Citation
Innovative Remediation Techniques for Treatments of Toxaphene Contamination

Material Information

Title:
Innovative Remediation Techniques for Treatments of Toxaphene Contamination
Creator:
CHEN, XIAOSONG ( Author, Primary )
Copyright Date:
2008

Subjects

Subjects / Keywords:
Alcohols ( jstor )
Biodegradation ( jstor )
Environmental technology ( jstor )
Experimental data ( jstor )
Fractions ( jstor )
Kinetics ( jstor )
Pentanols ( jstor )
Solubility ( jstor )
Solvents ( jstor )
Sorption ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Xiaosong Chen. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
2/28/2006
Resource Identifier:
495636947 ( OCLC )

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

INNOVATIVE REMEDIATION TECHNIQUES FOR TREATMENTS OF TOXAPHENE CONTAMINATION By XIAOSONG CHEN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Xiaosong Chen

PAGE 3

This document is dedicated to my family.

PAGE 4

ACKNOWLEDGMENTS I wish to express my deep and sincere gratitude towards Dr. Clayton J. Clark II, my supervisor and mentor during my Ph.D study, for his invaluable guidance and constant encouragement both professionally and personally throughout the duration of this work. I would like to thank the other members of my committee, Dr. Michael Annable, Dr. Angela Lindner, Dr. Kirk Hatfield, and Dr. Roy Rhue for their ideas, guidance, support, and encouragement. I would also acknowledge Dr. Jim Jawitz, Dr. Andy Ogram, and Dr. Huaguo Wang in the Department of Soil and Water Science, and Dr. Chou in the Department of Environmental Engineering, for their invaluable help, especially with regard to the column test and biodegradation. I wish to thank Yun Cheng and Padma Paan for their laboratory assistance. Finally, I also want to thank my family and friends for their emotional support and encouragement throughout my education. iv

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES...........................................................................................................viii LIST OF FIGURES.............................................................................................................x ABSTRACT.....................................................................................................................xiii CHAPTER 1 INTRODUCTION........................................................................................................1 2 COSOLVENT FLUSHING TECHNOLOGY FOR TOXAPHENE MASS REMOVAL...................................................................................................................9 Introduction...................................................................................................................9 Solubility.....................................................................................................................10 Background..........................................................................................................10 Methodolology....................................................................................................14 Materials.......................................................................................................14 Experimental procedure...............................................................................16 Analytical methods.......................................................................................17 Results and Discussion........................................................................................21 Comparison of estimated and experimental data.........................................21 Cosolvency power........................................................................................23 Sorption.......................................................................................................................26 Background..........................................................................................................26 Methodolology....................................................................................................28 Materials.......................................................................................................28 Sorption isotherm of toxaphene to soil........................................................28 Analytical methods.......................................................................................29 Results and Discussion........................................................................................32 Column Test................................................................................................................34 Background..........................................................................................................34 Methodology........................................................................................................37 Results and Discussion........................................................................................39 Modeling results...........................................................................................39 Relationship between sorption constant and cosolvent fraction..................41 v

PAGE 6

Relationship between sorption constant and reverse first order rate constant.....................................................................................................43 Relationship between reverse first order rate constant and cosolvent fraction......................................................................................................44 Relationship between fraction of instantaneous domain and cosolvent fraction......................................................................................................46 Conclusions.................................................................................................................47 3 DECHLORINATION TECHNOLOGIES OF TOXAPHENE..................................72 Introduction.................................................................................................................72 Iron Treatment.....................................................................................................73 Biodegradation....................................................................................................74 Comparison of Dechlorination Technologies......................................................77 Zero-Valent Iron Treatment........................................................................................77 Background..........................................................................................................77 Methodology........................................................................................................82 Results and Discussion........................................................................................83 Effects of Cosolvents on Dechlorination of Toxaphene by Iron................................86 Background..........................................................................................................86 Methodology........................................................................................................87 Results and Discussion........................................................................................89 Toxaphene dechlorination............................................................................89 Sorption isotherms........................................................................................90 Cosolvency power........................................................................................91 Kinetic analysis of the toxaphene degradation process................................92 Biodegradation of Toxaphene and Cosolvent Effects................................................95 Background..........................................................................................................95 Process pathway...........................................................................................95 Process products...........................................................................................96 Structures......................................................................................................98 Product Toxicity...........................................................................................99 Effects of cosolvent on biodegradation of toxaphene................................100 Methodology......................................................................................................100 Results and Discussion......................................................................................102 Conclusions...............................................................................................................103 4 SUMMARY AND DISCUSSION...........................................................................114 Conclusions...............................................................................................................114 Recommendations.....................................................................................................116 Surfactant Flushing and Its Effects on Iron Treatment .....................................116 Effects of Cosolvent/Surfactant Solutions with F raction of Lower than 5% on Biodegradation..........................................................................................117 Combine Biodegradation Technology w ith Iron Permeable Reactive Barrier (PRB).............................................................................................................118 vi

PAGE 7

LIST OF REFERENCES.................................................................................................119 BIOGRAPHICAL SKETCH...........................................................................................126 vii

PAGE 8

LIST OF TABLES Table page 1-1 Structure and designation of chlorobornanes and chlorocamphenes.......................2 1-2 Important physical and chemical properties of toxaphene......................................3 2-1 Estimated solubilities of toxaphene in methanol, ethanol, Propanol and IPA using theoretic estimation......................................................................................13 2-2 Chemical and physical properties of chosen solvents ( refer to the UAkron Chemical fact sheet)...............................................................................................15 2-3 Percent of difference of the predictions ( logSs,estd-logSs,expt)/ logSs,expt*100 using the log-linear and extended log-linear models for binary cosolvent systems...................................................................................................................22 2-4 Comparison of cosolvency powers from experimental data and from log-linear model and UNIFAC model....................................................................................24 2-5 The cosolvency power of different types of solvent..............................................25 2-6 The solvent-sorbent interaction and Kp,w value from plot of logKp-fc................34 2-7 Parameter values for binary cosolvent system.......................................................41 2-8 Parameter values for ternary cosolvent system (50% methanol + pentanol).........41 2-9 Comparison of the Kp value from column test and batch test................................42 2-10 Predicted versus experimentally obtained regression constants for binary and ternary cosolvent system flushing..........................................................................46 3-1 Results from field scale study of toxaphene degradation......................................75 3-2 Toxaphene dechlorination by various substrates containing iron at different volume-to-mass ratios and initial concentrations..................................................84 3-3 The chosen initial concentrations and fractions of cosolvent solutions.................87 3-4 Toxaphene degradation rate of power law function kinetic by zero-valent iron at different fraction of cosolvent solution and initial concentration......................89 viii

PAGE 9

3-5 Values of Kf and Nf in Frendliuch sorption isotherm and Ca0/Sc..........................90 3-6 Kinetic analysis of toxaphene degradation and sortion.........................................92 ix

PAGE 10

LIST OF FIGURES Figure page 1-1 GC/ECD chromatograms of toxaphene standards by many other companies.........6 1-2 HRGC/ECD chromatogram of technical toxaphene. I-V are the investigated congeners; II includes two coeluting isomers (IIa and IIb).....................................7 1-3 GIS map of toxaphene pesticide samples in north shore of Lake Apopka , FL......8 2-1 Change of toxaphene concentration with time.......................................................49 2-2 Comparison of the experimental data of toxaphene solubility in methanol, ethanol, and IPA with estimation...........................................................................50 2-3 Cosolvency power from the log-linear regression of the solubility to cosolvent fractions for binary solutions (methanol-water, ethanol-water, propanol-water, IPA-water)..............................................................................................................52 2-4 Cosolvency power from the log-linear regression of the solubility to cosolvent fraction for ternary solutions (butanol-methanol-water, isobutanol-methanol-water, pentanol-methanol-water, and hexanol-methanol-water)...........................54 2-5 Comparison of the cosolvency powers of different types of alcohol.....................56 2-6 Relationship of cosolvency power with the dielectric constant.............................57 2-7 Relationship of cosolvency power with the solubility of the POMSs...................57 2-8 The change of toxaphene concentration with time for batch test..........................58 2-9 Linear regression of S-Ce for 50% methanol binary cosolvent solution for batch test................................................................................................................58 2-10 The effects of ratio of solution: soil on the sorption constant Kp..........................59 2-11 Batch test of log linear regression of partitioning coefficient Kp to cosolvent fraction in cosolvent solutions...............................................................................60 2-12 IPA tracer breakthrough curve...............................................................................63 x

PAGE 11

2-13 Breakthrough curves for toxaphene using binary cosolvent system (methanol solution).................................................................................................................64 2-14 Breakthrough curves for toxaphene using ternary cosolvent system ( 50% methanol + pentaol solution).................................................................................65 2-15 Comparison of breakthrough curves for toxaphene using binary (50% or 60% methanol) and ternary cosolvent system (50% methanol+ 10% pentaol solution).................................................................................................................66 2-16 Comparison of breakthrough curves for toxaphene using binary (50% or 75% methanol) and ternary cosolvent system (50% methanol+ 25% pentaol solution).................................................................................................................67 2-17 LogKp-fc for binary system (methanol solution)....................................................68 2-18 LogKp-fc for ternary system (50% methanol +Pentanol solution).........................68 2-19 Log k2 – logKp for binary system (methanol solution)..........................................69 2-20 Log k2 – logKp for ternary system (50% methanol +Pentanol solution)...............69 2-21 Log k2 – fc for binary system (methanol solution).................................................70 2-22 Log k2 – fc for ternary system (50% methanol +Pentanol solution)......................70 2-23 F – fc for binary system (methanol solution).........................................................71 2-24 F– fc for ternary system (50% methanol +Pentanol solution)................................71 3-1 Three possible schemes of pathways for the reductive dechlorination under anoxic Fe0-H2O systems......................................................................................105 3-2 GC-ECD chromatograms of toxaphene dechlorination by zero-valent iron with different time periods...........................................................................................106 3-3 GC-ECD chromatograms of toxaphene dechlorination by Fe0, NiFe0, CuFe0 after 1 week..........................................................................................................107 3-4 Change of Cl concentration with the time...........................................................108 3-5 Toxaphene degradation rate as functions of methanol fraction under different initial toxaphene concentration C0:......................................................................109 3-6 Linearity of sorption isotherm to the Ca/Sc, which is changed with initial concentration of toxaphene and cosolvent fraction in solution...........................110 3-7 Sorption coefficients Kd onto the iron surfaces as functions of methanol fraction in the range of 0.1 to 0.5.........................................................................110 xi

PAGE 12

3-8 Comparison of experimental data and kinetic results..........................................111 3-9 Effects of 5% fraction of cosolvent solutions (methanol, ethanol, and IPA) on biodegradation of Toxaphene..............................................................................113 3-10 Effects of different fractions of methanol solution biodegradation of Toxaphene............................................................................................................113 xii

PAGE 13

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy INNOVATIVE REMEDIATION TECHNIQUES FOR TREATMENTS OF TOXAPHENE CONTAMINATION By Xiaosong Chen August 2005 Chair: Clayton J. Clark II Cochair: Kirk Hatfield Major Department: Civil and Coastal Engineering Due to its persistence in nature and its ability to bio-accumulate, toxaphene, a once prevalently used pesticide, has been a major concern for health and environmental officials. The study of toxaphene and methods to remediate it effectively and efficiently will continue to be a major focus for environmental engineers and scientists alike. This research explored different remediation approaches relevant to toxaphene contaminants and was divided into two parts: mass removal technology of toxaphene from a contaminated site using cosolvent flushing, and dechlorination technologies of toxaphene applying zero-valent iron treatment and the effects of cosolvent solutions on it. Biodegradation technology and its combination with cosolvent flushing were also discussed for the purpose of identifying opportunities and needs. The research revealed that the addition of cosolvent solution increased the solubility of toxaphene and decreased the sorption rate of toxaphene to the soil, thus reducing the retardation factor during cosolvent flushing. The alcohols with higher molecular weight exhibited a higher value of cosolvency power for toxaphene. Addition of Partial Miscible Organic Solvents xiii

PAGE 14

(PMOSs) into Complete Miscible Organic Solvents (CMOSs) to form a ternary system could enhance the solubility and the cosolvent flushing effects. Straight-chain carbon alcohols exhibited a higher potential to enhance the solubility of toxaphene than branched chain carbon alcohols. The first-order bi-continuum transport model provided good simulation of the non-equilibrium sorption process for a miscible displacement column test with toxaphene. The application of cosolvent flushing can reduce the time required to conduct sorption experiments for highly hydrophobic organic compounds such as toxaphene. The experimental data validated the log-linear relationship between the sorption constant Kp and cosolvent fractions fc, and between the reverse first-order rate k2 and fc. A log-log linear relationship was found between the k2 value and the Kp value, and the ratio of the k2 value as a function of the fraction of cosolvent could be estimated successfully using literature data and equations. The application of zero-valent iron and bimetallic substrates can be potentially used as a passive technique in the degradation and removal of toxaphene from the environment. A power law relationship had been shown to depict the decrease of toxaphene concentration related to iron treatment as a function of time. The application of bimetallic substrates has been used to improve the dechlorination rate. A nonlinear Freundlich sorption equation applied to experimental data indicated that when the saturation of solution Ca/Sc, which was the ratio of aqueous concentration to the solubility, is higher than 0.2, the sorption isotherm was nearly linear. Using the linear partitioning coefficient, a log-linear equation indicated the cosolvency power of toxaphene in methanol/water solutions was 3.45 for toxaphene sorption to iron surface, similar to that derived from the experimental batch solubility test of toxaphene in methanol solution. Experimental results showed that the dechlorination rate was reduced with increasing methanol fraction. A preliminary experiment assesing the effects of cosolvent solutions on the biological removal of toxaphene showed that the disappearance of toxaphene was inhibited with the depletion rates reducing from 35% to less than 14% compared with a control without cosolvent presence. xiv

PAGE 15

CHAPTER 1 INTRODUCTION Toxaphene, a polychloroterpene material, was developed in 1947 (U.S. EPA, 2003) and used mainly as a pesticide. The initial process used to manufacture it began with pine tar treated to make camphene, which was then treated with chlorine. Later, the definition of “technical toxaphene” was patterned after the Hercules Incorporated product (Hercules Code Number 3956) marketed under the trademark name of “Toxaphene” in the United States. But recently, many products with similar properties have become referred to as toxaphene (Worthing and Walker, 1987). Generally, toxaphene is an amber, waxy organic solid with odor similar to turpentine (ATSDR, 1996). Toxaphene is a complex mixture comprised of at least 2,300 closely related polychlorinated compounds, mostly chlorobornanes (C10H16–nCln) and chlorocamphenes (C10H18-nCln ) where n = 5, which produces well resolved peaks and many other unresolved components as well for GC/ECD chromatograms (Figure 1-1, Gray et al., 2000). Table 1-1 (Buser et al. ,2000) lists the toxaphene compounds in the reference mixture and the corresponding designation from Buser et al. (1994), Hainzl et al. (1994), and Stern et al.(1996). The IUPAC numbering method for the carbon skeleton of the bornane system was also employed. Six pure polychlorinated bornanes have been isolated from technical toxaphene, namely, Parlar 32 (toxicant B), Parlar 42a (toxicant A1), Parlar 42b (toxicant A2), Parlar 49a, Parlar 56, and Parlar 59. These were investigated in an anaerobic loamy silt, whose enantiomeric structures and high 1

PAGE 16

2 resolution GC/ECD (HRGC/ECD) chromatograms are presented as Figure 1-2 (Fingerling et al., 1996). Table 1-1. Structure and designation of chlorobornanes and chlorocamphenes Compound designationa Previous designationb Chemical name (IUPAC) P11 2,2,3-exo,8,9,10-hexachlorocamphene P12 2-exo,3-endo,8,8,9,10-hexachlorocamphene P15 2-exo,3-endo,7,8,9,10-hexachlorocamphene P21 2,2,5,5,9,10,10-heptachlorocamphene P25 TC3 2,2,3-exo,8,8,9,10-heptachlorocamphene P26 TOX8 2-endo,3-exo,5-endo,6-exo,8,8,10,10-octachlorobornane P31 TC9 2,2,3-exo,8,8,9,9,10-octachlorobornane P32 ToxicantB 2,2,5-endo,6-exo,8,9,10-heptachlorobornane P38 2,2,5,5,9,9,10,10-octachlorobornane P39 2,2,3-exo,5-endo,6-exo,8,9,10-octachlorobornane P40 TC6 2-endo,3-exo,5-endo,6-exo,8,9,10,10-octachlorobornane P41 TC6 2-exo,3-endo,5-exo,8,9,9,10,10-octachlorobornane P4 TC8,Toxicant A 2,2,5-endo,6-exo,8,8,9,10-octachlorobornane and 2,2,5-endo,6-exo,8,9,9,10-octachlorobornane P44 TC7 2-exo,5,,8,9,9,10-octachlorobornane P50 TOX9 2-endo,3-exo,5-endo,6-exo,8,8,9,10,10-nonachlorobornane P51 2,2,5,5,8,9,10,10-octachlorobornane P56 2,2,5-endo,6-exo,8,8,9,10,10-nonachlorobornane P58 2,2,3-exo,5,5,8,9,10,10-nonachlorobornane P59 2,2,5-endo,6-exo,8,9,9,10,10-nonachlorobornane P62 2,2,5,5,8,9,9,10,10-nonachlorobornane P63 2-exo,3-endo,5-endo,6-exo,8,9,9,10,10-nonachlorobornane P69 2,2,5,5,6-exo,8,9,9,10,10-decachlorobornane HxSedc 2-exo,3-endo,6-exo,8,9,10-hexachlorobornane HpSedc 2-endo,3-exo,5-endo,6-exo,8,9,10-heptachlorobornane B8-1412d TC5 2-endo,3-exo,5-endo,6-exo,8,8,9,10-octachlorobornane a Hainzl et al. (1995); b Buser et al. (1994); c Stern et al. (1996);d Klobes et al. (1997) Toxaphene is harmful if in contact with skin, toxic if swallowed, and irritating to the respiratory system (ATSDR, 1996). It has been reported to cause possible risk of

PAGE 17

3 irreversible effects, to be very toxic to aquatic organisms, and to potentially cause long-term adverse effects in the aquatic environment (ATSDR, 1996). Toxaphene is easily metabolized once ingested by birds and mammals, but fish are highly susceptible to poisoning as denoted by the bio-concentration factor (Table 1-2, ATSDR, 1996). Table 1-2. Important physical and chemical properties of toxaphene. Property Information Reference Molecular Weight 414 (average) DOT, 1978 Melting Point 65-90C Merck, 1989 Boiling Point N/A; dechlorinates at 155C Merck, 1989 Density at 25` 1.65 g/cm3 Worthing, 1979 Solubility in water 3ppm Wauchope et al., 1992 Worthing, 1979 Log Kow 3.30 U.S. EPA, 1981 Vapor Pressure 0.2-0.4 mm Hg at 20C 0.4 mm Hg at 25C 4*10-6 mmHg at 20C 3*10-7 mmHg at 20C 5*10-6mmHg at 20C Mackinson et al., 1981 Wauchope et al., 1992 Wauchope et al.,1992 Suntio et al.,1988 Bidleman et al.,1981 Log Koc 2.474 U.S. EPA, 1981 Henry’s Law constant 0.21 atm-m3/mol U.S. EPA, 1981 Bioconcentration Factor 3100 to 69,000 in fish U.S. EPA, 1981 The potential to cause cancer and birth defects led to toxaphene’s registration for most uses to be cancelled in 1982 by the U.S. Environmental Protection Agency and banned completely in 1986. U.S. EPA regulates toxaphene under the Clean Water Act (CWA); Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA); Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA); Food, Drug, and Cosmetic Act (FD&CA); Resource Conservation and Recovery Act (RCRA); Safe Drinking Water Act (SDWA); and Superfund Amendments and Reauthorization Act (SARA). For instance, FIFRA issued a rebuttable presumption against registration (RPAR), which established a maximum contaminant level (MCL) of 5 ppb and a maximum contaminant level goal (MCLG) of 0 ppb for toxaphene.

PAGE 18

4 Although it was banned in 1982, toxaphene is very persistent in nature and recent research on old samples has found that toxaphene's levels have remained very high (U.S. EPA, 2003). Toxaphene has been identified in at least 58 of the 1,430 current or former hazardous waste sites that have been proposed for inclusion on the U.S. EPA National Priorities List (NPL) (ATSDR, 1996). For instance, in Lake Apopk, Florida, the average concentration of toxaphene in sediments was 10977 ppb in 2000. The GIS map based on the data provided by St. John River Water Management District (SJRWMD) is shown in Figure 1-3. As a dense non-aqueous phase liquid (DNAPL), toxaphene is not expected to leach appreciably into groundwater or to be removed significantly by runoff unless adsorbed by sediments because of its low solubility in water (3 ppm). In accordance with this, toxaphene has been found to strongly adsorb to sediments (Log Kow = 3.3) and has been proven to be difficult to break down in the aerobic environment (Nash and Woolson, 1967). When released to the soil, toxaphene has been noted to persist for long periods of time, from 1 to 14 years (U.S. EPA, 2003). Thus, it is more likely to be found in air, soil, or sediment at the bottom of lakes and streams. Because toxaphene can be dispersed to remote areas by the atmosphere, its volatilization should be carefully monitored during large-scale removal. However, disagreement of it vapor pressure with a range from 4*10-6 to 0.4 mmHg at 20 C (Table 1-2) exists in the literature(Mackinson et al., 1981, Wauchope et al.,1992, Bidleman et al.,1981). Generally, toxaphene is regarded as semi-volatile. Potential by-products from toxaphene are potentially harmful as well (Parr and Smith, 1976). Due to its harmful nature and persistency in the environment, it has

PAGE 19

5 become necessary to study the degradation of toxaphene to remove or decrease its influence on the environment and surrounding ecological systems. This research will explore different remedial approaches for removal of toxaphene contaminants and will be divided into two parts: (1) immobilization technology of toxaphene from a contaminated site using cosolvent flushing; and (2) dechlorination technologies of toxaphene applying zero-valent iron treatment and the effects of cosolvent solutions on it. Biodegradation technology and its combination with cosolvent will also be discussed for the purpose of identifying opportunities and needs.

PAGE 20

6 Figure 1-1. GC/ECD chromatograms of toxaphene standards by many other companies

PAGE 21

7 Figure 1-2. HRGC/ECD chromatogram of technical toxaphene. I-V are the investigated congeners; II includes two coeluting isomers (IIa and IIb).

PAGE 22

8 Figure 1-3. GIS map of toxaphene pesticide samples in north shore of Lake Apopka , FL

PAGE 23

CHAPTER 2 COSOLVENT FLUSHING TECHNOLOGY FOR TOXAPHENE MASS REMOVAL Introduction Techniques for risk reduction highlight those that remove the contaminants. Many remediation technologies, including air/steam stripping, carbon absorption, thermal technologies, and physical barriers, have been used to remove chlorinated contaminants from the environment. Due to the unique physical properties of toxaphene, many of these will not be effective in toxaphene remediation. Cosolvent flushing can provide a rapid mass removal of DNAPLs at sites with moderate to good permeability. It involves pumping a mixture of water and one or more solvents through a contaminated zone to remove DNAPL by dissolution and/or mobilization. Alcohols are increasingly used in the remediation of subsurface contamination involving chlorinated organics, such as PCE and TCE, which are sparingly soluble chlorinated organic compounds (Jawitz et al., 1998). Due to its ability to both increase DNAPL solubility in water and decrease DNAPL-water interfacial tension, alcohols (or cosolvents) can be used for a solubilization flood, a mobilization flood, or both depending on alcohol selection and injection concentration. The solubility of a given component in a mixture may be altered by other components that may act as cosolutes or cosolvents. However, due to its complex structure,toxaphene can not be characterized by the individual components partition coefficients. In this research, toxaphene is taken as a single species. In order to evaluate and potentially implement cosolvent flushing as a remedial technology at a given site, a step-wise approach is needed, including the following 9

PAGE 24

10 primary elements: (1) initial assessment of applicability and approach; (2) additional site characterization; (3) laboratory studies; (4) numerical simulations; (5) field demonstrations; and (6)full scale implementation (Jafvert, 1996). This research first focused on the initial assessment of the feasibility of increasing the solubility of toxaphene in aqueous solutions and decreasing the sorption of toxaphene to the soil due to the addition of cosolvents, and subsequent focus of this work investigated the application of cosolvent flushing on toxaphene removal from soil. Solubility Background Cosolvent solubilization enhancement is described mathematically using a cosolvency power (Yalkowsky and Roseman, 1981). A log-linear equation, where the log solubility (moles/L) of a solute in a mixture solvent system (Ss,m) is the weighted sum of its log solubility in pure water (Ss,w) and in pure cosolvent (Ss,c), was provided to describe the phenomenon of the exponential increase of the solubility for nonpolar organic solute as the increase of cosolvent fraction (Yalkowsky and Roseman,1981 ). )log()1()log()log(,,,wsccscmsSfSfS 2-1 For a given cosolvent-water system, the difference between log solubilities of the solute in pure cosolvent and pure water was defined as the cosolvency power of this cosolvent (Yalkowsky and Roseman, 1981): )log(,,wscsSS 2-2 The solute solubility in the mixed solvent can be expressed as (Li and Yalkowsky, 1998a):

PAGE 25

11 iciwsmsfSS )log()log(,, (2-3) Where, fci is the volume fraction of ith cosolvent; i is the cosolvency power for solute in pure ith cosolvent. Similar to Raoult’s law, where the partial pressure of a component in a mixed liquid is the product of its mole fraction and its vapor pressure (ATSDR,1996), the log-linear relationship of solubility is based on an assumption of the ideality of the solutions, which means all of the components of the mixture behave identically. Practically, the mixture solution is not ideal and the log-linear model showed some deviations with the experimental data (Morris et al., 1988, Li et al., 1994). For non-ideal aqueous mixture, its excess free energy, gE, is a function of the activity coefficient of the component i, which is a measure of the extent of deviation from the ideal behavior. )ln(iiExRTg 2-4 where, R is the gas constant, T is the absolute temperature, and xi is the mole fraction of component ith in the solution. To predict the activity coefficient of a component in a non-electrolyte, non-polymeric aqueous mixtures, a tool UNIFAC (UNIQUAC functional-group activity coefficients) was employed (van Genuchten, 1981). The UNIFAC model assumed that the physical property of a fluid is the sum of contributions made by the molecule’s functional groups, therefore, divided the activity coefficient into combinational part c, representing the size and shape of the molecules, and residual portion r, reflecting the functional group. rclnlnln 2-5

PAGE 26

12 Using necessary parameters estimated from the number and type of functional groups and a series of equations, the values of c and r can be calculated (Gupte and Danner, 1987; Pinal et al., 1990; Powers et al., 2001; and Li, 2001). Applying the idea of activity coefficient to the cosolvency power in non-ideal mixture, the difference between the logarithmic value of equilibrium activity coefficient in the pure water and pure solvents is the cosolvency power of this given cosolvent-water system (Yalkowsky and Roseman, 1981): )log(cw 2-6 where, w and c are the activity coefficients in the pure water and pure solvent respectively. Similarly, the octanol-water partition coefficient of the solute (Kow) is (Morris et al., 1988): )log(logowowK 2-7 where, w and o are the activity coefficients in the pure water and octanol respectively. Considering the fact that the log(c) in a cosolvent is proportional to log(o) in octanol (Yalkowsky and Roseman, 1981; Morris et al., 1988), the cosolvency power can be derived from equation 2-6 and 2-7: bKaow )log( 2-8 The regression parameters a and b are independent of the solutes and specific for the cosolvent. The values of a and b for methanol, ethanol, and IPA have been provided by Morris et al.(1988). Using the toxaphene’s referenced data of 3.3 for logKow (U.S. EPA, 1981), the estimated values for their cosolvency powers are presented in Table 2-1.

PAGE 27

13 Li and Yalkowsky (1998a) extracted a total of 607 sets of thousands of solubility data from literature sources and obtained the values of a and b for fifteen CMOSs used widely in pharmaceutical and environmental activities. In this research, the cosolvency powers for methanol, ethanol , propanol and isopropanol are examined and compared with the equations provided by Morris et al.(1998) and Li and Yalkowshy (1998a) ( Table 2-1). Table 2-1. Estimated solubilities of toxaphene in methanol, ethanol, Propanol and IPA using theoretic estimation. (Morris et al., 1988) (Li and Yalkowsky, 1998) a b Cosolv Power a b Cosolv Power MeOH 0.68 1.07 3.31 0.89.02 0.360.07 3.16-3.43 EtOH 0.85 0.81 3.62 0.95.02 0.300.04 3.33-3.54 IPA 0.89 0.63 3.57 1.11.07 -0.05.18 2.75-3.57 ProOH N/A N/A N/A 1.09.05 0.010.13 3.31-3.90 To quantify the deviation caused by the nonideality of the solvent mixture, equation 2-3 was modified as (Rao et al., 1991) iciiwsmsfSS )log()log(,, (2-9) where, i is the empirical coefficient used to account for water-cosolvent interactions. In some CMOSs-water mixture solutions, such as methanol, ethanol and propanol, the interaction of water-cosolvent interaction is negligible, and the value of is always to be assumed to be 1 (Rao et al., 1991). However, because is not constant with the change of cosolvent fraction fc, the application of equation on the solubility prediction should be investigated carefully at a specific range of the cosolvent fraction. Li (2001) further provided an approach to accurately predict cosolvency, by extending the log-linear model with the activity coefficients of the system components. Pinal et al. (1991) proposed that a term 2.303 (fi log i), which is the analogue to (xi

PAGE 28

14 ln i), be added to simple log-linear model to account for the effect of the solvent nonideality. Li (2001) extended this equation to four equations with volume fraction fc or mole fraction x, and logarithms of the activity coefficient terms are based on 10-based or e-based: )log(logloglog,,wccwcwsmsffSS (2-10 a) )log(logloglog,,wcwcwsmsxfSS (2-10 b) )ln(lnloglog,,wccwcwsmsffSS (2-10 c) )ln(lnloglog,,wcwcwsmsxfSS (2-10 d) where x is the mole fraction of the cosolvent in the mixture solution. The objectives of this research included: (1) test the cosolvency powers for different types and fractions of cosolvent solutions with binary or ternary systems for toxaphene solubilization; (2) study the change of the cosolvency power of toxaphene in these solutions as a function of cosolvent molecular weight; and (3) compare the experimental data of cosolvency powers and solubilities of toxaphene for Methanol, Ethanol, propanol, and IPA with theoretical values estimated from literature (Table 2-1). Methodolology Materials The yellow waxy solid toxaphene was purchased from UltraSci company with Lot number 302-1258. Toxaphene solubility was examined in both binary solvents and ternary solvents mixtures. The binary solvent systems include methanol-water, ethanol

PAGE 29

15 water, propanol-water, and isopropanol-water at various fractions in solution. These solvents are classified as Completely Miscible Organic Solvents (CMOSs) in water. Some higher molecular alcohols are classified as Partially Miscible Organic Solvents (PMOSs) in water and their application in cosolvent flushing results in ternary solvent systems with methanol cosolvent solution to form a single liquid phase. The ternary systems examined in this research included butanol-methanol-water, isobutanol-methanol-water, pentanol-methanol-water, and hexanol-methanol-water. The chemical and physical properties of these solvents are presented in Table 2-2. The selection of the cosolvent solutions was based on properties such as their non-toxicity, biodegradability, economic efficiency, and common availability as reported by Chawla et al. (2001). Deionized and distilled water (DDI water) were both used in the solubility experiments. Table 2-2. Chemical and physical properties of chosen solvents ( refer to the UAkron Chemical fact sheet) Properties methanol Ethanol Isopropanol propanol isobutanol butanol pentanol hexanol CAS 67-56-1 64-17-5 67-63-0 71-23-8 78-83-1 71-36-3 30899-19-5 111-27-3 Formula CH4O C2H6O C3H8O C3H8O C4H10O C4H10O C5H12O C6H14O Fornula mass 32.04 46.07 60.10 60.10 74.12 74.12 88.15 102.18 Melting point C -98 -117 -89.5 -127 -108 -89 -117 -52 Boiling point C 64.6 78.3 82.4 97.2 107.9 117.6 131 156 Vapor pressure (25 C) 127 60 79 21 10 7.24 23 0.9 Evaporization number(diethyl lether=1) 6.3 8.3 11.0 17 24 33 N/A 0.05 Density g/mL (20 C) 0.7948 0.7895 0.786 0.8053 0.802 0.812 0.8155 0.8322 Solubility in water, ug/mL Miscible Miscible Miscible Miscible 100,000 63,200 22,658 25,745 Surface tension g/s2 (20 C) 22.61 22.75 21.19 23.75 23.0 26.28 N/A 25.73 Partition coefficient pKow -0.77 -0.31 0.05 0.25 0.76 0.88 N/A 2.03

PAGE 30

16 Table 2-2. Continued Properties methanol Ethanol Isopropanol propanol isobutanol butanol pentanol hexanol Dielectic constant (25 C) 32.6 24.3 20.3 20.1 17.7 17.8 14.7 13.3 Dipole moment Debye (20 C) 1.70 1.69 1.58 1.68 1.79 1.66 1.68 1.60 Experimental procedure Initially, 5 mg of pure toxaphene was placed into the 5 mL empty vials with a Teflon-coated, septum-lined cap. Different types and fractions of cosolvent solutions were added to the vials leaving no headspace. The vials were weighed before and after to keep accurate account of weight and volume of solutions. Different cosolvent fractions of 0% , 10%, 20%, 30%, 40%, 50%, 60%, and 75% were chosen for the binary systems. Due to the partial solubility of PMOSs, ternary systems were employed in which the various cosolvents were mixed with 50% methanol solution. The solubility of the solvent in the solution containing 50% methanol could be estimated from the equation 2-3, where the solubility for isobutanol, butanol, pentanol, and hexanol in water are 100*103 mg/L, 63.2 *103 mg/L, 22.658 *103 mg/L, 25.745 *103 mg/L respectively (Table 2-2). The cosolvency power for methanol was 3.43 based on the estimation from data provided by Morris et al. (1988). Methanol fraction for these experiments was 0.5, therefore the solubility of isobutanol, butanol, pentanol, and hexanol were estimated to be 5623 *103 mg/L, 3554 *103 mg/L, 1274 *103 mg/L, and 1447.7 *103 mg/L respectively, which seems to indicate that all of these solvents are miscible in 50% methanol solution. Volumes of cosolvent and water were measured separately and combined to avoid the volume change due to the volume shrinkage during mixture.

PAGE 31

17 The solutions were equilibrated on a rotator for 48 hrs at room temperature. Times of 24 hrs, 48 hrs, 1week, and 8 weeks for 50% methanol cosolvent solution were chosen to test the equilibrium of the batch tests. After rotation periods, there was still a yellow solid toxaphene phase visible at the bottom or the wall of the vials, indicating toxaphene in solution had reached its equilibrium limit. The results showed that the solubilities for 24hrs, 48 hrs, and 1week did not show significant difference, therefore, the equilibrium of the batch tests was assumed to be reached (Figure 2-1). If the micelle formed in the solution, the vials were centrifuged at 3000 rpm for 25 mins. A volume of 1 mL solution without a toxaphene phase was transfered into another vial. The vial was left in the hood for 24 hrs to remove most of the cosolvent, similar to the method used by James and Hiter (2002). To test the influence of the sample concentration procedure, comparison of toxaphene in pure water with and without exposing in the hood was conducted. The results showed that toxaphene concentration didn’t change significantly in the hood after 24 hrs. For analyses, a calibration curve was developed using known concentrations (0.5, 1, 2, 5, 10, 20, 50 g/mL) of toxaphene solutions in different fractions of methanol solutions utilizing same method of sample concentration. After concentration, 4 mL hexane was placed into the solution to extract toxaphene from the water. Since the concentrations might have been higher than the highest standards of the calculation curve, appropriate dilutions were made for analysis with the GC-ECD. Analytical methods The determination of trace amounts of toxaphene has been restricted to a limited number of analytical techniques. This restriction is necessary because of the difficulty in analyzing and characterizing toxaphene with complex and inconsistent structure. The

PAGE 32

18 most proven analytical methods include application of a gas chromatograph equipped with either an electron capture detector (GC/ECD), a micro-coulometric detector (GC/MC), or negative ion chemical ionization mass spectrometry (GC/NCIMS), and thin-layer chromatography (TLC). Each method provides its own applicability and advantages based on its specific method of analysis. Since the ECD has a unique high sensitivity for substituents such as halogens, GC/ECD is employed as the most prevalent analytical technique to determine trace amounts of toxaphene in biological and environmental samples (Griffith and Blanke, 1974; ATSDR, 1996). Vaz and Blomkvist (1985) used GC/NCIMS to quantitatively and selectively detect components of toxaphene at ppb (ng/g) levels in human breast milk. One of the more prevalent chromatographic analytical techniques, employment of a gas chromatograph with a mass spectrometer (GC/MS), is not recommended for toxaphene due to the complex multi-component nature of the chemical (U.S. EPA, 1995). The analysis methods for toxaphene in wastewater and soil have been suggested by the U.S. EPA (ATSDR, 1996). For analysis utilizing the GC/ECD, U.S. EPA Method 8081A is the standardized method for determining toxaphene in water and soil samples (U.S. EPA, 1995). Detection limits of 0.086 g/L in water and 5.7 g/kg in soil were reported for toxaphene by this method. Based on the optimization data, accurate quantification of technical toxaphene by Method 8081A would be possible at different analytical laboratories by standardizing the following: (a) sample preparation, (b) sample extraction procedures, (c) defining clean-up procedures, (d) selection of a reference standard, and (e) selecting peaks from the later retention time window of the toxaphene standard (back-half) where

PAGE 33

19 there are fewer interferences from matrix and co-contaminants effects. The sample preparation for soil will be discussed in next section. The complexity of the chromatograms produced by GC/ECD analysis of toxaphene has proven to be challenging to decipher. For this reason, there are various methods by which toxaphene has been quantified in past research. Of the varied methods, two are more generally agreed upon than all others and used by the U.S. EPA, as well as other laboratories: the total area pattern and multiple peaks pattern of analyzing toxaphene chromatographs. Toxaphene contains a large number of compounds, which produce well resolved peaks and many other unresolved components as well (Figure 1-1). This complex mixture results in the "hump" in the chromatogram that is characteristic of this mixture and contributes a significant portion of the area of the total response (Fingerling et al., 1996). To measure total area, the baseline of toxaphene is constructed in the sample chromatogram between the retention times of the first and last eluting toxaphene components in the standard (Paris et al., 1977). In order to use the total area approach, the pattern in the sample chromatogram must be compared to that of the standard to ensure that all of the major components in the standard are present in the sample. Otherwise, the sample concentration may be significantly underestimated. The usefulness of this method is found in its prevalent application by various researchers (Khalifa et al., 1976; Paris et al., 1977; Pearson et al., 1997; Gray et al., 2000). In evaluation of the bio-concentration of toxaphene in microorganisms, Paris et al. (1977) found recoveries of toxaphene to range from 77 to 91% using the total peak area.

PAGE 34

20 Similarly, Pearson et al. (1997) recorded toxaphene recoveries of 84% using the total peak area in evaluation of toxaphene found in Great Lake sediments. In multi-component samples usually found in the environment, there can be difficulties in recognizing toxaphene peaks present in the early portion of the chromatogram due to interferences from compounds, such as DDT and DDE. To solve this problem, toxaphene has also been quantified on the basis of 4 to 6 major peaks in various analytical settings. When using the 4to 6peaks approach, it is highly unlikely that the peaks will match exactly, but the chosen peaks’ relative sizes or areas appear to be disproportionally larger or smaller in the sample compared to the standard. The heights or areas of the 4 to 6 peaks are summed and used to determine the toxaphene concentration. With increasing sensitivity in the arena of GC peak separation, the multiple peak method has also seen frequent application in the quantification of toxaphene. Saleh and Casida (1978) used multi-peak pattern methodology in analysis of the production of various hepta-chloroboranes, components of toxaphene, and found this method was adequate in distinguishing between toxaphene and the components produced by reductive dechlorination. Mirsatari et al. (1987) used six peaks in the analysis of toxaphene for anaerobic microbial dechlorination and obtained recoveries of toxaphene ranging from 84% to 86%. Although GC/MS is not recommended for toxaphene because of limitations in sensitivity arising from the multi-component nature of toxaphene (U.S. EPA, 1995), this method can still be used to investigate the structure of toxaphene components and their stability during the process of degradation. Based on this investigation, the degradation pathways and schemes can be studied. Fingerling et al. (1996) used GC/MS and

PAGE 35

21 suggested the probable degradation pathway of selected chlorobornanes. Buser et al. (2000) also employed GC/MS for toxaphene analysis in respect to anaerobic degradation. GC/MS proved effective in the toxaphene quantification after the employment of multiple extraction and clean-up procedures to provide better separation of the desired peaks (Buser et al., 2000). In this research, The GC-17A Shimadzu was used for analysis of the hexane extracted aqueous phase. The total area approach was used for measuring the toxaphene present in the hexane extracted aqueous phase (Fingerling et al., 1996, Paris et al., 1977; Pearson, 1997). The operation conditions were as follows: column: DB-5 with 30m length*ID 0.32mm, film thickness 0.25 m; carrier gas: H2; injector temperature: 220 C; detector temperature: 300 C ; column temperature program: initial temperature 100 to 150 C at 30 C /min hold 2minto 300 C at 3 C /min hold 5min. For the higher molecular weight alcohols that were not easily concentrated by volatilization, the cosolvent solutions were analyzed by GC-FID directly after the rotation period. Results and Discussion Comparison of estimated and experimental data Using equation 2-3, the log-linear relationship of binary systems could be expressed as (Rao et al.,1991): cwmfSS )log()log( (2-11) The extended equations (equation 2-10) using the UNIFAC method (Li, 2001) were applied to analyze the solubility for binary system. The statistical analysis results for methanol, ethanol, isopropanol, and propanol using log-linear and extended log-linear models were illustrated in Figure 2-2 and summarized in Table 2-3, where the percentage

PAGE 36

22 difference between estimated Log(Ss,est) and experimental Log(Ss,expt) were calculated and compared, 100*log)log(log%exp,,exp,tseststsSSSdifference (2-12) Table. 2-3. Percent of difference of the predictions ( logSs,estd-logSs,expt)/ logSs,expt*100 using the log-linear and extended log-linear models for binary cosolvent systems. log-linear Equation a Equation b Equation c Equation d 10% -2.38692 1.316689 -0.56027 6.140957 1.819091 20% 15.28777 20.65135 18.25421 27.63788 22.11825 30% 2.722243 7.401085 5.623084 13.49567 9.401676 40% 4.442249 8.733797 7.4729 14.3239 11.42058 50% -1.89006 1.516849 0.876186 5.954639 4.479458 60% -4.81819 -2.19546 -2.33665 1.220862 0.895754 75% -1.55511 0.211947 0.551661 2.513685 3.295904 ave 1.685998 5.376608 4.26873 10.18394 7.632958 SD 6.78067 7.789711 7.070724 9.18467 7.474887 Max 15.28777 20.65135 18.25421 27.63788 22.11825 MeOH Min -4.81819 -2.19546 -2.33665 1.220862 0.895754 10% -27.6985 -23.7885 -28.0644 -15.2161 -25.0618 20% -5.056 0.360166 -5.78479 13.79897 -0.35031 30% -1.43004 3.007364 -2.45368 16.42654 3.852021 40% 3.467343 6.335371 2.106512 18.75229 9.014979 50% -7.71693 -6.94622 -9.26081 2.204176 -3.12537 60% -3.52485 -4.61236 -5.5965 2.7944 0.528319 75% -3.21536 -6.8941 -6.25683 -2.50358 -1.03622 ave -6.45347 -4.64832 -7.9015 5.179529 -2.3112 SD 9.980135 9.840449 9.58061 12.07313 10.78622 Max 3.467343 6.335371 2.106512 18.75229 9.014979 EtOH Min -27.6985 -23.7885 -28.0644 -15.2161 -25.0618 20% 12.296 25.65151 13.13289 49.43403 20.60885 30% 5.861828 17.16235 6.726988 38.82985 14.80154 40% 9.802747 19.01564 10.64162 38.82404 19.54215 50% 2.505652 8.325624 3.120713 23.57992 11.59517 60% 1.690982 4.585559 1.98091 16.24564 10.24821 75% 1.2665 0.178884 0.699646 6.915958 8.115056 ave 5.570618 12.48659 6.050462 28.97157 14.15183 SD 4.608634 9.685307 5.009455 16.06511 5.087008 Max 12.296 25.65151 13.13289 49.43403 20.60885 IPA Min 1.2665 0.178884 0.699646 6.915958 8.115056

PAGE 37

23 Table. 2-3. Continued. log-linear Equation a Equation b Equation c Equation d 20% 3.016232 15.14997 3.777705 36.81334 10.62772 30% -10.428 -1.00981 -9.70667 17.1367 -2.88858 40% -13.8287 -6.74673 -13.1737 8.605577 -6.19302 50% -7.10339 -1.98508 -6.54951 11.63594 1.12594 60% -6.58668 -4.07369 -6.3365 6.447124 1.236795 75% -5.3646 -6.46843 -5.90555 -0.2864 1.009675 ave -6.71585 -0.85563 -6.3157 13.39205 0.819756 SD 5.67589 8.173656 5.667993 12.83671 5.646781 Max 3.016232 15.14997 3.777705 36.81334 10.62772 ProOH Min -13.8287 -6.74673 -13.1737 -0.2864 -6.19302 From Table 2-3, the extended log-linear model (equation 2-10 d) for ethanol and propanol generated more improved solubility estimations than the simple log-linear model. However, for methanol and isopropanol, the extended models produced higher averaged percentage errors than the simple log-linear model. This phenomenon that the log-linear model fitted well with polar solvents such as methanol and IPA may be attributed to the more ideality state of the mixture of water and higher polar solvents, where the log-linear solubilization mostly prevails (Li, 2001). Cosolvency power The log-linear plots for the binary system were presented in Figure 2-3. The slope for this log-linear regression is the cosolvency power for the cosolvent. Results indicated that the experimental data mirrored the data estimated from theoretical log-linear equation (Morris et al., 1988). The experimental data were also compared with the UNIFAC model ((Li and Yalkowsky, 1998a). The comparison shown in Table 2-6 indicated that the UNIFAC modified log-linear model can not enhance the accuracy of the prediction of the cosolvency powers for the selected solvent when the solute is

PAGE 38

24 toxaphene. Although the UNIFAC calculation algorithm was base on theoretical derivation, its group interaction parameters were obtained empirically from different solutes (Li, 2001) which may cause inaccurate prediction due to the deviation. In this research, the simple log-linear regression produced acceptable predictions for the solubility in the selected cosolvent systems. The log-linear model can be modified with the solvent-water interaction constant (Rao et al., 1991). The slope of the log-linear regression of the log solubility versus cosolvent fraction represents the value of . Using the estimated cosolvency power, the values for were nearly 1(Table 2-4), which validated the assumption that the solvent-water interaction is negligible for binary system. Table 2-4. Comparison of cosolvency powers from experimental data and from log-linear model(Morris et al., 1988) and UNIFAC model (Li and Yalkowsky, 1998). (Morris et al., 1988) (Li and Yalkowsky, 1998) expt. est. %diff. est. %diff MeOH 3.43 3.31 3.38 1.04 3.30 3.88 EtOH 3.64 3.62 0.69 1.01 3.44 5.63 IPA 3.51 3.57 -1.62 0.98 3.16 9.89 ProOH 3.91 N/A N/A N/A 3.61 7.75 For the ternary systems, PMOSs were applied to 50% methanol solutions. The cosolvency power for toxaphene in the methanol-water binary system was 3.43 (Figure 2-3). Thus, the cosolvency power equation for ternary system could be expressed as cwmfSS 5.0*46.3)log()log( (2-13) The slope of the plots of the ternary equations for the PMOSs are presented in Figure 2-4 for isobutanol, butanol, pentanol, and hexanol. The interception is the value of LogSw +3.46*0.5 = 2.21, which had a very good agreement with the values shown in these Figures.

PAGE 39

25 The change of cosolvency power of different alcohol molecular weight for toxaphene solubilization is presented in Figure 2-5 and Table 2-5. The results indicated that the cosolvency power was enhanced with the increase of molecular of alcohol due to the increase of non-polarity. The PMOSs such as butanol, pentanol and hexanol can enhance the solubility to a great extent than CMOSs such as methanol, ethanol and propanol. Addition of PMOSs into the CMOSs cosolvent system to form ternary system can increase the solubility of toxaphene (Table 2-5). Table 2-5. The cosolvency power of different types of solvent MeOH EtOH IPA ProOHl IBA BtOH PtOH HexOH 3.43 3.64 3.51 3.91 5.94 4.56 7.59 8.74 The cosolvency power of propanol was 11.4% higher than isopropanol and that of butanol was 30.2% higher than isobutanol. This phenomenon of the straight carbon chain had higher cosolvency power than branch carbon chain could be attributed to the fact that the branch will increase the polarity of the molecular and therefore reduce the solubility of the highly hydrophobic toxaphene. Generally, the physical properties used to measure the polarity of a solvent include the dipole moment, the dielectric constant and the miscibility with water. Dipole moment is the product of the distance of a pair of separated opposite electric charges and the positive change. Dielectric constant or permittivity is an index of the ability of a substance to attenuate the transmission of an electrostatic force from one charged body to another. Molecules with larger dipole moments, higher dielectric constants and more miscible with water are considered more polar. The dipole moments, dielectric constants and solubility in water of the selected cosolvents are presented in Table 2-2.

PAGE 40

26 The cosolvency powers for the selected cosolvents increased with the decrease of the dielectric constants due to the increased non-polarity of the cosolvent (Figure 2-6). Because the dipole moments for the selected cosolvents were in a small range from 1.58 to 1.79, the relationship between the dipole moments and the cosolvency powers were not conclusive. For binary system, the cosolvency powers for these CMOSs with unlimited solubility in water were in the range from 3.43 to 3.91. The cosolvency powers for PMOSs increase with the decrease of the solubility of the solvent in water (Figure. 2-7). Sorption Background The aqueous concentration of a solute Ce,i (mg/L) follows Raoult’s law under ideal conditions (Banerjee, 1984): iwoiieSXC,, (2-14) where, Sw,i is the aqueous solubility of ith pure solute , and Xio is mole fraction of the ith component in the organic liquid. The aqueous concentration Ce,i can also be expressed using measurable parameters in commonly reported concentrations units (Dai,1997): )10)()()((3,wmixiieSMWMC (2-15) where, Mi is the concentration (mg/kg) of the ith organic compound of concerned in NAPL mixture, MWmix is the average molecular weight (g/mole) of the NAPL mixture. The partitioning coefficient of the organic compound Kw is the ratio of the concentration of interested compound in the organic phase to its concentration in aqueous phase (Cline et al.,1991):

PAGE 41

27 iwooiwoiowSMWSXCK,,,)( (2-16) where, MWo and o are the average molecular weight and density of the organic phase, respectively. Thus, the partitioning coefficient Kw and solubility Sw are expected to have an inverse log-log linear relationship: )log(loglogoowwMWSK (2-17) In the case of chemical compounds adsorbed to the soil, the partitioning coefficient can be expressed as a sorption constant Kp(mL/g), which is the ratio of the concentration in aqueous phase Ce (g/mL) to the concentration in sorbent S (g/g). Combining equation 2-9 and equation 2-17, a log-linear cosolvent model has been used to examine cosolvent effects on the sorption (Rao et al., 1985). icisiiwpmpfKK ,,,loglog (2-18) where Kp,m and Kp,w are the equilibrium sorption constants for the mixture cosolvent solution and pure water, respectively; fci is the volume fraction of ith cosolvent; i and i are the empirical coefficients used to account for solvent-sorbent interactions and water-cosolvent interactions respectively; s,i is the cosolvency power for solute in pure ith solvent. The objectives of this research are: (1) calculate the partition value Kp for different fractions of binary and ternary systems; (2) calculate the cosolvency power for methanol, ethanol, propanol, pentanol, and hexanol in soil absorption, and compare with solubility test, and discuss the value of ; (3) compare the effects of factors including different soil:solution ratio and time changes.

PAGE 42

28 Methodolology Materials Eustis #3 fine sand from Gainesville, Florida was used as soil to investigate the sorption isotherm of toxaphene. The Eustis #3 sand had a 0.39% organic carbon content and 96% of the sand fraction. Methanol, ethanol, propanol, pentanol, and hexanol were chosen to study the effects of cosolvent on the sorption isotherms of toxaphene to soil. Toxaphene solid was dissolved in the pure solvent solution to prepare the required concentration of cosolvent solution. Sorption isotherm of toxaphene to soil Three binary solutions including methanol-water, ethanol-water, and propanol-water were produced with alcohol fractions of 30%, 40%, 50%, 60%, and 75%. Ternary solutions were also developed with 50% methanol in solution with butanol, pentanol, and hexanol, whose fractions were from 0.05 to 0.25. CaCl2 with a concentration of 0.01 M was used as matrix. To determine the equilibrium sorption constant Kp for toxaphene and the values of each fraction of cosolvent solution, at least three different initial fractions of toxaphene at the range of 20% to 80% were prepared. Four grams of soil was placed into these 25mL vials with Teflon-coated, septum-lined caps, followed by addition of 12 mL of cosolvent solutions with different concentrations of toxaphene and performed in triplicate. The empty vials, vials+soil, vials+soil+solution were weighted respectively. Then, the vials were rotated at 40 rpm for 48 hrs. Preliminary studies of toxaphene concentration as a function of time for the binary and ternary solutions indicated that the systems would reach equilibrium after 1hr, and

PAGE 43

29 that Kp does not significantly change after this point (Figure 2-8). The 48 hrs used by other cosolvent solution test was enough to reach equilibrium state. After rotation, all the solutions were separated from the soil and centrifuged for 10min and extracted with hexane in preparation for GC-ECD analysis. To investigate the kinetics of sorption and determine the time required to attain the equilibrium state could reach, 50% methanol binary cosolvent solution with V:M = 3:1 was analyzed at the different time periods of 0.25, 0.5 ,1 , 2, 6, 12, 24, 48, 72, and 168 hrs. In order to study the effects of the ratio of solution to soil (volume: mass) on the Kp value, The ratio of solution to soil (mL:g) was studied utilizing 50% methanol binary cosolvent to extract with different ratios of V:M of 1:1, 2:1, 3:1, 5:1, 7:1 were used with 48 hr rotation time. Analytical methods The analysis of the toxaphene concentration in solution was the same as that discussed in the solubility section. For soil sample preparation, random samples were collected from different composite (based on depth) samples and were mixed together to form a single sample. The soil samples were air dried and ground prior to analysis by U.S. EPA Methods 8080 or 8081A. From this point, the sample matrices were ready for extraction or clean-up procedures. The quantification of toxaphene in soil samples is affected by moisture level, the physical processing, and extraction procedure used prior to analysis by U.S. EPA Method 8081A. U.S. EPA has provided many methods for extraction based on the matrix in which the toxaphene was analyzed. Liquid samples were extracted at neutral pH with

PAGE 44

30 methylene chloride employing Method 3510 (separatory funnel), Method 3520 (continuous liquid-liquid extractor), or other comparable techniques. Solid samples were extracted with hexane-acetone (1:1) or methylene chloride-acetone (1:1) using Method 3540 (Soxhlet), Method 3541 (automated Soxhlet), Method 3545 (pressurized fluid extraction), Method 3550 (ultrasonic extraction), or other appropriate technique. Gray et al. (2000) studied and compared different extraction methods, extraction solvent systems, and moisture levels to investigate their impact on toxaphene quantification in soil. Samples were extracted using three different methods (a) Soxhlet (U.S. EPA Method 3540C) using a 1:1 mix of methylene chloride and acetone, (b) wrist action shaker using a 1:1 mix of methylene chloride, and (c) acetone wrist action shaker using methylene chloride. Soxhlet extraction with methylene chloride and acetone was superior to less rigorous wrist-action extraction procedures studied, resulting in significantly higher levels of toxaphene being detected. Other studies comparing Soxhlet and ultrasonic-extraction methods produced similar results when the same extraction solvent was used (Gray et al., 2000). Lower moisture levels and physical processing of soil samples also had a significant effect on the toxaphene concentration measured. Highest levels of toxaphene were measured in soil samples that were dried and ground. The same trend was also observed for three other chlorinated co-contaminants present in the soil samples. Parr and Smith (1976), Mirsatari et al. (1987), and Pearson et al. (1997) have also successfully employed Soxhlet extraction in toxaphene analytical research. The analysis of toxaphene is often interfered with by contamination from widely prevalent polychlorinated biphenyls (PCBs), 1,l -dichloro-2-2-bis (chlorphenyl) -ethane (DDE), and other organochlorine pesticides, which are also complex multi-isomeric

PAGE 45

31 chemicals. Therefore, to get accurate quantification, a variety of clean-up steps may be applied to the extract prior to chemical analysis by U.S. EPA Method 8081A, depending on the nature of the matrix interferences and the target analytes (Matsumura et al. 1975; Nelson and Matsumura 1975; Gooch and Matsumura 1985). Suggested clean-ups include the employment of alumina (Method 3610), Florisil (Method 3620), silica gel (Method 3630), gel permeation chromatography (Method 3640), sulfur (Method 3660) and sulfuric acid (based on SW846 Method 3665A). Sulfuric or nitric acid clean-up is the most commonly used method and is used to remove a number of single component organochlorine and organophosphorus pesticides prior to PCB analyses by U.S. EPA Method 8082. Sulfuric acid clean-up was investigated as a means to remove co-contaminants in soil samples prior to toxaphene analysis by Method 8081A. A 500 ppm reference standard, containing twenty OCPs (organochlorine pesticides) measured by U.S. EPA Method 8081A, was subjected to sulfuric acid clean-up (U.S. EPA method 3660). Of the twenty components studied, DDE, endosulfan I, endosulfan II, endrin, endrin aldehyde, and methoxychlor were completely removed by the acid clean-up. Four components: aldrin, beta-BHC, endosulfan sulfate, and heptachlor epoxide were partially removed (losses between 47 and 64%). Lastly, ten components: alpha-BHC, delta-BHC, gamma-BHC, alpha-chlordane, gamma-chlordane, DDD, DDT, dieldrin, heptachlor, and toxaphene were not affected by the acid clean-up. This confirmed that toxaphene quantification would not be affected by sulfuric acid clean-up. As 10 of the 20 organochlorine pesticides were completely or partially removed by sulfuric acid clean-up, this method was considered a benefit for toxaphene quantification in soils co-contaminated with OCPs. Researchers

PAGE 46

32 including Fingerling et al. (1996), Gray et al. (2000) have successfully employed this procedure in the analysis of toxaphene. Since organic amendments are used in the toxaphene degradation, gel permeation chromatography (GPC) clean-up is also often employed during method optimization. GPC, employed by researchers as Buser et al. (2000) and Gray et al. (2000), is a size exclusion procedure commonly used on soil/sediment samples for the elimination of lipids, polymers, copolymers, proteins, natural resins, cellular components, viruses, steroids and dispersed high-molecular weight components in sample extracts. In those soil samples where GPC clean-up was used, lower analytical detection limits were achieved compared to those without the clean-up procedure. Toxaphene presents a complex GC/ECD profile and often includes the presence of chlorinated co-contaminants. To reduce the variability in toxaphene values reported within one or between different analytical laboratories, a suitable toxaphene reference standard would need to be selected and consistently used when trying to achieve reproducible and accurate toxaphene quantification. The variation in toxaphene reference standards was investigated by comparing GC/ECD profiles from four commercial sources under identical GC conditions. This research employed sulfuric acid clean-up technology, extracted through ASE ( Accelerated Solvent Extraction) with 4:1 methylene chloride: acetone solvent, and analyzed by GC/ECD using the same method discussed in the solubility section. Results and Discussion Using a linear sorption isotherm, the value Kp could be obtained from the slope. Previous studies suggest that the concentration of the solute in the solution affects the linearity of sorption isotherm (Bouchard, 2002). In order to obtain the linear sorption

PAGE 47

33 isotherm, the selection of the concentration of toxaphene in the solution should be evaluated carefully. The results of the batch test revealed a linear relationship between S, toxaphene concentration in soil (g/g) and Ce, toxaphene concentration in solution (g/mL). For 50% methanol binary solutions, the solubility of toxaphene from the solubility test was 135 g/mL. Different concentrations of toxaphene from 20 g/mL to 100 g/mL, at the range from 15% to 80% of the solubility, were chosen to study the relationship between S and Ce. The results showed a linear relationship of S-Ce among the concentration range of 15% to 80% of the solubility (Figure 2-9) and the slope of the regression line was the equilibrium sorption constant Kp. The studies of other cosolvent solutions also suggested linear relationship when the toxaphene concentration was among 20% to 80% of solubility. Experimental results suggested that the ratio of solution volume to mass of soil had an effect on the value of Kp (Figure 2-10). The Kp value reduced as the V:M ratio increased. In this case, the application of batch experiments on the characterization of the sorption properties of toxaphene to the soil should be considered carefully. Results also showed how sorption of toxaphene related to fc for different cosolvent solutions. Data shown in Figure 2-11 displayed the log-linear relationship for toxaphene sorption to soil as a function of fc. The increase of the molecular weight of the cosolvent in solution reduced the sorption of toxaphene due to the increase of toxaphene solubility in the aqueous phase.

PAGE 48

34 The solvent-sorbent interaction constants, , were lower than 1 (Table 2-6) , which may indicate swelling the soil matrix at the presence of solvent as reported by Park et al. (1953). Table 2-6. The solvent-sorbent interaction and Kp,w value from plot of logKp-fc methanol ethanol propanol butanol pentanol Hexanol Slope 2.93 3.17 3.27 4.67 5.5 6.95 a 3.5 3.6 3.93 5.87 7.312 8.16 0.84 0.88 0.83 0.79 0.75 0.85 Kp,w 8 10 8.91 3.16 4.26 4.07 a: values of were obtained from solubility tests Column Test Background Batch experiments were employed widely in the investigation of sorption kinetics such as equilibrium sorption constant Kp. However, from the previous batch test, the value of Kp changed greatly under different ratio of V:M and the contact time is quite uncertain to reach the equilibrium state for highly hydrophobic organic such as toxaphene. In this case, the application of batch test should be with causion. It is necessary to develop column tests to characterize the sorptive properties of solute and sorbents under different cosolvent systems. In addition to batch experiments, column tests were also conducted to characterize the sorptive properties of the toxaphene to soil for various cosolvent systems. For equilibrium process, there is an assumption that the sorption-desorption process is so rapid that it is instantaneous, that is, the point-wise equilibrium existes between the solution and sorbent phase. However, practically, the sorption-desorption processes are not instantaneous.

PAGE 49

35 For steady water flow, xCvxCDtStCwwww22 (2-19) where, D is the hydrodynamic dispersion coefficient along the x direction (L2/T) and v is the average pore-water velocity along the x direction (L). For non-equilibrium sorption, mathematical expressions were employed to describe the time rate of change of amount sorbed S/t ( Nkedi-kizza, et al. 1989, and Brusseau, et al., 1989). The somewhat simple non-equilibrium models with one site do not adequately describe the measured BTCs for reactive solute displacement even in homogeneous packed columns (Nkedi-kizza, et al. 1989). To study the sorption and transport of toxaphene in a miscible displacement column experiment, a non-equilibrium bi-continuum sorption model was employed, which has been used successfully by many other authors ( Nkedi-kizza, et al. 1989, and Brusseau, et al., 1989). In this model, the sorption was divided into two sites: S1, the instantaneous domain (g/g), and S2 the first order rate-limited domain (g/g). CFKSp 1 (2-20) 22112SkSkdtdS (2-21) where, F is the ratio of the instantaneous sorption to the total sorption domain. Kp is the equilibrium sorption constant (cm3/g). The value of k1 is the forward first-order rate constant (hr-1) and k2 is the reverse first-order constant ( hr-1). The governing equation is: XCXCPpSRpCRpC22)1()1()1( (2-22)

PAGE 50

36 )()1(SCpSR (2-23) where, concentration C, distance X, and time p are dimensionless: 0CCC (2-24) lxX (2-25) lvtp (2-26) DvlP (2-27) 02)1(CKFSSp (2-28) pKR)(1 (2-29) RKFp)(1 (2-30) vRlk)1(2 (2-31) where, l is the column length (cm), is the bulk density (g/cm3), is the volumetric soil-water content. In this model, five parameters needed to be calculated or estimated: T0, P, R, ,. The value of the input pulse T0 (hr) was measured in experiments. R, the retardation factor, was calculated from moment analysis (Brusseau, 1990); and P, Peclet number was derived from an IPA tracer test; while the values of , were estimated from the program, CXTFIT 2.1. is the fraction of the instantaneous domain retardation

PAGE 51

37 in the total sorption retardation. The Damkhole number, , is a measure of the mass transfer limitations which represents the ratio between hydrodynamic residence time and the characteristic time of sorption. The objectives of this research were to: (1) dervie the parameters of the non-equilibrium bi-continuum model: T0, P, R,,; (2) derive the relationship between various parameters including equilibrium sorption constant, the reverse first order rate constant, the fraction of instantaneous domain F and the cosolvent fraction in solution for methanol flushing of toxaphene contaminated soil; (3) compare the Kp value from column test and batch test; and (4) compare binary solvent system (methanol-water) and ternary system (methanol-pentanol-water) flushing. Methodology The column was packed with Eustis #3 fine sand. IPA was used as non-reactive tracer for the column flushing experiment because the IPA couldn’t be absorbed by soil and was used as non-reactive tracer by previous researchers (Brusseau, 1991). The experiments were conducted in 2.5cm ID * 15 cm length glass column (Kontes Glass Co.) fitted with Teflon o-rings and Teflon end pieces. The length of the column was adjusted to 5 cm which was employed by other previous researchers (Brusseau, 1991 and Dai, 1997). All tubing, fitting and valves in contact with the organic fluids were constructed with Teflon. The bulk density of this sandy column was 1.7 g cm-3 and the porosity was 0.39 cm-3. The miscible displacement experiment was similar to that used by Brusseau et al. (1991). A column packed with sandy soil was initially saturated with water by pumping 0.01N CaCl2 DI water with an upward flow rate of 0.5 mL/min. All solutions were prepared with 0.01 N CaCl2 matrix, and IPA was employed as non-reactive tracer

PAGE 52

38 introduced as a 2.3 pore volume pulse input. The results of the breakthrough curves (BTC) for the tracer experiments performed before and after the cosolvent flushing were not significantly different , suggesting that cosolvent flushing had little effect on the retardation capacity of the soil column during the experiment periods. The binary cosolvent solutions included methanol at fractions of 30%, 40%, 50%, 60%, and 75%, the remaining balance being water. Ternary cosolvent solutions included 50% methanol and pentanol at fractions of 0%, 5%, 10%, 25%, and the remaining balance as water. To test the effects of initial concentration C0 on the breakthrough curves, different initial toxaphene concentrations from 50g/mL to 100 g/mL in 50% methanol solution were analyzed. These concentrations were chosen because 50 g/mL is approximately 40% of the total toxaphene solubility, which was 135 g/mL in 50% methanol solution; and 100 g/mL was approximately 80% of the toxaphene solubility. The results showed the breakthrough curves using C/C0 were almost the same among the range of C0 from 40% to 80% of solubility, suggesting the initial concentration would not affect the breakthrough curves among this range. Thus, the initial concentrations, C0, for every fraction of cosolvent solution were within a range of 50% to 80% of the overall solubility in the selected cosolvent solutions. Approximately, 2.3 pore volumes of solution containing toxaphene was pulse input through the column with upward flow rate of 1mL/min (9.57 cm/hr) followed by the flushing of the cosolvent solutions with upward flow rate of 1mL/min (9.57 cm/hr). A switching valve was used to facilitate switching between the solution containing toxaphene and the cosolvent solution without toxaphene. The effluent from the column was collected every 1 min with an automatic collector. The collection time for 30%

PAGE 53

39 methanol solution was 240 min (25 pore vlume), for 40% methanol solution was 180 min (18.8 pore volume), and for others was 120 min (12.5 pore volume) to collect most of the toxaphene solute in the solutions. Mass recoveries were higher than 99.8% for all experiments. The computer modeling program CXTFIT 2.1 was applied to estimate transport parameters from laboratory column test using a nonlinear least-squares parameter optimization method. This research used the deterministic non-equilibrium CDE (convection-dispersion equation) to deal with the inverse problem (parameter estimation); and time and position are dimensional. Results and Discussion Modeling results The one-dimensional flow model based on the bi-continuum approach was employed and the five parameters (T0, P, R, , and ) were calculated. The IPA tracer experiment with the advective-dispersive local equilibrium solute transport model provided the value of P by using a nonlinear, least-squares optimization program (Van Genuchten, 1981). dXCddXCdPdpCd2)1( (2-32) The BTC for IPA tracer (Figure 2-12) was symmetric. The Peclet number acquired from fitting this model to the experimental data was 61.9. With the velocity of 31.33 cm/hr and 5 cm length, the diffusion coefficient was 2.53 cm2/hr. The Peclet number measured in this tracer experiment was comparable to other research with similar column apparatus (Dai, 1997) and was used for the entire cosolvent flushing experimentation. The value for R and thus Kp was obtained by moment analysis (Brusseau, 1990):

PAGE 54

40 015.0TMR (2-33) where, 001dpCpdpCM (2-34) M1 was the first moment. With the equation of 2-30, the value of Kp could get from R value. The values of and were fitting using CXTFIT 2.0. The value of reverse first order rate constant of concentration, k2, was obtained from equation 2-32. The modeling results are presented in Table 2-5 and 2-6, Figure 2-13 to 2-16. The BTC for cosolvent flushing with toxaphene exhibited some asymmetrical properties and had some degree of “tailing” which is characteristic of non-equilibrium caused by sorption of toxaphene to the soil. In order to be sorbed, the solute molecules must be transported to the soil surfaces during solute flow through soils. Physical processes controlling sorption, diffusion-controlled access to the sites delay the solute molecules reaching the sorption sites and lead to sorption non-equilibrium. These processes include diffusion within sorbent organic matter such as stagnant-film diffusion, and retarded pore diffusion in microporous mineral grains such as aggregated soil forming mobile-immobile zones and some structured /fractured media. It is difficult to discriminate between these two mechanisms.

PAGE 55

41 Table 2-7. Parameter values for binary cosolvent system (v=31.33 cm/hr, D= 2.53 cm2/hr, L=5.00 cm, P= 61.92, T0=2.30) fc 30% 40% 50% 60% 75% R 12.7 4.9 3.55 2.4 1.49 0.5020.007 0.7010.013 0.6840.11 0.6310.17 0.7860.3 1.36.06 0.7060.087 0.81.083 0.9580.113 0.6270.2 Kp (cm3/g) 2.684 0.895 0.585 0.321 0.112 k2 (hr-1) 1.3470.093 3.0190.52 4.5240.594 6.7781.048 12.321.433 F 0.4590.008 0.6240.016 0.56.15 0.3670.27 0.3490.091 Table 2-8. Parameter values for ternary cosolvent system (50% methanol + pentanol, v=31.33 cm/hr, D= 2.53 cm2/hr, L=5.00 cm, P= 61.92, T0=2.30) fc 0% 5% 10% 25% R 3.55 1.94 1.46 1.16 0.6840.01 0.754.018 0.861.04 0.95.03 0.81.083 0.616.112 0.426.218 0.275.203 Kp (cm3/g) 0.585 0.216 0.106 0.037 k2 (hr-1) 4.524.594 8.088.922 13.153.318 29.709.7 F 0.560.014 0.4920.037 0.559.127 0.6377 Relationship between sorption constant and cosolvent fraction As methanol fraction increased, retardation of toxaphene elution decreased. This result can be attributed to the increase of toxaphene solubility and decrease of its sorption to soil. The relationship of toxaphene sorption to cosolvent fraction can be seen in the plot of logKp vs fc for methanol shown in Figure 2-17 and 2-18. The slope is – , where is a constant representing the solvent-sorbent interactions, and is a constant representing the water-methanol interactions. The value of for methanol was 3.43, obtained from the solubility test. From Figure 2-17, the slope for methanol was 2.907, thus, the value of was 0.83. This value means the solvent-sorbent interactions in this situation are not negligible because this value was lower than 1. This value was consistent with those reported by Brusseau et al.(1991) from anthracene and naphthalene sorption for methanol system. The value of Kp,w would be extrapolated from the plot as 16.71mL/g, which was comparable to the batch test.

PAGE 56

42 The Kp values obtained from the experimental column test were higher than those from batch test (Table 2-7). These results may be attributed to the different solution to soil contact time and ratio. The batch tests showed that the Kp value increased with the reduction of V:M value(Table 2-7). The column contained nearly 9.57 mL solution volume and 25.45 g soil, therefore the V:M was nearly 0.376. Also, as the solution flowed through the column, the contact between the soil and solution increased. This phenomenon suggested the application of batch test directly to the column test is questionable due to the uncertainties associated with analytical resolution and translatability to dynamic systems (Brusseau et al. 1990). Table 2-9. Comparison of the Kp value from column test and batch test batch test (V:M) column test 1 2 3 5 7 Kp 0.59 0.31 0.26 0.21 0.18 0.17 The plot of logKpfc for ternary system (50% methanol + pentanol) is presented in Figure 2-18. The relationship between LogKp,m and fc is presented as: 22221111,,loglogccwpmpffKK (2-35) where, 1 1 1 was for methanol, which could be obtained from slope of the logKp vs fc regression in binary system (methanol/water) as 2.907. The fc1 employed here was 0.5, thus the value of 1 1 1fc1 was 1.45. The slope of the plot (Figure2-18) was the value of –2 2 2, and the interception was logKp,w -1 1 1fc1. The value of the 2 2 2 for pentanol was 4.534, and the value of 2 2 from solubility test was 7.59. Therefore the value of solvent-sorbent interactions was only 0.62. Comparing the BTC for 50% methanol only, 60% methanol only , and 50%methanol+10% pentanol (Figure 2-15), and 50% methanol only, 75% methanol only , and 50%methanol+25% pentanol (Figure 2

PAGE 57

43 16), the removal of toxaphene by cosolvent flushing was much more effective with the addition of pentanol solution. This was also the result of reducing the retardation, which was attributed to the increase in solubility, shown in solubility test results and reduction of the sorption constant Kp, shown in batch sorption test results. Relationship between sorption constant and reverse first order rate constant The value of k2 and Kp were found to be inversely related (Table 2-5 and 2-6). The value of k2 was derived from Damkohlor number and the value of Kp was calculated from retardation factor. Augustijin et (1994) developed a relationship between logklpgKoc using data for various soil types: ocKklog97.053.3log (2-36) where, Koc is the equilibrium sorption coefficient for organic carbon and can be derived from Kp,w/foc. Augustijn (1993) adjusted this equation according to the “aging effects” on the mass transfer rates: ocKklog69.074.1log (2-37) Brusseau et al.(1991) studied the k2 and Kp relationship using Eustis fine sand and methanol solution with Naphthalene and Anthracene as solute, and obtained the regression relationship as: pKklog61.079.0log (2-38) Using the same column apparatus and many polynuclear aromatic hydrocarbons (PAHs) as solute, Bouchard (1998) developed the relationship of k2Kp as: pKklog91.047.0log (2-39)

PAGE 58

44 Based on these previous research results, the value of , which is the slope of logk2 – logKp plot, were in the range among 0.61 to 0.97. In this research, from the log k2 – logKp regression line of binary (Figure2-19) and ternary system (Figure 2-20), the values of were 0.70 and 0.68 for binary and ternary system respectively, and therefore consistent with the values found in the literatures. pKklog70.045.0log for methanol flushing (2-40) pKklog68.048.0log for pentanol flushing (2-41) The interception is the value of logk2,w, where k2,w is the value of reverse sorption rate constant in pure water system. From the extrapolation of the plot, the value of k2,w for toxaphene was around 3 hr-1. Relationship between reverse first order rate constant and cosolvent fraction A log-linear relationship was found to exist between reaction and cosolvent fraction. The logk2-fc plot for methanol and pentanol are presented in Figure 2-22 and can be expressed as: 41.005.2log2 cfk for methanol (2-42) 72.014.3log2 cfk for pentanol (2-43) The results indicate that the reverse rate constants, k2, increased with the increasing fraction of cosolvent solution and the addition of pentanol enhanced the increase of reverse rate constants. This phenomena may be attributed to the fact that the increasing cosolvent could lead to a larger diffusion coefficient, therefore, increase the first-order rate constant. The value of the slope of the logk2-fc plot comprises four terms: , the slope of the logk2-logKp plot, , the solvent-sorbent interaction constant, , the solvent-solute interaction constant, and , the cosolvency power. The higher molecular weight

PAGE 59

45 alcohol such as pentanol have higher rates of non-polarity, which results in higher cosolvency power for cosolvent solution and contribute to the higher value of for pentanol. Based on literature information, the relationship between log k2 –fc can be estimated. The value of 0.61 from report by Brusseau et al. (1991) was used through studying several HOCs in aquaeous system (equation 2-39). The cosolvency power for methanol was estimated as 3.43 and for pentanol to be 7.59 as discussed in the solubility section. The value of could be assumed to be 1 and the value for was estimated to be 0.9 according to report of Brusseau et al.(1991). Based on these factors, the value of the slope could be predicted as 2.10 for methanol and 4.40 for pentanol. The value of logKoc for toxaphene is quite different from 2.47 (U.S. EPA ,1981) to 5.00 (Wauchope et al, 1992). The value of Kp,w = Kocfoc and the foc for Eustis find sand is 0.74%, thus, the Kp,w for toxaphene can be estimated within the range from 1.42 to 470. According to the equation 2-39 (Brusseau et al., 1991), the value of logk2,w for toxaphene should be between -0.8 to 0.7. The plot of logk2fc from the methanol cosolvent flushing column test is presented in Figure 2-21. This experimental data was compared with the predicted values and shown in Table 2-10. The methanol cosolvent flushing experimental and theoretical data were found to be within 2.5% difference. However, the value of for pentanol flushing was lower than theoretically estimated values, which may be attributed to the fact that the solvent-sorbent interaction constant was lower than 0.9 and the solvent-solute interaction constant may not be 1 for pentanol flushing. Considering the effects of solvent-sorbent and solvent-solute interaction from the batch tests, the

PAGE 60

46 relationship for logk2,w seems to be predictable according to the general empirical relationship and literature data for toxaphene. Table 2-10. Predicted versus experimentally obtained regression constants for binary and ternary cosolvent system flushing predicted experimental predicted logk2,w experimental logk2,w Methanol 2.1 2.05 -0.8 to 0.8 -0.41 Pentanol 4.4 3.14 -0.8 to 0.8 0.72 Relationship between fraction of instantaneous domain and cosolvent fraction The parameter of F describes the distribution of sorption between instantaneous to total sorptive domains and can be calculated from the value of according to equation 2-31. The relationship of F and fc for methanol and pentanol are presented in Figure 2-23 and 2-24. The results indicated the F value would be reduced with the increase of methanol fraction among the range of the 40% to 75% of fraction, while the F value would not have great change for pentanol below the 25% fraction. Brusseau et al. (1991) observed the same phenomena for up to 20% cosolvent that the F value would decrease with increasing cosolvent fraction. Agustijn et al. (1994) proposed the equation for this trend: )2.0(8.0)01.0(cwwmfFFF for 0.2< fc <1 (2-44) where the superscripts w and m indicated the water and cosolvent mixture respectively. Using this equation and the experimental data of methanol flushing, the Fw was nearly 0.8. This phenomena may be explained by the swelling polymer property (Park, 1952). At the first stage of swelling, the addition of cosolvent solution to the soil enhances the thickness of the matrix more quickly than the increase of the surface, which results in the

PAGE 61

47 relative increase of thickness and reduction of surface area to volume ratio. Therefore, with the increasing of the cosolvent fraction in certain range, the F value was reduced with the decrease of instantaneous surface to internal matrix. However, the experiment results showed that under the fraction of 40% methanol, the F value was not decreased with the increase of fraction, which may be due to some possible critical or threshold level of cosolvent to show appreciable swelling effect (Brusseau et al., 1991). The ternary system of 0%-25% pentanol +50% methanol + water flushing did not exhibit this trend. That may because the higher fraction of cosolvent solution would cause the second stage of swelling where the increase of the area may equal or exceed that of the thickness of the matrix. Conclusions The solubility of toxaphene can be increased by the addition of cosolvent solutions. Aqueous solubility of toxaphene in different cosolvent solutions was expressed as the cosolvency power which was estimated by log-linear regression of solubility to cosolvent fractions. This cosolvency powers for toxaphene in water systems including the cosolvents methanol, ethanol and isopropanol gathered from experimental data were found to be within a range of 92%-104% of theoretically derived data. The result showed that higher carbon number alcohols had higher values of cosolvency powers for toxaphene in cosolvent solutions. Addition of higher molecular PMOS into CMOS to form ternary system enhanced the solubility compared to the CMOS binary system alone. This was attributed to the increase of the non-polarity of PMOS.

PAGE 62

48 Straight chain carbon alcohols exhibited higher level of toxaphene solubility than branched chain carbon alcohols. The sorption constant of toxaphene to the soil, Kp, reduced with increase of the alcohol molecular weight due to the increase of the aqueous solubility of toxaphene. It was found that sorption process fro the miscible displacement of toxaphene from soil was best described by a non-equilibrium first-order bi-continuum model. The higher the fraction of the cosolvent in solution, the greater the reduction of the retardation factor of toxaphene in soil is. The additional of higher molecular alcohols such as pentanol enhance the flushing toxaphene from soil. The experimental data mirrored values in the literature for the log-linear relationship of sorption constant Kp to the cosolvent fractions fc. The results also showed log-linear relationship of the reverse first-order rate k2 to the cosolvent fractions fc. The log-log linear relationship was found between the k2 value to the Kp value. Experimental values k2 compared favorably with theoretical value estimated from literature data and equations.

PAGE 63

49 24 hr48 hr8 weeks1 week0306090120150180timeconcetnration of toxaphene (mg/L) Figure 2-1. Change of toxaphene concentration with time

PAGE 64

50 MeOh -10-50510152025303500.10.20.30.40.50.60.70.8fc for MeOh% differentce log-linear Eqn.a Eqn.b Eqn.c Eqn.d EtOh -40-30-20-10010203000.10.20.30.40.50.60.70.8fc for EtOh% differentce log-linear Eqn.a Eqn.b Eqn.c Eqn.d Figure 2-2. Comparison of the experimental data of toxaphene solubility in methanol, ethanol, and IPA with estimation.

PAGE 65

51 IPA 010203040506000.10.20.30.40.50.60.70.8fc for IPA% differentce log-linear Eqn.a Eqn.b Eqn.c Eqn.d ProOH -20-100102030405000.10.20.30.40.50.60.70.8fc for ProOH% differentce log-linear Eqn.a Eqn.b Eqn.c Eqn.d Figure 2-2. Continued.

PAGE 66

52 Methanollog(S) = 3.43f + 0.43R2 = 0.99170123400.10.20.30.40.50.60.70.8f for methanollog[Solubility(ppm)] ethanol-waterlog(S) = 3.64f + 0.56R2 = 0.98370123400.10.20.30.40.50.60.70.8f for Ethanollog[Solubility(ppm) ] Figure 2-3. Cosolvency power from the log-linear regression of the solubility to cosolvent fractions for binary solutions (methanol-water, ethanol-water, propanol-water, IPA-water)

PAGE 67

53 IPA-waterlog(S) = 3.51f + 0.42R2 = 0.97790123400.10.20.30.40.50.60.70.8f for IPAlog[Solubility(ppm) ] Propanol-waterlog(S)= 3.91f + 0.46R2 = 0.97970123400.10.20.30.40.50.60.70.8f for Propanollog[Solubility(ppm) ] Figure 2-3. Continued.

PAGE 68

54 Butanol-methanol-waterlog (S) = 5.94 f + 2.18R2 = 0.97300.511.522.533.544.500.050.10.150.20.250.3f for butanollog(S) Isobutanol-methanol-waterlog (S) = 4.56 f+ 2.19R2 = 0.959800.511.522.533.5400.050.10.150.20.250.3f of isobutanollog(S) Figure 2-4. Cosolvency power from the log-linear regression of the solubility to cosolvent fraction for ternary solutions (butanol-methanol-water, isobutanol-methanol-water, pentanol-methanol-water, and hexanol-methanol-water)

PAGE 69

55 Pentanol-methanol-waterlog (S) = 7.59 f + 2.21R2 = 0.983800.511.522.533.544.5500.050.10.150.20.250.3f of pentanollog(S) Hexanol-methanol-waterlog (S) = 8.74 f + 2.23R2 = 0.978800.511.522.533.544.500.050.10.150.20.25f of hexanollog(S) Figure 2-4. Continued.

PAGE 70

56 0123456789MeOHEtOHIPAProOHIBABtOHPtOHHeOHCosolvency Power Figure 2-5. Comparison of the cosolvency powers of different types of alcohol

PAGE 71

57 01234567891001020304dielectric constantcosolvency power 0 Figure 2-6. Relationship of cosolvency power with the dielectric constant 02040608010012002468Solubility (mg/L)cosolvency power 10 Figure 2-7. Relationship of cosolvency power with the solubility of the POMSs

PAGE 72

58 05101520250204060time(hr)S(mg/L) 80 Figure 2-8. The change of toxaphene concentration with time for batch test 0510152025Ce (ug/mL)S (mg/mL) Figure 2-9. Linear regression of S-Ce for 50% methanol binary cosolvent solution for batch test

PAGE 73

59 00.050.10.150.20.250.30.3512357V:M (mL/g)Kp Figure 2-10. The effects of ratio of solution: soil on the sorption constant Kp

PAGE 74

60 Figure 2-11. Batch test of log linear regression of partitioning coefficient Kp to cosolvent fraction in cosolvent solutions.

PAGE 75

61 Figure 2-11. Continued.

PAGE 76

62 Figure 2-11. Continued.

PAGE 77

63 Figure 2-12. IPA tracer breakthrough curve.

PAGE 78

64 Figure 2-13. Breakthrough curves for toxaphene using binary cosolvent system (methanol solution)

PAGE 79

65 Figure 2-14. Breakthrough curves for toxaphene using ternary cosolvent system ( 50% methanol + pentaol solution)

PAGE 80

66 Figure 2-15. Comparison of breakthrough curves for toxaphene using binary (50% or 60% methanol) and ternary cosolvent system (50% methanol+ 10% pentaol solution)

PAGE 81

67 Figure 2-16 Comparison of breakthrough curves for toxaphene using binary (50% or 75% methanol) and ternary cosolvent system (50% methanol+ 25% pentaol solution)

PAGE 82

68 Figure 2-17. LogKp-fc for binary system (methanol solution) Figure 2-18. LogKp-fc for ternary system (50% methanol +Pentanol solution)

PAGE 83

69 Figure 2-19. Log k2 – logKp for binary system (methanol solution) Figure 2-20. Log k2 – logKp for ternary system (50% methanol +Pentanol solution)

PAGE 84

70 Figure 2-21. Log k2 – fc for binary system (methanol solution) Figure 2-22. Log k2 – fc for ternary system (50% methanol +Pentanol solution)

PAGE 85

71 Figure 2-23. F – fc for binary system (methanol solution) Figure 2-24. F– fc for ternary system (50% methanol +Pentanol solution)

PAGE 86

CHAPTER 3 DECHLORINATION TECHNOLOGIES OF TOXAPHENE Introduction Cosolvent flushing provides a technology for mobilization of toxaphene from the subsurface of a contaminated field site. However, during an in situ flushing process, the structure of toxaphene is not transformed. Remediation technologies of chlorinated contaminants, especially DNAPLs, include processes such as dissolution, displacement or volatilization, as well as those that destroy the contamination through chemical reaction or biodegradation. The possible processes for transformation of chlorinated compounds include: (1) substitution (hydrolysis); (2) dehydrochlorination; (3) reductive dechlorination such as hydrogenolysis, dichloro-elimination, and coupling; (4) oxidation, such as hydroxylation, halosyl oxidation, epoxidation, and bihalogenation. Research showed that the dechlorination of toxaphene is through reductive dechlorination (Parr et al., 1976; Gillham et al., 1994 and ; Clark et al.,2003). The word “degradation” was used here in the broad sense, that is, to describe any measurable change in the pesticide. This would either partial degradation to other organic intermediate or complete degradation to inorganic end products (Parr et al., 1976). In this research, certain remediation technologies that are more likely to enhance the dechlorination of toxaphene contamination were examined including zero-valent iron treatment, biodegradation, and combination of these methods with cosolvent flushing processes. 72

PAGE 87

73 Iron Treatment Iron treatment has been used widely in the reductive dechlorination of many chlorinated contaminants (Gillham et al., 1994; and Clark et al., 2003). Many studies have utilized iron treatment as field scale test with permeable reactive barriers (PRB) or with pump-and-treat technologies, suggested that iron treatment could be an effective method in chlorinated contaminant degradation (Gillham et al., 1994). EnviroMetal Technologies, Inc. (ETI) applied a metal-enhanced dechlorination technology to remediate chlorinated volatile organic compounds (VOCs) such as chlorinated methanes, ethanes, and ethenes in aqueous media. An in-situ application of the technology was demonstrated under the U.S. Environmental Protection Agency’s (U.S. EPA) Super-fund Innovative Technology Evaluation (SITE) Program at a confidential site in central New York state (U.S. EPA, 1998). Some primary objectives of the SITE demonstration included : determination of whether treated groundwater from the in-situ, permeable treatment wall meets groundwater standard for the critical contaminants including PCE, TCE, l,l, l-trichloroethane (TCA), cDCE, tDCE, and VC; and determine the removal efficiency (RE) of critical contaminants from groundwater. Some secondary objectives include: determine concentration gradients of critical contaminants; examine total metals, chloride, sulfate, nitrate, bicarbonate, and non-critical VOC concentrations; document geochemical conditions (specific conductance, Eh, pH, DO, and temperature; Examine biological microorganism growth in the reactive iron medium and in upgradient and downgradient groundwater; and document operating and design parameters (initial weight, volume, and density of the reactive iron medium, groundwater flow velocity).

PAGE 88

74 The result of SITE demonstration showed that average critical contaminant concentrations for the down gradient well were all below the target standards and the minimum overall average REs were high for all critical parameters. The total costs were estimated to be about $18 per 1,000 gallons of groundwater treated for a continuous wall, and $20 per 1,000 gallons treated for a full-scale funnel and gate system. However, little research has been conducted on application of iron treatment to the toxaphene removal (Clark et al., 2005). Biodegradation Since 1991, the U.S. EPA Environmental Response Team Center (ERTC) and Response Engineering Analytical Contact (REAC) have used anaerobic bioremediation technology to conduct many studies on toxaphene removal. Their medium development studies in bench-scale reactors verified that blood meal can promote the degradation of toxaphene under anaerobic conditions. Field-scale studies were eventually carried out at four places: Navajo Vats, Sanderes Aviation, Ojo Caliente Dip Vat and Gila River Indian Community (GRIC) and presented by Allen et al. (2002). Comparison of the degradation rate of these sites is presented in Table 3-1. In a site of Navajo Vats in Nazlini and Whippoorwill (Allen et al., 2002), four 1228-liter plastic tanks of pilot reactors with 1591kg of contaminated soil were buried in-ground. Full-scale studies were conducted at 22 additional sites with dug pits, which were covered to allow the venting of gases. During full-scale studies, stockpiled contaminated soil, water and nutrient supplements (sodium phosphate, sheep manure, limestone and blood meal at a rate of approximately 10% wt/wt) were added to the test pits and slurried in a cement mixer. Blood Meal is an all natural organic fertilizer that is a fast source of organic nitrogen. At site of Sanderes Aviation in Tempe, Arizona (Allen

PAGE 89

75 et al., 2002), three 1.5 meters square by 0.46 meters deep pits were constructed and lined with plastic liners. A mass of 818 to 1045 kg of contaminated soil, 378 liters of water and nutrient supplements (sodium phosphate, 1% wt/wt; limestone, 5% wt/wt; and blood meal, 1% or 2.5% wt/wt) were mixed in a cement mixer and added to the pits. At the site of Ojo Caliente Dip Vat , Zuni reservation near Gallup, New Mexico (Allen et al., 2002), 18 meter long, 5.4 meter wide, and 1.5 meters deep pits were constructed and lined with a doubled 6-mil plastic liner. In the pits, 146 cubic meters of contaminated soil and nutrient supplements (sodium phosphate, 1% wt/wt; limestone, 2.5% wt/wt; and blood meal, 2.5% wt/wt) were added in layers and hydrated with water and then loaded to the pits. For Gila River Indian Community (GRIC) field site, using similar methodology as that seen in the Ojo Caliente Dip Vats, the results were collected in Table 3-1. Table 3-1. Results from field scale study of toxaphene degradation. Site Name Init. conc. (ppm) Final conc. (ppm) Time (days) Degradation (%) Nazlini 291 71 108 75.6 Whippoorill 40 17 110 57.5 Blue Canyon Rd. 100 17 106 83 Jeddito Island 22 3 76 76.5 N avajo Vats Site Poverty Tank 33 8 345 75.8 Ojo Caliente 14 4 14 71.4 Laahty Family 29 4 31 86 Ojo Caliente Dip Vat Site Henry O 23 8 76 68

PAGE 90

76 Table 3-1. Continued. Site Name Init. conc. (ppm) Final conc. (ppm) Time (days) Degradation (%) Unit 1 59 4 272 93.5 Unit 2 31 4 272 86.6 Unit 3 29 2 272 94.4 Gila River Indian Community Site Unit 4 211 3 272 98.4 Gray et al. (2000) reported on the use of the XenoremTM process of the Stauffer Management Company (SMC) at these two sites. The U.S. EPA supported this bioremediation process in the site Record of Decision (ROD) as the preferred technology to treat the pesticide contaminated soil at Tampa sites in Florida (Gray et al., 2000). Initially, random samples were collected from two highly contaminated (hot) zones at the Tampa based sites. The microcosms were carried out over a 6-week period, representing one anaerobic/ aerobic cycle (4 weeks/2 weeks). The results demonstrated the potential of the enhanced anaerobic/aerobic composting technology to degrade toxaphene. Toxaphene was analyzed by two methods : (1) the degradation of the technical toxaphene and its metabolic breakdown products; (2) only the technical toxaphene. The results showed that the degradation of toxaphene and metabolic by-products has rapid degradation rate, giving rise to an overall loss greater than 90%. Finally, a total of 4000 yd3 of soil was treated as a full-scale operation in a similar fashion as the demonstration study and demonstrated greater than 90% removal of toxaphene from soil after treatment with the XenoremTM technology. In the case of the full-scale operation, 95% toxaphene removal was achieved after 17 weeks. This demonstrated that the process could remediate 1000 yd3 demonstration and 4000 yd3 full-scale soil to acceptable levels.

PAGE 91

77 Comparison of Dechlorination Technologies Technology Advantages Disadvantages Zero-valent iron enhanced dechlorination Relative high degradation rate (needs hours or days); Generate no air emissions and no secondary waste; no chemical (such as O3 or H2O2) required; maintenance and operation passively. potential for reactivity of iron and formation of by-products; Potential for gradual loss of hydraulic conductivity ; geologic conditions may preclude the use of PRB Biodegradation Low air emissions; can be constructed without obtrusive surface structures Slow rates of removal (needs weeks or months); Sludge treatment and disposal required; maintenance required Zero-Valent Iron Treatment Background Iron treatment has been successful in treating chlorinated organics (Clark et al. 2003). The Redox couple formed by zero oxidation state metallic iron has a standard reduction potential of -0.440 V (equation 3-1). The estimated standard reduction potentials for the half-reaction for various alkyl halides (equation 3-2) is from +0.5 to +1.5V at pH=7. Thus, the net reductive dechlorination by iron (equation 3-3) is favorable thermodynamically. 022FeeFe (3-1)

PAGE 92

78 XRHHeRX2 (3-2) XRHFeHRXFe20 (3-3) The corrosion of iron under aerobic and anaerobic conditions is quite different. Equations 3-4 and 3-5 are the rapid corrosion with dissolved oxygen presence, while equations 3-6 and 3-7 are under anaerobic conditions. OHeOHO44222 (3-4) OHFeOHOFe42222220 (3-5) OHHeOH22222 (3-6) OHHFeOHFe222220 (3-7) The formation of iron hydroxide precipitation may produce a surface layer and inhibit the further dissolution. Thus, the rapidly corrosion under aerobic condition should be avoided. Matheson (1994) provided three possible schemes of pathways for the reductive dechlorination under anoxic Fe0-H2O systems (Figure 3-1) (Matheson et al., 1994). (A) The first pathway is the direct electron transfer from iron metal at the metal surface. 2Fe0 2Fe2+ + 4e 3H2O 3H+ + 3OH2H+ + 2e H2 X-Cl + H+ + 2e X-H +Cl2Fe0 + 3H2O + X-Cl 2Fe2+ + 3OH+ H2 + X-H + Cl(3-8)

PAGE 93

79 (B) The second pathway is through the reduction by Fe2+ which results from corrosion of the metal. XRHFeHRXFe3222 (3-9) (C) The third pathway involves the hydrogen produced as a product of corrosion with water. XHRHRXH2 (3-10) Because the H2 is not easy to reduce without an effective catalyst, this reaction is not highly possible. Matheson (1994) performed a variety of control experiments and treatment studies to identify the dominant process pathway under anaerobic condition. From the results, neither H2-saturated water nor 5-100mg/L FeCl2 produced measurable dechlorination over 15 days in the absence of the metal. Amendment of Fe0 with additional Fe2+ or H2 did not affect the dechlorination rate greatly, and the addition of 0.5mM EDTA, which can form Redox-inactive complex with Fe2+, did not exhibit obvious effect. Considering all of these, the pathway (A) is the most important contributor to the dechlorination process. Various researchers have studied the kinetics surrounding the use of iron for treatment of chlorinated organics (Gillham and O’Hannesin, 1994; Clark et al., 2003). These relationships are presented below. The psuedofirst order relationship has been found to describe Fe0 treatment of some chlorinated organics (Gillham and O’Hannesin, 1994; Clark et al., 2003). This pseudo first-order relationship is:

PAGE 94

80 tkCCww10)ln( (3-11) The slope of the pseudo first-order relationship was defined by: isissaKKkk 11 (3-12) This relationship was found to be present in PCE degradation in contact with iron (Loraine, 2001; Clark et al., 2003). Johnson (1996) summarized kinetic data from 18 different batch and column experiments of some simple chlorinated compound degradation by iron metal and found out that the range in reported degradation rate for batch experiments corresponds to t1/2 values from 10min to 30 days with a typical value of roughly 15hrs, while the t1/2 from column studies range from 30s to 30hrs with a typical value of roughly 1.5hrs. That means first-order rate constants (k1) from both batch and column studies vary widely and without meaningful correlation. However, the surface area-normalized constants (ksa), which were obtained by normalization of these data to iron surface area concentration, yields a specific rate constant that varies by only 1 order of magnitude for individual chlorinated compound (Johnson, 1996). In order to quantify the effect of variation in the specific reactivity of the metal surface, Johnson (1996) also formulated an empirical second-order kinetic model in terms of reactive surface sites wsswCCkdtdC2 (3-13) This formulation is most appropriate at distinct sites such as impurities, pits, kinks, steps, other surface defects, or crystal faces.

PAGE 95

81 Many other kinetic models have been formulated for the chlorinated compounds degradation in the iron surface. Using the assumption that partitioning equilibrium conditions were reached and the rate of the surface reaction was the rate-controlling step, Loraine (2001) set up a model: wawawCKCkKdtdC1 (3-14) where, k is the rate of reaction per surface area and Ka is the absorption equilibrium constant per gram iron. Johnson (1996) took the kinetics of surface association and effect of site saturation into account and formulated a kinetic model for a steady-state concentration of the surface complex: wwmwCKCVdtdC2/1 (3-15) where, Vm = krCss and K1/2=(kb+kr)/kf, kr is the rate constant for reduction of the surface complex, and kf is the forward rate constant association with the surface by diffusion, adsorption, or surface complexation , and kb is the backward disassociation from the surface by diffusion or desorption. Vm can be regarded as the maximum reaction rate for a particular type and amount of iron metal, and K1/2 is the concentration of P at Vm/2. The objectives of this research include: (1) apply the iron treatment technologies to the toxaphene and explore the degradation rate; (2) study the effects of bimetallic species on toxaphene degradation.

PAGE 96

82 Methodology To evaluate dechlorination efficiency of toxaphene in aqueous solutions contacted with zero-valent iron, individually sealed batch reactor experiments were performed and analyzed over a one week intervals. A group of experiments were conducted with solutions containing 1 or 2.5 ppm of toxaphene contacted with zero-valent iron and bimetallic substrates containing zero-valent iron including Fe0, nickel-plated zero valent iron (NiFe0), and copper plated zero-valent iron (CuFe0). Toxaphene was obtained in high purity from AccuStandard and was used as received, in concentration of 1000 ppm in methanol. The Fe0 substrate used in this study was obtained from Fischer Scientific and the NiFe0 and CuFe0 substrates were produced by plating copper and nickel electrochemically onto Fe0. The electrochemical plating was conducted in the laboratory of Egwu Kalu at FAMU/FSU college of Engineering and followed the methodology found in Kalu(1998) and Kuruganti et al.(2001). Each batch reactor was prepared in a 5mL glass vial fitted with a Teflon-lined septa screw-top cap. Initially the vials along with the caps were weighed empty, and following the addition of iron substrates, the entire vials were weighed. The vials were then filled with aqueous toxaphene solutions allowing no headspace at the top and reweighed. A 4:1 volume/mass ratio which was used by Gillham and O’Hannesin (1994) and Muftikian et al. (1995) was adopted for the batch tests in this research. A mass of 1.25 g of substance was added to each of 5mL vials. In order to study the influence of the amount of substrate on the degradation rate, the mass of Fe0 was raised to 2.5g, which reduced the volume-to-mass ratio to 2:1.

PAGE 97

83 The vials were then placed on a rotator (40 revolutions per minute) for the desired sampling times. The samples were then analyzed after hexane extraction. The experiment was run in triplicates for the aqueous sorbed phase. The remaining aqueous volume above the iron was removed and the vial reweighed. For the determination of mass of toxaphene on the iron, toxaphene was desorbed from the iron and analyzed separately. Toxaphene was desorbed by adding methanol (MeOH) to the vial after removing the aqueous solution. The vials were kept on a rotator for 12 hrs and then weighed and the contents representing the sorbed phase concentration transferred to 0.5mL vial insert for analysis. The hexane extractions for the aqueous phase were conducted as they were analyzed by a GC/ECD as discussed in Chapter 2. Results and Discussion The results of the batch experiments showed that toxaphene degraded when in contacted with zero-valent iron and by bimetallic substrates containing zero valent iron. The degradation seemed to display a power law function in the decrease of toxaphene concentration (C/C0) as a function of time, t(h), (Clark et al., 2005): 10kaatCC (3-16) A linear function can be produced plotted on a natural log-natural log system: tkCCaalnln10 (3-17) The experimental results allowed for calculation of the k1 value for toxaphene degradation(Table 3-2). The half-lives for toxaphene in these batch tests were from 7 hrs to 120 days, compared with the aerobic degradation of toxaphene in soil with a half-life

PAGE 98

84 of up to 14 years (Nash and Woolson, 1967) and anaerobic biodegradation technology with half-lives between one to several months (Fingerling et al., 1996). Table 3-2. Toxaphene dechlorination by various substrates containing iron at different volume-to-mass ratios and initial concentrations Substrate V:M initial Conc.(mg/L) -k1 R2 Fe 4:1 1 0.088 0.94 Fe 2:1 1 0.113 0.96 Fe 4:1 2.5 0.061 0.89 Cu-Fe 4:1 1 0.245 0.94 Ni-Fe 4:1 1 0.219 0.93 The effects of the amount of iron substrate and the initial toxaphene concentration on the degradation of toxaphene were also explored. All the results also fit to the power law relationship (equation 3-17). The reduction of the volume-to-mass ratio to 2:1 by doubling the mass of iron to 2.5 grams in the system increased the degradation rate k1 by 29% compared to that of the 4:1 volume-to-mass ratio experimental runs (Table 3-2). These results corresponded with the comparable research of PCE and TCE degradation with iron treatment (Loraine, 2001; Clark et al., 2003) and seemed to indicate that greater substrate mass included provided a higher rate of toxaphene dechlorination as a consequence of the increase of the reactive sites in the system. The increased initial concentration to 2.5ppm resulted in 30% lower degradation rate k1 than that of 1ppm at the same volume-to-mass ratio system (Table 3-2). This outcome was somewhat expected due to the competition for adsorption sites on the iron substrate by the higher concentration of toxaphene in solution, and agreed with the similar data from dechlorination of PCE or TCE by iron (Clark et al., 2003). These effects suggested that the sorption of toxaphene to the surface of the iron is the key controlling step for toxaphene degradation.

PAGE 99

85 The experiments with bimetallic substrates denoted the enhancement of degradation rate by nickel-plated and copper-plated zero-valent iron, which increased the degradation rate k1 by 150% and 180%, respectively. This enhancement of degradation paralleled earlier results of PCE and TCE treatment by bimetallic substrates (Clark et al., 2003, Muftikian et al., 1995, and Liang et al., 1997), which is attributed to the bimetallic substrates’ ability to facilitate electron transfer in the reduced environment (Korte et al., 1997) and reduce the formation of oxides on the substrate surfaces (Wang et al., 1997). The change of the chromatograms peak distribution indicated that as contact time increased, overall peak area of toxaphene decreased, meanwhile, later eluting peaks decreased or disappeared (Figure 3-2). Howdeshell et al. (1996) suggested that the higher molecular weight component of toxaphene might be dechlorinated into lower weight components, which would cause the decrease of the retention time on the chromatograms. In addition, the bimetallic substrates displayed smaller peak area on the chromatograms after one week contact, which indicated more aggressive degradation of toxaphene by bimetallic substrates (Figure 3-3). Chloride concentration in solution was found to increase over experimental run time, and a chloride mass balance showed that from 5.8 to 8.4 mole/mL of chloride was produced for 1mole/mL toxaphene degradation. Dechlorination is also presented to increase over time for each of the iron substrates until equilibrium was reached (Figure 3-4).

PAGE 100

86 Effects of Cosolvents on Dechlorination of Toxaphene by Iron Background As discussed in Chapter 2, cosolvent flushing provides rapid mass removal of DNAPLs chlorinated contaminants by dissolution and/or mobilization. Treatment of this waste effluent by Fe0 seems to be an option for post in situ flushing. However, the ability of toxaphene to be adsorbed to the iron surface will be reduced by presence of cosolvents (Clark et al., 2003), which may cause a decrease in the degradation rate. Therefore, the level to which the cosolvent fractions will affect not only the sorption of toxaphene to the iron but also its degradation by the iron requires further exploration. Loraine (2001) studied the reduction of TCE and PCE by iron in DI water, MeOH (57% v/v), IPA (57% v/v), and EtOH (28, 57, and 100% v/v) and none of the alcoholic solutions improved the removal rate of either TCE or PCE. However, degradation of TCE was not significantly inhibited at the relatively low concentration of ethanol (28%), therefore, the application of combining the cosolvent flushing and zero-valent iron might have been practical if TCE was efficiently solubilized. Clark et al. (2003) used ethanol (10, 30, and 50% v/v) to test the degradation of PCE on the surface of untreated iron, nickel iron and treated iron. This degradation displayed a pseudo first-order kinetic relationship. For untreated iron, with low concentration of 2 g/mL of PCE, the pseudo first-order rate constant reduced from 3.74 to 3.52 hr-1 when the partition of ethanol increased from 0%-10%, while this constant reduced quickly to 1.36 hr-1 when the ethanol increased to 30%. The results were similar to the study of Loraine (2001) and implied a direct correlation between the degradation rate and the volume of solution per reactive site in a given mass of iron.

PAGE 101

87 The objectives of this research include: (1) examine the effects of various types and fractions of cosolvent solutions on dechlorination of toxaphene by Fe0; (2) examine the sorption isotherm of toxaphene on iron surface and how it changes with cosolvent in solution; (3) set up a kinetic model to describe the process. Methodology Toxaphene was purchased from UltraSci (Lot 302-1258). Solid toxaphene was dissolved in methanol solution resulting in a solution with a toxaphene concentration of 238.4 g/mL. This solution was used to prepare required concentrations from 2 g/mL to 100 g/mL. The type and fraction of cosolvent and toxaphene initial concentration were applied as followed: 10% methanol with toxaphene initial concentration 2g/mL, 5g/mL; 20% methanol with 2g/mL, 10g/mL; 30% methanol with 2,5,15g/mL; 40% methanol with 2,15,20g/mL; 50% methanol with 2,5,10,15,20,50,100g/mL; 60% methanol with 2,15g/mL; 75% methanol with 2,15,100g/mL; 100% methanol with 15g/mL(Table 3-3). Water with toxaphene initial concentration of 2g/mL and 50% ethanol and 50% IPA with toxaphene initial concentration of 15g/mL respectively were also chosen to compare (Table 3-3). The sampling intervals were 0, 1, 3, 8, 18, 42, 72, 168 hrs. Table 3-3. The chosen initial concentrations and fractions of cosolvent solutions 2mg/L 5 mg/L 10 mg/L 15 mg/L 20 mg/L 50 mg/L 100mg/L 0% Y 10% Y Y 20% Y Y 30% Y Y Y

PAGE 102

88 Table 3-3. Continued. 2mg/L 5 mg/L 10 mg/L 15 mg/L 20 mg/L 50 mg/L 100mg/L 40% Y Y Y 50% Y Y Y Y Y Y Y 60% Y Y 75% Y Y Y 100% Y EPA50% Y IPA50% Y * Y means this initial concentration and fraction of methanol was chosen The mass of iron used was 1.25g contacted with a 5 mL volume of solution. The toxaphene was desorbed from the iron by adding 100% methanol and analyzed separately to determine the mass sorbed to the iron. The weight of the vial + iron, w1, the total weight of vial+iron+solution in the system, w2, and that of vial+iron after remove of solution, w3 were measured during the experiment. Toxaphene in the system was divided into two parts, that is the concentration of toxaphene in aqueous, Ca(g/mL), and the concentration of toxaphene adsorded by iron, Cs(g/mL). aleftaleftextextsVVCVVCC )( (3-18) where, Vext was the methanol volume used to extract toxaphene from the iron (mL). Cext was the concentration of toxaphene in the extraction methonal solution (g/mL). Va (mL) was the total volume of the solution in the system: ccaffWWV112 (3-19) was the density of methanol, fc was the methanol fraction.

PAGE 103

89 Because the process of removal of solution from the iron couldn’t separate the solution and the solid completely, there still was a residual volume of solution left in the vials , Vleft (mL) , before the process of extraction by methanol. ccleftffWWV113 (3-20) CT is the total toxaphene concentration in the system (g/mL) CT =Ca+Cs (3-21) Results and Discussion Toxaphene dechlorination The experimental results showing the decrease in toxaphene in contact with Fe0 is presented in Figure 3-5. Applying the power law function of equation 3-17 (Clark et al., 2005), the decholorination results for different volume fraction of cosolvent are presented in Figure 3-6 and Table 3-4. The results suggested that the adsorption of contaminants as well as the degradation rate were reduced with the increase of cosolvent fraction. However, the reduction of degradation rate in some low fractions of cosolvent with the high concentration of toxaphene solution is negligible (from 0.094 in 0% methanol with 2g/mL initial concentration to 0.069 in 30%methanol with 15g/mL initial concentration, Table 3-4). Pertaining to cosolvents’ ability to increase aqueous solubility results indicated that cosolvents had a definite effect on toxaphene dechlorination. Table 3-4. Toxaphene degradation rate of power law function kinetic by zero-valent iron at different fraction of cosolvent solution and initial concentration 2ppm 5ppm 0% 10% 20% 30% 40% 50% 60% 75% 10% 30% 50% k1(*10-2) 9.38 8.59 7.68 6.88 5.93 4.58 3.03 1.27 8.31 6.12 4.22 r2 0.948 0.948 0.885 0.939 0.886 0.859 0.807 0.725 0.909 0.847 0.88

PAGE 104

90 Table 3-4. Continued. 10ppm 15ppm 20ppm 50ppm 20% 50% 30% 40% 50% 60% 75% 100% 40% 50% 50% k1(*10-2) 8.88 4.04 6.9 5.24 3.46 2.04 1.18 0.71 5.29 3.72 3.42 r2 0.969 0.771 0.949 0.974 0.98 0.836 0.883 0.857 0.919 0.768 0.852 100ppm 15ppm 50% 75% E30% I30% k1(*10-2) 2.71 0.71 7.31 7.06 r2 0.793 0.857 0.948 0.963 Sorption isotherms The sorption of toxaphene to iron surface is shown in Figure 3-6 and Table 3-5 Table 3-5. Values of Kf and Nf in Frendliuch sorption isotherm and Ca0/Sc Ct0 5 5 10 10 15 15 15 15 20 20 fraction 10% 30% 20% 50% 30% 40% 50% 60% 40% 50% lnKf -0.93 -1.52 -1.18 -2.02 -1.48 -1.75 -2.37 -2.51 -1.90 -2.19 Kf 0.39 0.22 0.31 0.13 0.23 0.17 0.09 0.08 0.15 0.11 Nf 1.00 0.88 0.99 0.79 0.96 0.95 0.85 0.70 1.00 0.89 Ca0/Sc 0.74 0.15 0.67 0.06 0.45 0.20 0.09 0.04 0.27 0.12 Ct0 50 100 100 15 15 fraction 50% 50% 75% 30% 30% lnKf -2.32 -2.26 -2.56 -1.67 -1.73 Kf 0.10 0.10 0.08 0.19 0.18 Nf 1.01 0.99 0.86 0.98 0.93 Ca0/Sc 0.30 0.59 0.08 0.42 0.33 From the results, the sorption isotherm of toxaphene to iron was not always linear. In order to study the linearity of sorption, the nonlinear Freundlich sorption equation was applied: nafsCKC (3-22) The results showed that the linearity was related to the saturation of solution (Ca/Sc), where Sc represented the solubility of toxaphene in solution (changed with

PAGE 105

91 different fc). Chapter 2 provided the solubility of toxaphene in different fraction and type of cosolvent solutions. Enhancement of cosolvent solubilization is described quantitatively using the cosolvency power. The value of cosolvency power for the methanol is between 3.31 (Morris, 1988) and 3.43 (Figure 2-3). The value of Sc increased with the increase of cosolvent fraction, while Ca/Sc decreased. From Table 3-5 and Figure 3-6, sorption isotherm displayed nonlinear characteristics when the value of Ca/Sc was in the range of 0.04 to 0.20. The sorption was almost linear for higher Ca/Sc value (>0.2). This phenomenon was consistent with observation of previous research (Bouchard, 2002 and Xia, 2001). Under the same initial concentration of toxaphene in solution, the linearity of sorption isotherm was reduced with the increase of cosolvent fraction. This nonlinearity may be attributed to the decrease in the marginal adsorption energy with increasing related surface concentration saturation, which is lower in high cosolvent fraction. Cosolvency power It was found that the presence of cosolvent reduced toxaphene sorption on the iron surface. Equation 2-18 was applied to analyze the cosolvent power (only the Kf values with linearity n=1 were chosen here). 0,,loglogdcndKfK (3-23) Analysis of the values acquired among the 0.1 to 0.5 methanol fraction showed that cosolvent power for toxaphene in these solutions is 3.45 and the Kd,0 ( adsorption rate in 0% colvent solution) is 0.58 from extrapolation. (Figure 3-7). The cosolvent power was consistent with that from solubility test.

PAGE 106

92 The result indicated that adsorption was hindered by increasing methanol fractions, which may contribute to the decrease of degradation rate (Figure 3-5). This phenomenon suggested that sorption to the iron surface is the major and limiting factor for toxaphene degradation, which was also observed in previous PCE and TCE degradation with iron treatment ( Clark, 2003 and Loraine, 2001). Kinetic analysis of the toxaphene degradation process The nonlinearity of experimental data describing toxaphene degradation indicated that the reaction is not always pseudo first order. Since the reaction mechanisms involved in toxaphene degradation with iron treatment are complex, nT was employed as the reaction order when only the total toxaphene concentration was considered. Burris (1995) used equation 3-24 to analyzed the PCE and TCE degradation rate with iron treatment: TnTTTCkdtdC (3-24) Integrating equation 3-24 and employing initial total toxaphene concentration CT0 yields: TTnnTTTTCtknC11)1(0])1[( (3-25) The nonlinear regression results of nT and kT are presented in Table 3-6. Table 3-6. Kinetic analysis of toxaphene degradation and sortion C0(g/mL) fc nT kT(10-6) ln(k') K' n' ka(10-3) Ns 5 10% 5.89 1.91 1.27 3.56 0.99 3.43 5.834 30% 6.44 1.24 1.837 6.28 0.99 175 6.376 10 20% 5.39 7.1 1.476 4.38 0.981 2.02 5.288 50% 4.612 8.96 2.614 13.65 1.207 154 5.567

PAGE 107

93 Table 3-6. Continued. C0(g/mL) fc nT kT(10-6) ln(k') K' n' ka(10-3) Ns 15 30% 4.84 7.89 1.803 6.07 0.946 4.9 4.582 40% 4.094 8.45 2.048 7.75 0.956 3.7 3.914 50% 4.081 6.8 2.767 15.91 1.1 54.5 4.489 60% 3.331 18.49 3.405 30.11 1.063 156 3.541 20 40% 3.834 7.7 2.064 7.88 0.962 2.11 3.689 50% 3.684 7.89 2.555 12.87 1.053 9.66 3.879 50 50% 2.83 12.5 2.451 11.6 0.961 1.29 2.72 100 50% 2.7 8.45 2.468 11.8 0.968 0.663 2.614 75% 1.84 106.4 3.199 24.51 1.036 3.83 1.906 From the analysis of the cosolvent effect on toxaphene degradation, it seems that the decrease of the adsorption also reduced the toxaphene degradation rate. It is reasonable to assume that the degradation of toxaphene would dominate on the solid phase, as stipulated by previous research (Clark et al., 2003). Considering the change of Cs, equation 3-26 can be applied to include both the degradation and desorption: asnsssCCCkdtdCs21 (3-26) The rate of change of toxaphene concentration in aqueous phase is due to the transport process to the solid surface. asaCCdtdC21 3-27 Combining equation 3-26 and 3-27 results in the following equation: dtdCCkdtdCanssss (3-28) Since CT = Ca+Cs, equation 3-28 can be expressed as: snssTCkdtdC (3-29) Applying reaction equation 3-25 and 3-28 to study the sorption isotherm:

PAGE 108

94 sTnssnTTTCkCkdtdC (3-30) Equation 3-30 can be transformed to : nsTCkC (3-31) Where, TnTskkk1)( (3-32) Tsnnn (3-33) Using linear regression to derive the values of k and n: )ln(ln)ln(kCnCssT (3-34) The value of ks and ns can be derived from equations 3-21 and 3-30. The results of k, n, ks and ns are presented in Table 3-6. The sorption isotherm equation with coefficients derived from kinetic data could be expressed as: sNNsnTsaCCkkCTsT1)( (3-35) From equation 3-21 and 3-35, nnnnTTTskCtknCTT1)1(1)1(0)(])1[( (3-36) Thus nnnnTTTnnTTTakCtknCtknCTTTT1)1(1)1(011)1(0)(])1[(])1[( (3-37)

PAGE 109

95 A comparison of this equation with the experimental data of toxaphene aqueous concentration is presented in Figure 3-8. The agreement of concentration given by equation 3-37 with the experimental data support the validity of the assumption made to derive equation 3-26. Biodegradation of Toxaphene and Cosolvent Effects Background The most important reaction for toxaphene degradation is reductive dechlorination, a process that involves two-electron transfer with the release of a chloride ion (Cl-) and its replacement by a hydrogen ion (H+) (Stern et al., 1992, Fingerling et al., 1996, and Buser et al., 2000). Process pathway Generally, significantly fewer toxaphene components are found in biota and environmental samples, indicating the original mixture undergoes extensive alteration. The most important of these reactions is reductive dechlorination (Stern et al., 1992, Fingerling et al., 1996, and Buser et al., 2000). Fingerling et al. (1996) observed reductive dechlorination of chlorobornanes, chlorocamphenes, and related compounds. This process is denoted by substitution of Clby H+ under anaerobic conditions. Buser et al. (2000) studied the anaerobic degradation of toxaphene, and observed more rapid reductive dechlorination in sewage sludge compared to soil. The predominant degradation mechanism of toxaphene is a process consisting of consecutive reactions, such as the following: (decachloro) nonachloro octachloro heptachloro hexachloro (furtherproducts)

PAGE 110

96 Buser et al. (2000) used nonachloro homologs of toxaphene (such as P58 and P59) to test the degradation rate in the non-sterilized anaerobic sludge. The results showed that the nonachloro compounds are most rapidly degraded, followed by the octaand heptachloro compounds. The hexachloro compounds, identified as “dead-end” metabolites of technical toxaphene, showed an initial increase due to the reductive dechlorination of higher chloro-homologs. This immediate increase is then followed by a decrease as the higher chloro-homologs continue to degrade. Process products Buser (2000) employed ECNI SIM chromatograms to study the isomeric composition of all homologs of toxaphene congeners (decachloro, nonachloro, octachloro, heptachloro and hexachloro) in the anaerobic sludge prior to t = 0 and after 16 d of incubation. The data show the significant degradation for decachloro and nonachloro-compounds except P63. Stern (1992) also observed the rapid degradation of nonachloroand decachlorocompounds. It was also shown that most octachloro compounds, including P31, P39, P42 and TC5, degrade rapidly while compounds such as P40/P41, P44, and some later-eluting unknown congeners remained. Many hexachloro and heptachloro compounds including HxSed, Compound X1, HpSed, TC2, and other late-eluting heptachloro compounds, remained after the 16 day incubation period. In the meantime, the P32 congener (a heptachloro compound) degraded rapidly. Buser et al. (2000) found that HxSed, HpSed, TC2, P40/P41 and P44 remained after incubation. HxSed and HpSed were persistent metabolites of degraded technical toxaphene and the predominant toxaphene components in the sediments of toxaphene-treated lakes (Stern et al., 1996) and from the Baltic Sea (Haglund et al., 1998).

PAGE 111

97 Saleh and Casida (1978) had previously reported about the transformation of HxSed from P32 in various chemical, photochemical, and metabolic systems. Fingerling et al. (1996) also asserted that HxSed was formed via the loss of the 2-endo chlorine atom from several chlorobornanes, including P32. Furthermore, the precursors of HxSed could be P39, P42, P56, P59, P63, P69, or some other octachloroor nonachlorocompounds. P42, an octachlorobornane with 2,2'-gem dichloro substitution, can rapidly degrade and form HxSed. Geminal, or gem, represents the structure of two chloro groups located in the same carbon. Assuming an initial loss of a chlorine from the 2-endo position, degradation of P42 leads to a pair of C-7 epimeric heptachlorobornanes, and eventually to formation of HxSed. These epimeric structures indicated that the two-chloro groups were attached to either the C-8 or C-9. Additionally, Fingerling et al. (1996) reported that the two C-7 epimeric nonachlorobornanes, P56 and P59, also eventually led to HxSed. Buser et al. (2000) reported on another persistent metabolite, HpSed, which formed by the rapid degradation of octachlorobornane P39. Buser stated that the loss of a chlorine atom from the 2-exo position would usually occur. Other possible precursors of HpSed are B8-1412, an octachlorobornane (Stern et al., 1998); TC5 (Buser et al., 1994), a major component in technical toxaphene; P40; and P50 (Buser et al., 2000). Buser et al. (2000) also found the formation of P44, an octachlorobornane with 5,5'-gem dichloro substitution, and suggested it might be the result of dechlorination of the 2,2',5,5'-gem chloro substituted nonachlorobornane P62, a major component in technical toxaphene. Stern (1996) employed an isocratic (constant composition) solvent system including an acetonitrile/water mixture (65:35, v/v), GC-ECNIMS, and HPLC on a Nova-Pak HR C18 preparative column to isolate HxSed and HpSed from toxaphene extraction

PAGE 112

98 solution. The HRGC-ECD chromatograms of the fractions containing HxSed and HpSed confirmed the presence of the isolates as the major peaks in their respective chromatograms, while the smaller peaks may be the impurities. Structures The position of the eliminated chlorine under anaerobic conditions depends on the individual substitution of the congener. The more rapidly degrading compounds of toxaphene, including P32, P39, P42, P49, P58, P56, and P59, share common structural features. These compounds all have 2,2'-gem dichloro substitution on the 6-member carbon ring, which seems to make these congeners more susceptible to anaerobic degradation. The first step of reductive dechlorination most often occurs at the C-2 atom, where the pair of labile geminal ring dichloro groups exists, and preferentially from the endo-position. This is then followed by reductive dechlorination of an exo-position. Different congeners of toxaphene have different ratios of the endo to exo positions. For the congener known as P32, the ratio is 1:2.8, as the ratios are indicated by endo:exo; 1:3 for P42; 1:4.8 for P49a; 1:95 for P59; and 1:98 for P56. Each of these ratios stands for the differing congeners of toxaphene. Further, the degradation rates increase for higher chlorinated toxaphene congeners, which have a comparatively lower stability. Thus, after anaerobic degradation, the retention time for the GC peak pattern will become shorter, and was displayed by Fingerling (1996) with the HRGC/ECD chromatogram of technical toxaphene. Likewise, the more persistent compounds of toxaphene, including HxSed, HpSed, P26, P40/P41, P50 and P63, have the structural similarity of having a staggered (e.g. 2-exo, 3-endo, 5-exo, 6-endo or 2-exo, 3-endo, 6-exo) chloro substitution on the 6-member carbon ring (Buser et al., 2000). These effects have been supported by the similar photo

PAGE 113

99 degradation reaction breakdown of toxaphene, in which chlorine loss takes place exactly from the geminal dichloro group in C-2 position (Mueller, 1988). The reason that this chlorine loss occurs at this position could be that toxaphene is synthesized from camphene through a photoinduced chlorination process, whose primary step is the formation of a geminal dichloro group in the C-2 position. However, it is more likely that the loss of chlorine from the C-2 atom occurs because it is the weakest C-Cl bond due to steric interaction of the geminal chlorine atoms. Product Toxicity Two of toxaphene’s major metabolites were identified as 2-exo,3-endo,6-exo,8,9,10hexachlorobornane (hexachlorobornane III/HxSed) and its isomer, 2-exo,3-endo,6endo,8,9,10-hexachlorobornane (hexachlorobornane II). Comparing the 24-h LD50 values for these two metabolites with the value for toxicant B (C10H10Cl7) showed that the ratio of LD50 values for toxicant B: hexachlorobornane III/HxSed: hexachlorobornane II was 36:18:1 for goldfish and 21:7:1 for houseflies. In the experiment of houseflies with piperonyl butoxide pretreatment, a reduction in toxicity to of toxicant B and two dechlorination products were observed (Saleh and Casida, 1978). Miskimmin et al. (1995) observed bioaccumulation of HxSed and HpSed in 2-year-old rainbow trout muscle from both Peanut and Chatwin lakes, which suggested that these components were being transferred from the sediment to the fish. Although the “dead-end” metabolites (such as HxSed) have a reduced toxicity compared to their precursors, they exhibit toxicity to the environment and need to be further reduced. Suspected metabolites of HxSed, such as tetraand pentachlorocamphere, are not found prevalently in the literature.

PAGE 114

100 Khalifa (1976) examined the iron(II) protoporphyrin systems with toxaphene and two of its most toxic components, Toxicants A (C10H10Cl8) and B (C10H10Cl7). The studies of toxaphene reaction with reduced hematin in neutral aqueous medium and analyzed with GC-CI-MS revealed that two or more processes form products from toxicants A and B: reductive dechlorination, dehydrochlorination, and vicinal chloride (chloro groups on adjoining carbon atoms) elimination. The complex structure of toxaphene provides many possible sites of chlorine for initial attack triggering further degradation. The pathway of breakdown could depend on the ease of the C-Cl bond scission at each stage, as well as stereochemical selectivity. The experiment in iron(II) protoporphyrin systems also confirmed that the initial process is likely to be reductive dechlorination due to the action of the geminal dichloro group (Khalifa et al.,1976). Effects of cosolvent on biodegradation of toxaphene Addition of cosolvents could act as electron donors in the anaerobic reductive dechlorination process. However, the toxicity of alcohols to microorganisms has been reported in literature (Ingram,1984 and Powers,2001). It is necessary to investigate the effect of cosolvents in biodegradation of toxaphene to balance the toxicity of cosolvent on biodegradation and enhancement through acting as electron donor and carbon sources. This research conducted preliminary experiments to investigate the biodegradation of toxaphene in the presence of different type and concentration of cosolvents and assess the effect of cosolvent solutions on the biodegradation of toxaphene. Methodology Methanol, ethanol, iso-propanol with concentration from 5 to 40%(v/v) was used in an anaerobic enrichment mixed culture originally from highly toxaphene contaminated

PAGE 115

101 soil at a landfill site at Lake Apopke . The energy source and electron donor was hydrogen. The mass of 1.125g of KH2PO4, 9.092g of Na2HPO4, 0.417g of NH4NO3, 1.7 mL trace elements, and 1.7mL of resazurin was placed into 1170mL water, boiled and then cooled in N2 flow to remove most oxygen in the solutions. The solution was added with 0.2g/L of MgSO47H2O, 0.0286 g/L of Ca(NO3)24H2O and then removed to vials for N2 blow to remove the oxygen, followed by autoclave. The cosolvent solution of 5%, 10%, 20%,30%, 40% methanol, 5% ethanol, 5% IPA was put into vials with 5 grams of contaminated soil respectively. The solution without cosolvent addition was used as controls. Every treatment will be performed in triplicate. The blank controls used 5g soil in the serum vials without addition of any electron donor and carbon sources, and followed by autoclave. The blank controls showed that after 8 weeks, the concentration of toxaphene in the system was degraded 1.78-3.54%. Hydrogen was injected as electron donor. The microorganism was cultured in 30 for 8 weeks and infill H2 every 4-5days. After 8 weeks incubation, the solution was removed and centrifuged. A volume of 2mL solution was placed in vial and extracted the toxaphene into hexane. The toxaphene concentration was analyzed by GC in solution. The soil was left in hood for 24hrs, and then toxaphene was extracted into methyl chloride: acetone= 4:1 solution with ASE and analyze the toxaphene concentration in soil. Toxaphene is not a compound with a simple single peak spectral output, as has been noted in the literature (Fingerling et al., 1996; Pearson et al., 1997), but a complex

PAGE 116

102 mixture that produces a wide GC spectral area output indicative of a complex elution output. Therefore, integration of the output peak over the entire elution time was the method used for analysis, similar to Pearson et al. (1997) and Clark et al. (2005). This provided complete analysis of the complicated spectra produced in the analysis of toxaphene. Integration of the eluted area was based on standard calibration curves, which detailed how the area of toxaphene elution changed with respect to initial calibration concentrations as a function of output time. These calibration curves were developed from the toxaphene standard. Results and Discussion Reduction of toxaphene degradation was observed for the high cosolvent concentrations, which may be attributed to the toxicity of cosolvent to the microorganisms. When concentrations of alcohols are higher than 100,000mg/L (10%v/v), most species of bacteria will be killed due to the disruption of the cellular permeability barrier as stated by Ingram and Buttke (1984). Although cosolvent is useful in increasing the removal of toxaphene from the soil and transport with the groundwater, its toxicity to microorganisms is still a factor in its application. In concentration of 50,000mg/L (5%v/v) , the degradation of toxaphene was inhibited with the degradation rate reducing from 35% to less than 14% compared to the control without cosolvent presence. Lower concentration of alcohols can be degraded in both aerobic and anaerobic condition at a fast rate (Sulflita and Mormile, 1993). The results also suggested that alcohols with higher molecular weight were more potential inhibitors to toxaphene biodegradation. The average biodegradation rates for toxaphene in 5%v/v methanol and ethanol are 13.3% and 10.2%, respectively. Harold (1970) suggested that the longer chain of alcohol with its larger hydrocarbon tail

PAGE 117

103 functional group reduces its polarity and causes the higher partitioning in the hydrophobic cell membrane. In contrast, the larger polarity of shorter chain of alcohol will reduce its concentration within the membrane and decrease its toxicity compared to longer chain alcohol. The toxicity of the alcohols is reported to be dominated by physicohemical properties of alcohols which have potential to affect the hydrophobic ad electrostatic interactions in the cytosolic and envelope components of cells (Powers et al., 2001), which include cell membrances, conformations of enzymes and macromolecules, activity coefficients of metabolites, ionization potentials, pKa values of functional groups, and intracellular pH (Ingram and Buttke, 1984; Yaacobi and Ben-Naim, 1974). The synthesis of various organells including the cell wall, RNA, DNA, and proteins appear to be inhibited under high alcohol concentration( Blumberg and Strominger, 1974; Mitchell and Lucas-Leonard, 1980; Osztovics et al., 1981; Haseltine et al., 1972). Conclusions Based on the batch tests results, the application of zero-valent iron and bimetallic substrates containing Fe0 is effective in degrading toxaphene in solution by dechlorination. A power law relationship depicted decrease of toxaphene concentration in solution in contact with zero-valent iron as a function of time. The application of bimetallic substrates improved the dechlorination of toxaphene in solution up to 278% of the dechlorination rate. A nonlinear Freundlich sorption equation indicated that the linearity of the Freundlich sorption isotherm was related to the saturation of solution, which was the ratio

PAGE 118

104 of aqueous concentration to the solubility (Ca/Sc). When Ca/Sc > 0.2, the sorption isotherm was almost linear. Using the linear partitioning coefficient, a log-linear relationship of sorption constant to cosolvent fraction was employed and indicated the cosolvency power of methanol was 3.45 for toxaphene sorption to iron surface, similar to the value of 3.43 derived from the experimental batch solubility test of toxaphene in methanol solution. Experimental results showed that the degradation rate reduced with the increase of methanol fraction. A mathematical modeling was set up to explain the sorption and dechlorination of toxaphene by the iron solid. The experimental data had a good agreement with the theoretical data. In 5% fraction of cosolvent solutions, the degradation of toxaphene was inhibited with the degradation rate reducing from 35% to less than 14% compared to the control without cosolvent presence. The longer chain alcohols were much more potent inhibitors to toxaphene biodegradation. No obvious reduction of toxaphene was observed for the high fractions (>10%v/v) of methanol solutions.

PAGE 119

105 Figure 3-1. Three possible schemes of pathways for the reductive dechlorination under anoxic Fe0-H2O systems.

PAGE 120

106 Figure 3-2. GC-ECD chromatograms of toxaphene dechlorination by zero-valent iron with different time periods.

PAGE 121

107 Figure 3-3. GC-ECD chromatograms of toxaphene dechlorination by Fe0, NiFe0, CuFe0 after 1 week.

PAGE 122

108 Figure 3-4. Change of Cl concentration with the time

PAGE 123

109 Figure 3-5. Toxaphene degradation rate as functions of methanol fraction under different initial toxaphene concentration C0:

PAGE 124

110 Figure 3-6. Linearity of sorption isotherm to the Ca/Sc, which is changed with initial concentration of toxaphene and cosolvent fraction in solution Figure 3-7. Sorption coefficients Kd onto the iron surfaces as functions of methanol fraction in the range of 0.1 to 0.5.

PAGE 125

111 Figure 3-8. Comparison of experimental data and kinetic results

PAGE 126

112 Figure 3-8. Continued.

PAGE 127

113 Figure 3-9. Effects of 5% fraction of cosolvent solutions (methanol, ethanol, and IPA) on biodegradation of Toxaphene 0%10%20%30%40%50% Figure 3-10. Effects of different fractions of methanol solution biodegradation of Toxaphene

PAGE 128

CHAPTER 4 SUMMARY AND DISCUSSION Conclusions This research explored remediation of toxaphene with mass removal technology and dechlorination technologies. The specific conclusions of this research are: The study of the aqueous solubility of toxaphene in different cosolvent solutions revealed that the higher molecular weight of the alcohol, the higher value of cosolvency power of toxaphene in this solution is. Addition of higher molecular PMOS into CMOS to form ternary system could enhance the solubility greatly compared to the CMO binary system only. Straight chain carbon exhibited higher potential to enhance the solubility of toxaphene than branched chain carbon cosolvents. Batch tests for sorption isotherm of toxaphene to soil indicated that the Sorption constant, Kp, will reduce with increase of the carbon number due to the increase of the solubility of toxaphene in solutions. The first-order bi-continuum model provided good simulation of the non-equilibrium sorption process of the miscible displacement column test for toxaphene. The application of cosolvent flushing can reduce the time required to conduct the sorption experiment for some highly hydrophobic organic such as toxaphene. The higher the fraction of the cosolvent in solution, the greater the reduction of the retardation factor of toxaphene flushing from soil is. The addition of higher carbon number alcohols such as pentanol can enhance the flushing effect greatly. The experimental data validated the log-linear relationship of sorption constant Kp to the cosolvent fractions fc, and reverse first-order rate k2 to the cosolvent fractions fc. the log-log linear relationship was found between the 114

PAGE 129

115 k2 value to the Kp value. The k2 value to the fraction of cosolvent could be estimated successfully based on the literature data and equations. Based on the batch tests results, the application of zero-valent iron and bimetallic substrates could be potentially used as a passive technique in the degradation and removal of toxaphene from the environment. A power law relationship had been proven to depict the decrease of toxaphene concentration related to iron treatment as a function of time. The application of bimetallic substrates has been treated to improve the dechlorination rate. Batch tests with different fractions of methanol and different initial concentration of toxaphene were conducted to investigate the effect of cosolvent on the adsorption of toxaphene from iron surface. A nonlinear Freundlich sorption equation indicated that the linearity of the sorption isotherm was related to the saturation of solution, which was the ratio of aqueous concentration to the solubility (Ca/Sc). When Ca/Sc > 0.2, the sorption isotherm was almost linear. Using the linear partitioning coefficient, a log-linear equation was employed and indicated the cosolvency power of toxaphene in methanol/water solutions was 3.45 for toxaphene sorption to iron surface, similar to the value of 3.43 derived from the experimental batch solubility test of toxaphene in methanol solution. Experimental results showed that the dechlorination rate will reduce with the increase of methanol fraction. The preliminary experiment for the effects of cosolvent solutions on the biodegradation of toxaphene showed that the degradation of toxaphene was inhibited with the degradation rate reducing from 35% to less than 14% compared to the control without cosolvent presence.

PAGE 130

116 Recommendations The remediation techniques discussed in this research could be used to improved remedial operations at hazardous waste sites; however, further work is recommended. Suggested future work includes: Surfactant Flushing and Its Effects on Iron Treatment The role of surfactant solution on the toxaphene treatment has still not been fully understood. The expected result of surfactants application in iron treatment is the increase of surface concentrations and the possible following increase of degradation rate. However, the coverage of surfactants on the iron surface may reduce the contact of toxaphene with iron and affect the degradation rate. The effects of surfactants need to be further investigated and some experiments need to be conducted. The aqueous surfactant-iron system consists of five phases: (1) the aqueous phase, (2) mobile surfactant micelles, (3) surfactant hemimicelles or double layers on the iron surface, (4) reactive binding sites on the iron surface, and (5) non-reactive graphite inclusions on the iron. (Loraine, 2001). The concentrations of the chlorinated compounds can be divided into two states as mobile, which can exist as solubilized and micellar concentrations, and bound, which is sorbed to iron active sites, iron graphite, and iron surfactant phase. Recently, studies of iron reduction of simple chlorinated organics with the surfactants or surfactant/alcohol mixtures observed the increase of degradation rates with increase of surface concentrations. The Triton X-100 enhanced PCE reduction by as much as 40% and the [TX]bound increased at the same time (Loraine ,2001). This correlation indicates that Triton X-100 hemimicelles on iron may have acted as a reactive phase for PCE removal and increase of the surface coverage will cause the increase of

PAGE 131

117 degradation rate. When the TX concentration reached a level (CMC), the unbound micelles formed and the degradation rates declined due to the competition of sorption into surfactant layer and the partition to the mobile phase. It seems that the concentration of TX at nearly CMC could get the best degradation rate. Generally, surfactants are used for extraction of NAPLs from soil. For iron treatment, two possible methods could be used. The first one is PRB. If it is combined with surfactants flushing, a suitable type and concentration of surfactants may be found to reduce the possible negative effect of surfactants in toxaphene degradation on iron surface. Another method is pump-and-treat with surfactants flushing and then using iron treatment as ex situ treatment. In this case, if suitable surfactants are chosen or the concentration of surfactants could be changed, it may even have a positive effect with iron treatment. Actually some researches of surfactant modified zeolite (SMZ) with zero-valent iron have shown promising results in contaminant removal (Zhang, 2002). Surfactant modified zeolite (SMZ) with zero-valent iron on the toxaphene degradation needs further investigation. Effects of Cosolvent/Surfactant Solutions with F raction of :ower than 5% on Biodegradation From the preliminary experiment of the effects of cosolvent solution with fraction higher than 5%, the biodegradation of toxaphene was inhibited. However, the further study are still needed to be conducted to explore the potential of cosolvent as electron donor and carbon source for growth of microbe with the presence of low concentrations of cosolvent. Surfactants were also reported to enhance the growth of the anaerobic microbe and needs to be further investigation.

PAGE 132

118 Combine Biodegradation Technology with Iron Permeable Reactive Barrier (PRB) The H2 produced during the iron treatment could provide the electron donor for many microbe and the consumption of H2 could enhance the permeability in turn. Iron reducing baceria could reduce the ferric oxide and hydroxide coating within the barrier to ferrous state and enhance the iron treatment efficiency (Scherer, 2001). The problem needed to be investigated is the long-term reactivity and permeability of the PRB with biofilm.

PAGE 133

LIST OF REFERENCES Allen, H.L., Mandel, R.M., Dunkleman, T., Shane, D., Torres, M., Crouse, D.G., Miller, T.F., and Hopinkah, D. 2002. Anaerobic bioremediation of toxaphene-contaminated soil. 17th WCSS, August 14-22, 2002, Thailand. Annable, M., Jawitz, J., Rao, P., Dai, D. Kim, H., and Wood, A. 1998. Field evaluation of interfacial and partitioning tracers for characterization of effective NAPL-water contact areas. Ground Water, 36, 495-498. Annable, M., Rao, P., Hatfield, K., Graham, W., Wood, A., and Enfield, C. 1998. Partitioning tracers for measuring residual NAPL: field-scale test results. Journal of Environmental Engineering, 124,498-503. ATSDR. Agency for Toxic Substances and Disease Registry. Toxicological Profile for Toxaphene. Atlanta, GA: ATSDR, Public Health Service, U.S. Department of Health and Human Services. 1996. 231 pp. NTIS Accession No. PB97-121057. Augustijn, D.C.M. 1993. Chemodynamics of complex waste mixtures: applications to contamination and remediation of soils and aquifer media. Ph.D. Dissertation. University of Flordia. Augustijn, D.C.M., D. Dai, P.S.C. Rao, and A.L. Wood. 1994. Solvent flushing dynamics in contaminated soils. In Transport and reactive processes in aquifers. A.A. Balkema/Rotterdam. Brookfield. Pp.557-662. Banerjee S, Yalkowsky S H, 1988. Cosolvent-induced solubilization of hydrophobic compounds into water. Analytical Chemistry, 60, 2153-2155. Bidleman TF, Christensen EJ, Harder HW. 1981. Aerial deposition of organochlorines in urban and coastal South Carolina. Atmos Pollut Nat Waters 481-508. Blumberg, P. and J. Strominge. 1974. Interaction of penicillin with the bacterial cell: penicillinbinding proteins and pencillin-sensitive enzymes. Bateriol. Rev. 38,291. Buser, H.R., Haglund b , P., Mller, M.D., Poiger a , T., and Rappe b , C. 2000. Rapid anaerobic degradation of toxaphene in sewage sludge. Chemosphere, 40,1213-1220. 119

PAGE 134

120 Buser, H.R. and Mller, M.D. 1994. Isomerand enantiomer-selective analyses of toxaphene components using chiral high-resolution gas chromatography and detection by mass spectrometry/mass spectrometry. Environmental Science and Technology, 28, 119-128. Brusseau, M.L., R.E. Jessup, and P.S.C. Rao. 1991. Influence of organic cosolvents on the sorption kinetics of hydrophobic organic chemicals. Environmental Science and Technology, 25, 903-910. Brusseau, M.L., A.L. Wood, and P.S.C. Rao. 1990. Sorption kinetics of organic chemicals: evaluation of gas-purge and miscible-displacement techniques. Environmental Science and Technology, 24, 727-735. Bounchard, D.C. 1998. Sorption kinetics of PAHs in methanol-water systems. Contaminant Hydrology. 34, 107-120. Bounchard, D.C. 2002. Cosolvent effects on sorption isotherm linearity. Contaminant Hydrology. 56, 159-174. Chawla R C, Doura K F, McKay D, 2001. Effect of alcohol cosolvents on the aqueous solubility of trichloroethylene. Proceeding of the 2001 conference on environmental research. Kansas state university, Manhattan, Kansas. Cho J, Annable M D, Rao P S C, 2003. Residual alcohol influence on NAPL saturation estimates based on partition tracers. Environmental Science and Technology, 37, 1639-1644. Clark,C.J. II, Chen, X. and Babu, S. 2005. Degradation of toxaphene by zero-valent iron and bimetallic substrates. In review. Clark,C.J. II, Rao, P.S.C., and Annable, M.D. 2003. Degradation of perchloroethylene in cosolvent solutions by zero-valent iron. Journal of Hazardous Materials B, 96, 65-78. Cline, P.V., J.J. Delfino, and P.S.C. Rao. 1991. Partitioning of aromatic constituents into water from gasoline and other complex solvent mixtures. Environmental Science and Technology 25: 914-920. Dai,D. 1997. Solubility and partitioning behavior of organic compounds in complex NAPL mixtures. Ph.D. Dissertation. University of Florida. Fingerling , G., Herkorn, N., and Parlar, H. 1996. Formation and spectroscopic investigation of two hexachlorobornanes from six environmentally relevant toxaphene components by reductive dechlorination in soil under anaerobic conditions. Environmental Science and Technology, 30, 2984.

PAGE 135

121 Gavaskar ,A., Gupta ,N., Sass, B., Janosy, R., and Hicks,J. 2000. Design Guidance for Application of Permeable Reactive Barriers for Groundwater Remediation. SERDP, F08637-95-D6004. Gooch JW, Matsumura F. 1985. Evaluation of the toxic components of toxaphene in Lake Michigan lake trout. J Agric Food Chem 33:844-848. Gillham,R.W. and O’Hannesin, S.F. 1994. Enhanced degradation of halogenated aliphatics by zero-valent iron. Ground Water, 32, 958-967. Gray, N.C.C., Cline, P.R., Moser, G.P., and Gannon, D.J. 2000. Bioremediation of toxaphene contaminated soil. Second International Conference on Remediation of Chlorinated and Recalcitrant Compounds, Monterey, California. 251-258 Griffith Jr., F.D. and Blanke, R.V. 1974. Microcoulometric determination of organochlorine pesticides in human blood. J Assoc Off Anal Chem. 57, 595-603. Gupte P A, Danner R P, 1987. Prediction of liquid-liquid equilibria with UNIFAC: a critical evaluation. Industrial & Engineering Chemistry Research, 26, 2036-2042. Haglund, P., Andersson, R., Buser, H.R., Wagman, N., and Rappe, C. 1998. A non-destructive cleanup procedure for chlorobornanes in water and sediment. Organohalogen Compounds 35, pp. 275. Hainzl D, Burhenne J, Parlar H. 1994. Theoretical consideration of the structural variety in the toxaphene mixture taking into account recent experimental results. Chemosphere 28(2):245-252. Haseltine, W.A., R. Block, W. Gilbert, and K. Weber. 1972. MSI and MSII made on ribosome in idling step of proteins synthesis. Nature. 238,381. Hayworth, J.S. and Burris, D.R., 1997. Nonionic surfactant-enhanced solubilization and recovery of organic contaminants from within cationic surfactant enhanced sorbent zones. 1. Experiments. Environmental Science and Technology 31, p. 1277. Hillery, B. R.; Basu, I.; Sweet, C. W.; Hites, R. A. 1997. Temporal and Spatial Trends in a Long-Term Study of Gas-Phase PCB Concentrations near the Great Lakes. Environmental Science and Technology, 31, 1811-1816. Ingram, L., and Buttke,T. 1984. Effects of alcohols on micro-organisms. Advances in Microbial Physiology, 25, 253 Jafvert C T, 1996. Technology evaluation report: surfactants/cosolvents. Ground-water remediation technologies analysis center, Pittsburgh, PA. www.gwrtac.org James R R and Hites R A, 2002. Atmospheric transport of toxaphene from the southern united states to the great lakes region. Environmental Science and Technology, 36, 3474-3481.

PAGE 136

122 Jawitz, J.W., Annable, M.D., Rao, P.S.C. and Rhue, R.D., 1998. Field implementation of a winsor type 1 surfactant/alcohol mixture for in situ solubilization of a complex LNAPL as a single-phase microemulsion. Environmental Science and Technology 32, 523. Johnson ,T. L., Scherer ,M. M., and Tratnyek ,P. G. 1996. Kinetics of Halogenated Organic Compound Degradation by Iron Metal.. Environmental Science and Technology, 30, 2634-2640. Kalu, E.E. 1998. Aging effects of electroless cobalt batch on microstructure and magnetic properties of Co-P films. Plating and surface Finishing, 85, 74-78. Khalifa, S., Holmstead, R.L. and Casida, J.E., 1976. Toxaphene degradation by Iron (II) protoporphyrin systems. J. Agric. Food Chem. 24, 277. Korte, N., Liang, L., Muftikian, R., Grittini, C., and Fernando, Q. 1997. The dechlorination of hydrocarbons. Plantinum Metals Review, 41, 2-7. Kuruganti, A.S., Chen, K.S., and Kalu E.E. 2001. Evaluation of a printable catalyst for use in flex-circuit and PCB application. Planting and Surface Finishing, 88, 60-66. Lampron, K. J., Chiu ,P. C. and Cha , D. K 2001. Reductive Dehalogenation of Chlorinated Ethenes with Elemental Iron: The Role of Microorganisms. Water Research, 35, 3077-3084. Li A, 2001. Predicting cosolvency. 3. Evaluation of extended log linear model. Ind. Eng. Chem. Res. 40, 5029-5035. Li A, Andren A W, 1994. Solubility of Polychlorinated biphenyls in water/alcohol mixtures. 1. Experimental data. Environmental Science and Technology, 28, 47-52. Li A, Yalkowsky S H, 1998a. Predicting cosolvency. 1. Solubility ration and solute log Kow. Ind. Eng. Chem. Res. 37, 4470-4475. Li A, Yalkowsky S H, 1998b. Predicting cosolvency. 2. Correlation with solvent physicochemical properties. Ind. Eng. Chem. Res. 37, 4476-4480. Li P, Zhao L, Yalkowsky S H, 1999. Combined effect of cosolvent and cyclodextrin on solubilization of non polar drgs. Journal of pharmaceutical sciences, 88 (11), 1107-1111. Lee C M, 2001. Non aqueous phase liquids in subsurface. EnviroChemLibrary case study. National science foundation. www.ces.clemson.edu/ees/lee Lee L S, Bellin C A, Pinal R, Rao P S C, 1993. Cosolvent effects of sorption of organic acids by soils from mixed solvents. Environmental Science and Technology, 27, 165-171.

PAGE 137

123 Loraine, G.A. 2001. Effects of alcohols, anionic and nonionic surfactants on the reduction of PCE and TCE by zero-valent iron. Water Research, 35, 1453-1460. Mackinson FW. 1981. Occupational health guideline for chlorinated camphene. Pub1 No. 81-123 DHHS (NIOSH) Washington, DC. Matheson, L. J.; Tratnyek, P. G. 1994. Reductive dehalogenation of chlorinated methanes by iron metal. Environmental Science and Technology, 28, 2045-2053. Matsumura F, Howard RW, Nelson JO. 1975. Structure of the toxic fraction A of toxaphene. Chemosphere 5:271-276. Mirsatari SG, McChesney MM, Craigmill AC, et al. 1987. Anaerobic microbial dechlorination: anapproach to on-site treatment of toxaphene-contaminated soil. J Environ Sci Health B22:663-690. Mitchell, J. and J. Lucas-Leonard. 1980. The effect of alcohols on uanosine 5-diphosohate-3-diphosphate metabolism in stringent and relaxed Escherichia coli. J. Biol. Chem. 255,6307. Morris K R, Abramowitz R, Pinal R, Davis P, Yalkowsky S H, 1988. Solubility of aromatic pollutants in mixed solvents. Chemosphere, 17, 285-298. Muftikian, R., Fernando, Q., and Korte, N. 1995. A method for the rapid dechlorination of low molecular weight chlorinated hydrocarbons in water. Water Research, 29, 2434-2439. Nash , R.G. and Woolson, E.A. 1967. Persistence of chlorinated hydrocarbon insecticides in soils. Science, 157, 924. Nelson JO, Matsumura F. 1975. Separation and comparative toxicity of toxaphene components. J Agric Food Chem 23:984-990. Nkedl-Klzza, P., P.S.C. Rao, and A.G. Hornsby. 1987. Influence of organic cosolvents on leaching of hydrophobic organic chemicals through soils. Environmental Science and Technology, 21, 1107-1111. Osztovics, M., S. Lgali, A. Antal, and P. Veghelyi. 1981. Alcohol is not mutagenic. Mutat. Res. 74,247. Paris, D.F., Lewis, D.L., and Barnett, J.T. 1977. Bioconcentration of toxaphene by microorganisms. Bulletin of Env. Contamination & Toxicology, 17, 564-572. Park G.S. 1952 The Determination of the Concentration Dependent Diffusion Coefficient for Methylene Chloride in Polystyrene by a Steady State Method. Trans. Faraday Society 48:11-21.

PAGE 138

124 Parr, J.F. and Smith, S. 1969. A multi-purpose manifold assembly: use in evaluating microbiological effects of pesticides. Soil Sci., 107, 271-276. Parr, J.F. and Smith, S. 1976. Degradation of toxaphene in selected anaerobic soil environments. Soi. Scenc., 121, 51-57. Pearson, R.F., Swackhamer, D.L., Eisenreich, S.J., and Long, D.T. 1997. Concentrations, accumulations, and inventories of toxaphene in sediments of the Great Lakes. Environmental Science and Technology, 31, 3523-3529. Pinal R, Lee S L, Rao P S C, 1991. Prediction of the Solubility of Hydrophobic Compounds in Nonideal Solvent Mixtures, Chemosphere, 22, 939-951. Pinal R, Rao P S C, Lee L S, Cline P V, 1990. Cosolvency of partially miscible organic solvents on the solubility of hydrophobic organic chemicals. Environmental Science and Technology, 24, 639-647. Powers S E, Hunt C S, Heermann S E, Corseuil H X, Rice D, Alvarez P J J, 2001. The transport and fate of Ethanol and BTEX in groundwater contaminated by gasohol. Critical review in Environmental Science and Technology: 31, 79-123. Rao P S C, Annable M D, Sillan R K, Dai D, Hatfield K, Graham W D, Wood A L, Enfield C G, 1997. Field-scale evaluation of in situ cosolvent flushing for enhanced aquifer remediation. Water resources research, 33, 2673-2686. Saleh MA, Casida JE. 1978. Reduction dechlorination of the toxaphene component 2,2,5-endo,6-exo,8,9,10-heptachlorobornane in various chemical, photochemical, and metabolic systems. J. Agric Food Chem 26:583-599. Stern, G.A., Loewen, M.D., Miskimmin, B.M., Muir, D.C.G. and Westmore, J.B., 1996. Characterization of two toxaphene components in treated lake sediment. Environmental Science and Technology 30, 2251. U.S. EPA. 1981. Basis for listing hazardous waste. U. S. Environmental Protection Agency. Code of Federal Regulations. U.S. EPA. 1995. U.S. Environmental Protection Agency: Method 8081A: Chlorine pesticides by capillary column gas chromatography. Test methods for evaluating solid waste, Vol IB: Laboratory Manual Physical/Chemical methods. Revision 0, January 1995. U.S. EPA. 1998a. Metal-enhanced dechlorination of volatile organic compounds using an in-situ reactive iron wall U.S. EPA/540/R-98/501. Contract No. 68-C5-0037. U.S. EPA, 1998b. Method 8081B: Organochlorine pesticides by gas chromatography. Test methods for evaluating solid waste, physical/chemical methods. Revision 2, January 1998.

PAGE 139

125 U.S. EPA, 1999. In Situ Enhanced Source Removal. U.S. EPA/600/C-99/002. U.S. EPA. 2003. Public health goal for chemicals in drinking water. Wang, C. and Zhang, W. 1997. Synthesizing nanoscale iron particles for rapidand complete dechlorination of TCE and PCBs. Environmental Science and Technology, 31,2154-2156. Wauchope RD, Buttler TM, Homsby AG, et al. 1992. The SCS/ARS/CES pesticide properties database for environmental decision-making. In: Ware GW ed. Reviews of Environmental Contamination and Toxicology 123:34, 149 Worthing CR, Walker SB, eds. 1987. The pesticide manual: A world compendium. 8th ed. Thornton Heath, UK: The British Crop Protection Council. Van Genuchten, M. Th. 1981. Analytical solution for chemical transport with simultaneous absorption zero-order production and first-order decay. J. Hydrol., 49, 213-233. Xia, G., and J.J. Pignatello. 2001. Detailed sorption isotherm of polar compounds in a high-organic soil. Environmental Science and Technology. 35, 84-94. Yaacobi,M. and A. Ben-Naim. 1974. Solvophobic interaction. J. Phys. Chem. 78,175. Yalkowsy, S.H. and T. Roseman. 1981. Solubilization of drugs by cosolvents IN: Techniques of Solubilization of Drugs, S.H. Yalkowsky (Ed.). Marcel Dekker, Inc., New York, NY, 91-134.

PAGE 140

BIOGRAPHICAL SKETCH Xiaosong Chen was born in 1977 in Fujian, China. He did his undergraduate studies in fine chemistry and received a Bachelor of Chemical Engineering degree from Shanghai Jiaotong University in 1999. He then attended the Department of Biological Technology in Shanghai Jiaotong University, where he received a Master of Engineering degree in environmental science in March 2002. In August 2002, he joined the University of Florida for his Ph.D. degree in the Department of Civil and Coastal Engineering under the tutelage of Dr. Clayton J. Clark II. 126