67.01 8.78 64.58 9.86 60.99 8.94 66.51 7.90 73.51 8.70 71.60 8.26 75.66 7.87 71.36 7.60 82.86 7.90 80.28 7.73 74.80 7.57 84.68 8.08 76.77 7.41 78.53 7.60 82.74 7.33 85.95 7.26 78.25 7.57 83.75 7.47 86.64 7.00 81.10 7.72 83.62 6.57 92.31 6.90 83.03 7.42 6.98 91.29 6.70 84.07 5.71 94.25 4.55 98.33 6.68 84.39 6.19 94.16 4.31 98.55 6.25 86.61 6.48 93.03 5.74 89.41 5.93 92.85 5.59 87.92 10 ppm As(V), 0.001 N 10 ppm As(V), 0.1 N 10 ppm As(V), 0.1 N Final pH % sorbe d Final pH % sorbe d Final pH % sorbe d 9.50 36.70 8.71 37.27 9.81 35.97 9.68 41.07 7.11 47.19 9.92 39.20 8.68 48.01 6.89 46.17 9.38 39.58 8.46 6.65 47.50 8.73 45.57 7.96 49.53 6.63 47.21 7.58 54.42 8.85 45.89 6.02 46.92 7.76 51.28 7.49 54.49 6.01 44.68 6.69 61.73 7.78 57.33 6.00 47.04 5.29 69.71 7.83 54.44 5.27 58.82 5.88 65.96 5.46 66.80 5.22 58.18 5.91 68.10 4.32 72.02 5.20 59.44 6.11 65.16 5.00 61.38 4.48 72.07

PAGE 146

132 Appendix E (continued) Table E. 2: 5 ppm As(III) sorption data on 0.1 g/L Kemiron. 10 ppm As(III), 0.001 N sodium nitrate 10 ppm As(III), 0.1 N sodium nitrate Final pH % As(V) removed pH % As(V) removed 9.7 61.5 8.5 55.0 9.2 65.2 7.9 59.4 6.6 55.6 7.8 56.9 7.0 61.8 7.5 54.9 7.0 58.9 7.3 57.4 7.3 63.1 6.1 50.2 7.4 62.6 5.9 44.3 8.0 69.0 5.9 44.6 7.1 59.5 5.1 41.7 6.6 56.7 4.8 39.8 6.2 53.4 4.5 38.3 5.7 41.0 4.3 35.0 4.7 36.7 4.6 31.1 4.6 30.3 4.2 27.3 4.1 26.9 1 ppm As(III), 0.1 N sodium nitrate 8.9 98.4 8.7 98.9 7.7 98.2 5.9 98.4 5.8 96.1 5.8 98.3 5.8 97.5 5.6 95.1 5.1 92.4 5.1 96.8 4.7 91.1 4.5 88.4 4.3 83.5 4.1 86.6 5 ppm As(III), 0.1 N sodium nitrate 9.6 78.1 9.6 81.9 9.7 80.7 8.7 87.4 7.3 81.6

PAGE 147

133 Appendix E (continued) Table E.2 (continued). 6.9 78.4 6.0 69.1 6.3 71.8 6.6 74.3 6.8 76.4 4.0 38.2 4.2 43.8 Table E. 3: 5 ppm As(III) Isotherm data on 0.1 g/L Kemiron, I = 0.01 N NaNO3 pH Ceq(mg/l) q(mg/g) pH Ceq(mg/l) q(mg/g) 6 0.06 9.28 7 0.02 9.61 6 1.30 34.34 7 1.00 37.18 6 4.63 52.32 7 4.18 56.74 6 13.20 66.13 7 40.22 95.97 6 21.23 83.26 6 30.49 92.84 6 39.29 105.03 pH Ceq(mg/l) q(mg/g) pH Ceq(mg/l) q(mg/g) 8 0.01 9.75 9 0.01 9.75 8 0.75 39.46 9 0.61 40.80 8 11.49 82.79 9 3.64 61.98 8 29.95 98.12 9 11.59 81.80 8 38.15 116.19 9 19.99 94.12 9 30.46 104.76 9 39.27 113.00

PAGE 148

134 Appendix E (continued) Table E. 4: 5 ppm As(V ) Sorption on 0.1 g/L Bayoxide, I = 0.01 N NaNO3 0.1g/L BayOxide 5 ppm initial As(V) concentration Final % As(V) pH 5 ppm 9.3 18.2 9.6 18.5 8.9 8.8 21.9 8.2 24.0 7.0 26.4 7.0 28.5 6.7 30.0 6.8 31.2 6.3 29.1 6.6 31.8 6.4 34.6 5.4 37.6 5.3 36.3 4.7 41.1 10 ppm initial As(V) concentration Final % As(V) pH 10 ppm 8.7 3.6 8.3 3.0 7.7 5.4 7.1 9.3 6.7 9.8 6.5 6.4 12.3 6.4 11.3 6.4 14.6

PAGE 149

135 Appendix E (continued) Table E. 5: 5 ppm As(V) S orption to 0.1 g/L Kemiron in the presence of Se(IV), I = 0.001 N NaNO3 w 5 ppm Se(IV) Final % As(V) pH removed 8.97 53.12 8.93 54.51 8.14 56.90 8.28 57.56 7.74 64.11 7.67 65.76 7.19 68.86 6.72 71.11 6.41 73.55 5.05 75.32 5.73 75.57 5.84 75.08 5.05 4.65 79.73 w 0.5 ppm Se(IV) Final % As(V) pH removed 8.27 70.88 8.28 67.82 8.96 63.50 8.09 74.92 7.82 76.61 7.73 78.57 7.58 79.16 7.21 86.64 7.07 85.18 6.96 87.60 6.57 89.04 6.35 88.79 5.99 97.07

PAGE 150

136 Appendix E (continued) Table E. 6: 5 ppm As(V) S orption to 0.1 g/L Kemiron in the presence of Ca2+, I = 0.001 N NaNO3 W 0.1 ppm Ca 2+ Final % As(V) pH removed 3.81 99.99 5.38 99.95 6.09 99.98 6.55 99.99 5.29 99.96 5.92 100.00 6.34 100.00 6.28 100.00 6.91 100.00 5.96 100.00 6.28 99.65 4.83 100.00 4.33 100.00 w 0.001 ppm Ca 2+ Final % As(V) pH removed 8.95 66.98 8.05 76.42 8.35 71.93 8.06 75.49 8.08 76.15 8.14 75.15 7.54 80.95 6.81 86.49 7.43 81.90 7.21 84.43 6.84 88.42 6.43 91.06 6.55 91.19 6.34 92.66 5.85 94.45

PAGE 151

137 Appendix E (continued) Table E. 7: 5 ppm As(V) sorption to 0.1 g/L Kemiron in the presence of CO3 2 -, I = 0.001 N NaNO3 0.1 ppm CO32 Final pH % As(V) sorbed 8.27 70.1 7.52 77.9 7.42 79.7 7.37 80.6 3.59 96.6 3.91 95.6 6.14 91.2 6.52 88.8 6.54 90.8 6.88 89.0 6.89 86.8 6.91 89.6 7.13 86.7 6.32 91.5 5.75 94.0 1 ppm CO 3 2 9.27 65.3 9.28 65.8 9.16 66.8 8.42 72.1 7.73 80.3 7.53 82.3 7.55 83.0 7.23 86.8 7.33 84.6 6.83 90.1 6.93 88.1 6.73 90.4 4.09 97.3 5.65 93.7 5.21 95.9 100 ppm CO 3 2 9.99 58.0 9.91 61.0 9.42 66.8 7.99 82.1 8.14 82.6

PAGE 152

138 Appendix E (continued) Table E. 7 (continued). 0.1 ppm CO32 Final pH 8.11 % As(V) sorbed 83.1 7.88 86.0 7.49 86.6 7.31 90.2 6.96 90.2 6.73 91.4 6.33 93.5 5.96 94.8 3.82 97.7 Table E. 8: 5 ppm As(V) sorption to 0.1 g/L Kemiron in the presence of SO4 2 -, I = 0.001 N NaNO3 1 ppm SO 4 2 Final pH % As(V) sorbed 8.57 70.3 8.79 69.7 8.33 73.9 8.00 76.4 7.75 79.0 7.01 86.3 7.32 83.9 6.97 87.7 6.95 87.4 6.34 92.2 6.26 92.9 5.91 94.5 5.69 95.7 5.48 95.8 4.69 97.7 100 ppm SO 4 2 7.80 78.9 8.79 70.3 7.85 79.8

PAGE 153

139 Appendix E (continued) Table E. 8 (continued) 7.50 83.5 7.14 87.6 5.76 92.0 4.96 91.7 6.90 86.2 6.79 89.0 4.71 95.1 6.73 88.6 4.28 94.6 6.42 89.9 Table E. 9: 5 ppm As(V) or As(III) sorption to 0.1 g/L Kemiron in the presence of 5 ppm Ni. I = 0.001N NaNO3 As(V) pH %As(V) Sorbed 7.93 76.9 7.76 78.3 7.57 81.0 7.25 84.0 7.15 86.0 6.94 88.9 6.59 90.3 6.43 91.9 5.74 94.4 6.97 90.2 4.46 96.7 As(III) pH % As(III) Sorbed 8.13 79.8 7.72 76.5 7.59 74.8 7.43 74.2 7.35 73.5 6.93 71.8 6.84 68.8 5.70 56.3 7.03 71.1 7.03 72.6 7.24 74.2 7.86 82.2

PAGE 154

140 Appendix E (continued) Table E. 10: 5 ppm As(V) sorption to 0.1 g/L Kemiron ( ORP in a synthetic landfill leachate solution. Final Final % As(V) Final Final % As(V) ORP pH removed ORP pH removed 336 10.6 89 234 10.2 36 334 12.0 87 234 8.6 60 325 10.2 56 235 9.1 53 308 10.8 99 243 8.4 64 248 8.4 49 253 9.4 45 245 8.5 36 259 9.0 52 235 7.8 56 259 9.2 48 230 10.2 38 263 9.9 38 230 8.8 34 265 8.5 67 230 8.8 34 277 8.4 59 217 8.1 54 288 7.5 93 86 11.6 99 289 9.9 41 8 11.1 100 299 9.3 47 9 11.3 100 300 8.1 81 72 11.8 90 301 7.8 90 83 10.9 100 310 8.8 56 110 10.0 55 317 9.2 51 110 10.0 42 319 7.5 96 112 10.4 65 322 7.9 85 115 10.9 100 323 8.1 71 122 11.2 100 332 7.2 98 131 9.9 45 334 7.6 94 132 10.3 41 345 8.1 75 135 10.1 37 350 8.3 70 135 11.2 100 351 8.3 75 135 11.2 100 351 8.1 76 138 10.0 43 355 8.3 75 139 11.4 99 356 8.3 69 160 10.0 46 356 8.3 67 170 10.0 37 357 8.3 64 196 10.4 56 359 8.2 75 197 8.4 69 363 8.3 73 202 8.5 61 366 8.3 73 210 7.9 91 372 8.2 72 221 8.6 60 375 7.7 85 221 7.8 91 381 9.2 51 223 10.2 38 391 8.8 59 231 9.3 47 233 9.8 42

PAGE 155

141 Appendix F: Non L inear R egression of Freundlich I sotherm disp('Gauss Newton Method of nonlinear Regression of Freundlich isotherm, pH 6') disp('y=(a(1)*x^a(2))') x=[0.0574205 1.3029017 4.631712 13.204807 21.22822 30.493129 39.291775]; y=[9.277358 34.34369 52.3225 66.13327 83.26322 92.84054 105.0341]; x y n=20 a=[10 0.1]' for i=1:n disp( ') i a dfda1=x.^a(2); dfda2=(a(1).*x.^a(2)).*log(x); DFDB=[dfda1' dfda2'] D=[(y (a(1).*x.^a(2)))'] B=(inv(DFDB'*DFDB))*(DFDB'*D) a=a+B; end disp( ') a x1=(0:0.5:45); ytheo=(a(1).*x1.^a(2)); plot(x,y, '*') hold on plot(x1,ytheo, 'r' )

PAGE 156

142 Appendix G: Non L inear R egression of Langmuir I sotherm disp('Gauss Newton Method of nonlinear Regression of Langmuir isotherm') disp('y=(a(1)*x*a(2))/(1+a(1)*x)') x=[0.8 5.30505 14.46464 25.01477 31.86075 43.2572]; y=[18 46.2556 54.18855 59.7002 64.21011 67.5292]; x y n=50 a=[1 100]' for i=1:n disp( ') i a dfda1=((a(2).*x)./((1+a(1).*x)))(a(1).*x.*a(2).*x)./((1+a(1).*x)).^2; dfda2=a(1).*x./(1+a(1).*x); DFDB=[dfda1' dfda2'] D=[(y (a(1).*x.*a(2))./(1+a(1).*x))'] B=(inv(DFDB'*DFDB))*(DFDB'*D) a=a+B; end disp( ' ) a x1=(0:0.5:50); ytheo=(a(1).*x1.*a(2))./(1+a(1).*x1); plot(x,y, 'O') hold on plot(x1,ytheo, 'r' )

PAGE 157

About the Author Douglas Oti received a Bachelors Degree in Geological Engineering from the University of Science and Technology in 1998 in Ghana and a n M. S in Civil Engineering with a focus on Water Resources from North Carolina Agricultural and Technical State University in 2004. He was a T.A of various undergraduate Water Resources Courses while in the Masters program and entered the Ph.D. program at the University of South Florida in 2004. While in the Ph.D. progra m Douglas was very active in various Student Associations including FSAWWA, ASCE and Engineers for a Sustainable World. In ESW he held a position of Vice President for projects in 20072008. He has presented his work at the Florida Air & Waste Management A ssociation (A&WMA) conference as well as local college and university research symposia. He has so far published one peer reviewed article in the Journal of Environmental Science and Health, Part A: Toxic/Hazardous Substanc e and Environmental Engineering He was also involved in teaching and mentoring undergraduate students in classrooms and in the lab for Environmental and Hydraulic s courses He was always the first to be in the lab and the last to leave.