EPA 600 3-80 041 United States Environmental Protection Agency Environmental Research Laboratory Athens GA 30605 EPA-600/3-80-041 April 1980 Research and Development Sorption Properties of Sediments and Energy-Related > ' Pollutants OAK ST. HDSF RESEARCH REPORTING SERIES Research reports of the Office of Research and Development, U.S. Environmental Protection Agency, have been grouped into nine series. These nine broad cate- gories were established to facilitate further development and application of en- vironmental technology Elimination of traditional grouping was consciously planned to foster ter' The nine series a r >iated fields. 1. Enviro 2 Enviro 3 Ecoloc 4. Enviro 5. Socioe 6 Scienti 7. Interac 8 "Speci 9 Miscel This report has be describes resean cies, and materic ences. Investigat mine the fate of pc for setting standa aquatic, terrestris EPA Hassett, John J. SORPTION 600 PROPERTIES OF SEDIMENTS 3-80 AND ENERGY-RELATED 041 POLLUTANTS. DATE DUE WMRC Library One Hazelwood Drive Champaign, IL 61820 (217) 333-8957 library@wmrc.uiuc.edu DIMCO This document is available to the public through the National Technical Informa- tion Service, Springfield, Virginia 22161 EPA-600/3-80-041 April 1980 SORPTION PROPERTIES OF SEDIMENTS AND ENERGY-RELATED POLLUTANTS by a b John J. Has sett and Jay C. Means Co-Principal Investigators a c Wayne L. Banwart and Susanne G. Wood a Department of Agronomy, University of Illinois at Urbana- Champaign, Urbana, Illinois 61801 Chesapeake Biological Laboratory, University of Maryland Solomons, Maryland 20688 c Institute for Environmental Studies, University of Illinois at Urbana-Champaign, Urbana, Illinois 61801 Contract No. 68-03-2555 Project Officer David S. Brown Environmental Processes Branch Environmental Research Laboratory Athens, Georgia 30605 ENVIRONMENTAL RESEARCH LABORATORY OFFICE OF RESEARCH AND DEVELOPMENT U.S. ENVIRONMENTAL PROTECTION AGENCY ATHENS, GEORGIA 30605 DISCLAIMER This report has been reviewed by the Environmental Research Laboratory, U.S. Environmental Protection Agency, Athens, Georgia, and approved for pub- lication. Approval does not signify that the contents necessarily reflect the views and policies of the U.S. Environmental Protection Agency, nor does mention of trade names or commercial products constitute endorsement or recom- mendation for use. ii FOREWORD Environmental protection efforts are increasingly directed towards prevention of adverse health and ecological effects associated with specific compounds of natural or human origin. As part of this Laboratory's research on the occurrence, movement, transformation, impact, and control of environ- mental contaminants, the Environmental Processes Branch studies the micro- biological, chemical, and physico-chemical processes that control the trans- port, transformation, and impact of pollutants in soil and water. Efforts to achieve our national goal of energy independence will require increasing use of our country's vast domestic coal reserves. The combustion of coal or its conversion to a gaseous or liquid fuel, however, can release numerous organic compounds that are potentially toxic, carcinogenic, or muta- genic. This report examines the sorption properties of several energy-rela- ted pollutants on sediments. Information on these properties is needed to predict the movement of the compounds in aquatic systems so that potential environmental problems can be anticipated. David W. Duttweiler Director Environmental Research Laboratory Athens, Georgia iii ABSTRACT This report describes the factors that determine the extent of sorption of organic compounds that are representative of coal conversion waste streams. The compounds, all radiolabeled, were acetophenone , 1-naphthol, pyrene, 7 , 12-dimethy lbenz [a] anthracene , 3-methy lcholanthrene, dibenz [a,/z] anthracene , acridine, 2,2 '-biquinoline, 13#-dibenzo [a, i] carbazole , dibenzothiophene, benzidine, 2-aminoanthracene , 6-aminochrysene , and anthracene-9-carboxylic acid. Batch equilibrium isotherms were determined for each compound on four- teen sediments and soils that had been collected from the Missouri, Illinois, Mississippi and Ohio rivers and their watersheds. Laboratory procedures for determining oc tanol-water partition coefficients and water solubilities were developed and then performed on the compounds. The sorption constants were correlated with soil and sediment proper- ties and with the water solubilities and octanol-water partition coefficients of the compounds. Regression equations were developed that allow prediction of a hydrophobic compound's linear partition coefficient from knowledge of the compound's octanol-water partition coefficient or its water solubility and the organic carbon content of the sediment or soil. Regression equations were tested on independent data sets from the literature for the adsorption of parathion and a variety of halogenated hydrocarbons. Observed values for these compounds were in good agreement with values predicted by the regression equations . This report was submitted in partial fulfillment of Contract No. 68- 03-2555 by the Department of Agronomy and the Institute for Environmental Studies of the University of Illinois at Urbana-Champaign under the sponsor- ship of the U.S. Environmental Protection Agency. The report covers the period from July 1, 1977 to December 31, 1979. CONTENTS Foreword iii Abstract iv Figures , vii Maps ix Tables x Acknowledgments xiv SECTION 1. INTRODUCTION 1 SECTION 2. CONCLUSIONS 4 SECTION 3. RECOMMENDATIONS 12 SECTION 4. LABORATORY INVESTIGATIONS 13 4.1. Acetophenone 13 4.2. 1-Naphthol 18 - 4.3. Benzidine 27 4.4. Pyrene 35 4.5. 7 , 12-Dimethylbenz [a] anthracene 41 4.6. Dibenz [a, h] anthracene 46 4.7. 3-Methylcholanthrene 51 4.8. Dibenzothiophene 56 4.9. Acridine 62 4.10. 2,2'-Biquinoline ..... 67 4.11. 13#-Dibenzo [a 3 i>] carbazole 71 4.12. 2-Aminoanthracene 76 4.13. 6-Aminochrysene 80 4.14. Anthracene-9-Carboxylic Acid 84 SECTION 5. EXPERIMENTAL METHODS 89 5.1. Preparation and Tritiation of Compounds 89 5.1.1. Chromatographic Analyses 89 5.1.1.1. Thin- Layer Chromatography 89 5.1.1.2. Gas-Liquid Chromatography 91 5.1.2. Purity and Purification Procedures 91 5.1.2.1. Determination of Specific Activity and Radiochemical Purity 91 5.1.2.2. Purification of Compounds , 93 5.1.3. Tritiation 95 5.1.4. Radioactivity Measurements ... 96 5.2. Octanol-Water Partitioning Procedure .... 97 5.3. Water Solubility Determination 99 5.4. Degradation Studies 102 5.4.1. Sorption Isotherms , 102 v CONTENTS Continued 5.4.1.1. Soil/Sediment Phases , , , . . 102 5.4.1.2. Aqueous Phases , 104 5.4.2. Octanol-Water Partitionings 104 5.4.2.1. Octanol Phases 104 5.4.2.2. Aqueous Phases 105 5.4.3. Water Solubilities 105 SECTION 6. SAMPLE SELECTION AND CHARACTERIZATION 106 6.1. Criteria for Sample Selection 106 6.2. Sample Characterization 110 6.2.1. Instrumentation Neutron Activation Analysis (INAA) Ill 6.2.2. Clay Mineral Analysis Ill 6.2.3. Amorphous Al, Fe and Si 116 6.3. Effect of Sample Pretreatment on Sorption of Acetophenone by Sediments 119 6.3.1. Experimental Methods 119 6.3.2. Results and Discussion 121 6.3.2.1. Sample Pretreatment 121 6.3.2.2. Effect of Extent of Grinding 123 6.3.2.3. Organic Matter Removal 124 SECTION 7. LITERATURE CITED 127 vi FIGURES Number Page 2.1 Relationship between Koc and Octanol-Water Partition Coefficient (Kow) of Energy-Related Organic Pollutants .... 7 2.2 Relationship between Koc and Water Solubility (S) of Energy-Related Organic Pollutants 9 4.1 Representative Isotherms for the Sorption of Acetophenone by Soils and Sediments 15 4.2 Representative Isotherms for the Sorption of 1-Naphthol by Soils and Sediments 22 4.3 Relationship between Molar Kd (for the Sorption of 1-Naphthol) and %0C for Ten Soil and Sediment Samples Showing High Correlation between the Two Parameters and for Six Samples that Gave High Koc Values 23 4.4 Relationship between Koc (for the Sorption of 1-Naphthol) and the %OC/Montmorillonite Ratio of Soils and Sediments for Ten Samples Showing High Correlation between Kd and %0C and for Six Samples that Gave High Koc Values 24 4.5 Effect of Varying the Ethanol:Water Ratio in the Solvent System on Soil TLC R^ Values for Dicamba, a-Naphthol and 3-Methylcholanthrene 26 4.6 Representative Isotherms for the Sorption of Benzidine by Soils and Sediments 29 4.7 Effect of pH on the Sorption of Benzidine by Soils and Sediments 33 4.8 Representative Isotherms for the Sorption of Pyrene by Soils and Sediments 37 4.9 Representative Isotherms for the Sorption of 7 , 12-Dimethyl- benz [a]anthracene by Soils and Sediments 43 4.10 Representative Isotherms for the Sorption of Dibenz [a s h]- anthracene by Soils and Sediments 48 vii FIGURES Continued Number Page 4.11 Representative Isotherms for the Sorption of 3-Methyl- cholanthrene by Soils and Sediments 52 4.12 Representative Isotherms for the Sorption of Dibenzo- thiophene by Soils and Sediments 58 4.13 Representative Isotherms for the Sorption of Acridine by Soils and Sediments 65 4.14 Representative Isotherms for the Sorption of 2,2'-Biquinoline by Soils and Sediments 68 4.15 Representative Isotherms for the Sorption of 13#-Dibenzo [a, i] carbazole by Soils and Sediments 72 4.16 Representative Isotherms for the Sorption of 2-Amino- anthracene by Soils and Sediments 77 4.17 Representative Isotherms for the Sorption of 6-Amino- chrysene by Soils and Sediments 81 4.18 Representative Isotherms for the Sorption of Anthracene-9-Carboxylic Acid by Soils and Sediments 85 5.1 Protocol for Determining the Octanol-Water Partition Coefficient of a Hydrophobic Organic Compound 98 5.2 Protocol for Determining the Water Solubility of a Hydrophobic Organic Compound 101 6.1 Comparison of Isotherms for the Sorption of Aceto- phenone by Fresh and Air-Dried Samples of Crane Island Sediment 122 6.2 Comparison of the Linear Partition Coefficients (Kp) for the Sorption of Acetophenone by Soil and Sediment Samples and the Percent Clay in the Samples after Organic Matter Oxidation 126 viii MAPS Number Page 6.1 Areas of High Potential for Coal Gasification Development 108 6.2 Sampling Sites on the Missouri, Ohio, Wabash, Illinois and Mississippi River Systems and their Watersheds ..... 109 ix TABLES Number Page 1.1 Compounds Selected for Study 3 2.1 Koc Values and their Correlation Coefficients for the Sorption of Energy-Related Organic Pollutants by Soils and Sediments; Octanol-Water Partition Coefficients and Water Solubilities of the Same Compounds 6 2.2 Predicted and Measured Kd Values for the Sorption of Parathion 10 2.3 Predicted and Measured Koc Values for the Sorption of Halogenated Hydrocarbons by Willamette Silt Loam (1.6% Organic Matter, 0.84% Organic Carbon) 10 4.1 Physical Properties of Acetophenone 13 4.2 Acetophenone Sorption Isotherm Data ..... 16 4.3 Linear Partition Coefficients (Kp) and Koc Values for the Sorption of Acetophenone by Soils and Sediments 17 4.4 Physical Properties of 1-Naphthol 18 4.5 1-Naphthol Sorption Isotherm Data 20 4.6 Freundlich Sorption Constants, r 2 Values and Koc Values for the Sorption of 1-Naphthol by Soils and Sediments 21 4.7 Physical Properties of Benzidine 27 4.8 Freundlich Constants (Kd and 1/n) and r 2 Values for the Sorption of Benzidine by Soils and Sediments (Molar Basis) ... 30 4.9 Benzidine Sorption Isotherm Data 31 4.10 Modified Freundlich Partition Constants (1/n = 0.5) for the Sorption of Benzidine (Kd 2 ) and Ionized Benzidine (Kd3) by Soils and Sediments (Molar Basis) 32 4.11 Physical Properties of Pyrene 35 x TABLES Continued Number Page 4.12 Pyrene Sorption Isotherm Data 38 4.13 Freundlich Sorption Constants and Correlation Coefficients (Kd, 1/n and r 2 ) and the Modified Freundlich Partition Constants (Kp, 1/n = 1) for the Sorption of Pyrene by Soils and Sediments 39 4.14 Linear Partition Coefficients (Kp) and Koc Values for the Sorption of Pyrene by Soils and Sediments 40 4.15 Physical Properties of 7 , 12-Dimethylbenz [a] anthracene .... 41 4.16 7, 12-Dimethylbenz [a] anthracene Sorption Isotherm Data .... 44 4.17 Linear Partition Coefficients (Kp) and Koc Values for the Sorption of 7 , 12-Dimethylbenz [a] anthracene by Soils and Sediments 45 4.18 Physical Properties of Dibenz [a 3 h~\ anthracene 46 4.19 Dibenz [a 3 h] anthracene Sorption Isotherm Data 49 4.20 Linear Partition Coefficients (Kp) and Koc Values for the Sorption of Dibenz \a s h] anthracene by Soils and Sediments ... 50 4.21 Physical Properties of 3-Methylcholanthrene 51 4.22 3-Methylcholanthrene Sorption Isotherm Data 53 4.23 Linear Partition Coefficients (Kp) and Koc Values for the Sorption of 3-Methylcholanthrene by Soils and Sediments ... 54 4.24 Physical Properties of Dibenzothiophene 56 4.25 Dibenzothiophene Sorption Isotherm Data 59 4.26 Linear Partition Constants (Kp) and Their r 2 Values, and Molar Freundlich Constants (Kd and 1/n) and Their r 2 Values for the Sorption of Dibenzothiophene by Soils and Sediments 60 4.27 Physical Properties of Acridine 62 4.28 Acridine Sorption Isotherm Data 64 4.29 Modified Freundlich Partition Constants (Kp, 1/n = 1) and Koc Values for Sorption of Acridine by Soils and Sediments . . 66 xi TABLES Continued Number Page 4.30 Physical Properties of 2, 2 ' -Biquinoline 67 4.31 2, 2 '-Biquinoline Sorption Isotherm Data 69 4.32 Modified Freundlich Partition Constants (Kp, 1/n = 1) and Koc Values for Sorption of 2, 2' -Biquinoline by Soils and Sediments 70 4.33 Physical Properties of 13#-Dibenzo [a., -£] carbazole 71 4.34 13#-Dibenzo[a, t] carbazole Sorption Isotherm Data 73 4.35 Modified Freundlich Partition Constants (Kp, 1/n = 1) and Koc Values for Sorption of 13#-Dibenzo [a 3 i ] carbazole by Soils and Sediments 74 4.36 Correlation (r 2 ) of 13#-Dibenzo[a, £] carbazole Kp with Selected Soil/Sediment Properties for the 14 Soils and Sediments Studied 75 4.37 Physical Properties of 2-Aminoanthracene . . . 76 4.38 2-Aminoanthracene Sorption Isotherm Data 78 4.39 Linear Partition Coefficients (Kp) and Koc Values for the Sorption of 2-Aminoanthracene by Soils and Sediments . . 79 4.40 Physical Properties of 6-Aminochrysene 80 4.41 6-Aminochrysene Sorption Isotherm Data 82 4.42 Linear Partition Coefficients (Kp) and Koc Values for the Sorption of 6-Aminochrysene by Soils and Sediments 83 4.43 Physical Properties of Anthracene-9-Carboxylic Acid 84 4.44 Anthracene-9-Carboxylic Acid Sorption Isotherm Data .... 86 4.45 Linear Partition Coefficients (Kp) and Koc Values for the Sorption of Anthracene-9-Carboxylic Acid by Soils and Sediments 87 5.1 Typical Rf Values for 14 Energy-Related Compounds in Selected Thin-Layer Chromatographic Solvent Systems 90 5.2 Typical Gas-Liquid Chromatographic Conditions and Retention Times for 14 Energy-Related Compounds 92 xii TABLES Continued Number Page 5.3 Radiochemical Purity Information for 14-Energy-Related Compounds Used in this Study 94 5.4 Octanol-Water Partition Coefficients (Kow) of Energy-Related Organic Pollutants 100 5.5 Water Solubilities of Energy- Re la ted Organic Pollutants . . . 103 6.1 Field Notes 106-107 6.2 Characteristics of Soils and Sediments 110 6.3 Elemental Limits of Detection by Instrumental Neutron Activation 112 6.4 Instrumental Neutron Activation Analysis of Soils and Sediments 113-114 6.5 Qualitative Determination of the Different Clay Minerals in the less-than-2u Fraction of the Soil and Sediment Samples 115 6.6 Semiquantitative Determination of Clay Minerals in the Soils and Sediments 117 6.7 Sodium Hydroxide- Extractable Si and Al and Citrate- Dithionate-Extractable Fe in the Soil and Sediment Samples . . 118 6.8 Correlation Matrix of Selected Sample Characteristics .... 118 6.9 Selected Chemical and Physical Properties of River Sediments . 120 6.10 Effect of Sample Pretreatment on Modified Freundlich Partition Coefficients (Kp) 121 6.11 Effect of Extent of Grinding on Sorption (Kp) of Acetophenone 123 6.12 Organic Carbon Contents and Modified Freundlich Partition Coefficients (Kp, 1/n = 1) Before and After Organic Matter Removal, and Expandable (2:1) Clay Material in Soil and Sediment Samples 124 xiii ACKNOWLEDGMENTS The authors gratefully acknowledge the guidance and counsel of Dr. David S. Brown, Project Officer, of the Environmental Research Laboratory, Athens, Georgia. The authors also express their appreciation to Dr. Samuel W. Karickhoff, Environmental Research Laboratory, Athens, Georgia, for valuable advice regarding some technical difficulties encountered during the investi- gation. Special acknowledgment is made of the innovative technical contribu- tions of Drs. Adam Khan and Syed Ali. The authors are grateful to John J. Ameel, Sandra K. Dick, David D. Ellis, Carolyn T. Hanson, Elsa K. Tong, David L. Zierath and Kathleen A. Brinkman for their excellent technical assistance, and to V. Jean Clarke for typing this report. xiv SECTION 1 INTRODUCTION The contract contained two basic tasks. The first was to perform a literature review covering the theory of sorption and the sorption properties of energy-related compounds. This literature review, Adsorption of Energy- Related Organic Pollutants (EPA-600/3-79-086) , has been published by the U.S. Environmental Protection Agency. The second task involved determination of sediment/soil and pollutant properties that control sorption of compounds typical of coal conversion effluent streams. With the current proposed increase in the use of coal for energy production, there is concern for the environmental effects of pollutants that are produced at various stages of coal mining and processing. Coal is an extremely complex organic polymer interlaced with inorganic trace impurities. The following figure shows a typical chemical representation of the coal polymer (1). H H It is apparent that the various fragments of the coal structure include a tremendous variety of polycyclic aromatic, heterocyclic aromatic, phenolic, amine, quinone, sulfur, nitrogen and other compounds. The charac- terization of some of the organic wastes from coal conversion has been re- ported for pilot plant studies. Forney et al. (2) identified some of the major organic constituents of tars produced by the Synthane coal gasification process. Schmidt et al. (3) has analyzed process water for major organic constituents. 1 Coal conversion processes result in extremely large gaseous and aqueous effluent streams (4). In addition to the wastes produced directly by the conversion process, unspecified large quantities of water are produced constituting the leachate from coal storage, solid wastes and particulates. When effluent streams this large are produced, significant quantities of constitutents can be placed into the environment. Introduction of the effluent streams into the environment will result in exposure of the organic pollutants to sediments or soils and their subsequent sorption to the extent that it is chemically or physically dictated. Sorption results in lower aqueous concentrations of the pollutants. Hence, potential physiological activities (5) and release of mobilities within the environment as well as other effects may be decreased. The extent of sorption of organic materials by sediments and soils is dependent upon the nature of both the sorbent (soil or sediment) and the sorbate (pollutant). In a number of studies of organic pesticide sorption by soils, sorption has been correlated with humus content, clay mineral type and content, texture, pH and hydrous oxide content. The nature of the organic molecule, whether it is a cation, anion or neutral molecule, its water solubility or octanol-water partition coefficient, and its polarizability are a few of the properties of the sorbate that interact with the solid phase to determine the amount of sorption. The sediments and soils used in this research were collected from the Missouri, Mississippi, Illinois and Ohio rivers and their watersheds. These samples provided a wide range in properties such as organic matter content, clay content and type, and hydrogen ion activity (pH) that are known to affect sorption. Sampling sites were in close proximity to potential coal gasifica- tion areas (6) . The organic compounds selected for study (Table 1.1) are representative of many of the classes of compounds found in coal conversion waste streams. In addition, they encompass a wide range of compound properties that have been shown to affect sorption. 2 TABLE 1.1. COMPOUNDS SELECTED FOR STUDY Polynuclear Aromatic Hydrocarbons Pyrene 7,12-Dimethylbenz [a] anthracene Dibenz [a 3 h] anthracene 3-Methylcholanthrene Aromatic Amines Benzidine 2-Aminoanthracene 6-Aminochrysene Aromatic Alcohols 1-Naphthol Nitrogen and Sulfur Heterocyclics Dibenzothiophene Acridine 2,2' -Biquinoline 13#-Dibenzo [a 3 £ J carbazole Aromatic Ketones Acetophenone Organic Acids Anthracene-9-carboxylic Acid 3 SECTION 2 CONCLUSIONS Sorption results when a solution component is concentrated at an inter- face (7). For sediment-water or soil-water systems the interface of interest has primarily been the solid-liquid interface. Sorption occurs when the forces of attraction between the sorbing species and the solid surface overcome both the forces of attraction between the sorbing species and the solvent (8) and any repulsive forces between the sorbate and sorbent (9,10). The sorbing species is called the solute when it is in solution and the sorbate when it is sorbed to the sorbing surface (sorbent) . For organic compounds there are two general cases where the affinity of the sorbate for the sorbent is greater than the affinity of the solute for the solvent and significant sorption results. In the first case, there is a strong specific interaction (coulombic attraction, ligand exchange or hydrogen bonding (11)) between the sorbate and the sorbent, and these forces of attraction overcome even a fairly strong attraction of the solute to the solvent. The sorption of organic cations or polar organic molecules by swelling clay minerals (12) is an example of this type of sorption. The sorption of benzidine (Section 4.3) from acid solutions is an example from this research (13). Sorption that is characteristic of the second case results not because of a large specific sorbate-sorbent interaction but rather because of a weak solute-solvent interaction. In this case, even a weak positive sorbate- sorbent interaction can overcome an extremely weak solute-solvent interaction and result in the compound being removed from solution. The weak solute-solvent interaction, that is, the low water solubility or hydrophobic nature of many organic molecules, is the result of a large decrease in entropy of the system upon solvation (14), coupled with an absence of hydrophilic functional groups or at least a dominance of the hydrophobic portion of the molecule. The attraction between a hydrophobic organic molecule and the sorbent is not the result of specific interactions as in the first case, but rather of general interactions between the sorbate and sorbent such as van der Waals forces (14). The sorption of aromatic hydrocarbons by soil or sediment humic materials is an example of this type of sorption (15,16,17,18). This second case of sorption has been referred to as hydrophobic sorption because of the emphasis on the role of the weak solute-solvent (water) interaction (14,16). Hydrophobic sorption increases as compounds become less and less polar, that is, as molecular weights, molecular volumes or carbon numbers increase 4 or as water solubilities decrease (15,18). Hydrophobic sorption has been shown to be highly correlated with the organic carbon content of the soils or sediments while at the same time relatively independent of other sorbent properties (15,16,17,18). When sorption of hydrophobic compounds is expressed as a function of the organic carbon content of the soil or sediment, a constant, Koc, is generated which is a unique property of the compound being sorbed (11,15,16,17,18) : = Kd " %0C (decimal equivalent) (2-1) where Koc is equal to the Freundlich constant (Kd) divided by percent organic carbon (decimal equivalent) in the respective soil or sediment. Thus, for these compounds, Kd values may vary dramatically from soil to soil or sediment to sediment, but the Koc values converge toward a value that is constant across all soils and sediments. For linear isotherms (1/n = 1) Kd may be expressed on a mass basis or a molar basis; for nonlinear isotherms (1/n 4 1) Kd must be expressed on a molar basis (19). For linear isotherms the best fit is often achieved by forcing 1/n to be equal to one. This has the effect of expressing all the variation between isotherms in one constant (Kd) . Small changes in 1/n can result in large differences in Kd, masking any potential correlation of Kd with sorbate or sorbent properties. The partitioning of organic compounds between water and an organic solvent, usually octanol, has been correlated with the extent of hydrophobic sorption. Increasing octanol-water partition coefficients (Kow values) for a series of compounds have been related to increased sorption when sorption is expressed in terms of Koc values (15,18): Kow = cone, cpd in octanol/conc . cpd in water (2-2) where the octanol and water phases have been equilibrated to allow partition- ing of the compounds between the two phases. A summary of the results of the sorption experiments is given in Table 2.1. The Koc values were calculated from a regression (20) of the compounds' linear or molar partition coefficients against the respective organic carbon contents (decimal equivalents) of the 14 sediment and soil samples used in the sorption experiments. The Koc values ranged from a low of 35 for acetophenone to a high of 1,668,800 for dibenz [a 3 h] anthracene. The water solubilities varied from 5,440 yg/ml for acetophenone to 0.00249 yg/ml for dibenz [a } h] anthracene. The octanol-water partition coefficients went from a low of 38.6 for acetophenone to a high of 3,170,000 for dibenz [a 3 h] anthra- cene. Individual Kp values and a more detailed discussion of the factors affecting the sorption of the respective compounds are given in the following sections. Figure 2.1 presents the relationship of Koc and Kow for the compounds studied as part of this contract with the exception of benzidine which is protonated at low pH and then behaves as an organic cation (see Section 4.3), The sorption data for the compounds studied by Karickhoff et at. (15) are also included in order to expand this relationship to a greater number and 5 o >■> . U -H o Q C/l —A ~~_> rv. o O o o O o o o o O o o o O o O M X o O o o O CU 4-1 ft co rH CM o CM rH CM CM O CM o 4J c s d rH ON 00 O ft H c/3 1 ■u o +1 +i +1 +1 + 1 +i +1 +l +1 +1 +1 +i +i +1 rH •H H +1 O 4-1 >4-| O o o O o o o o vO O O o o o C 5-1 M-l O o o O o o o o O o o o Rj CO (1) O o o o O o CM CM o 00 <£> -sT CO 4-J a o ft ft CO ft o a crj 0) CJ s o CO CO u cu cu 3 43 H 4J CO O > CO U ^ T3 C 3 O Cu B o u CO rH O O +1 m co rH CM 00 rH ON ON ON 00 00 00 oo ON O o o o O o o o o o O o O O 00 o X o o S X 4-1 CU cu c cu J3 a o •H x: 4-1 o N c cu X cu c •H X) •H >-i O <: cu (3 •H rH O c •H 3 cr •H co I o N CO X U «0 U J3 o N c cu X •H o I CO CU c o G CU X Cu o 4J CU u < o X 4J X a co I CU (3 •H •H N c cu cu (3 CU CJ CO u x 4-1 c CO o a cu f3 CU CO >N rl X o o a CM T3 •H a & o X u CO CJ I ON I CU c cu o CO u X 4-1 c < CO •H o cu u o s^s 4-1 CO f3 •H CO 00 CO a, (3 O •H CO CO CU 5-1 00 cu rl T3 CU 4J CO rH 3 a rH co o CO 4-1 3 CU •H O •H CU O O 3 O •H 4- 1 CO rH CU 5- 1 U o a TJ 3 cO • CO 4-1 CU 3 3 cu rH rH tO CO > > •H O 3 O cr ^ cu CO 4-> CU O X CU Q o cu 00 r( cu o o 4-J o 3 •d •H T3 CU 3 rH CO > CJ T3 3 3 O 3, o CM co 'd m m cu X 3 3 5-1 o O O •H •H CO 4-1 4-1 CJ O to cu cu iH C/D rH CO 3 3 O •H •H •H X 3 3 O CU CU X $ > Cu •H o 00 00 5-1 H3 cu CU >■> 5-) 5-1 X 3 3 1 X) T) 3 cu CU O CJ o 3 o o 5-1 5H Cu Cu CU iH CU CU X X X CO 4-1 4-1 CJ •H !^ to rH X X Cu Cu -o CO cu CU 3 3 4-J •H •H o e 3 cu CU II < 6 7 variety of compounds. A similar but inverse relationship between Koc and water solubilities (S) was found for both sets, as illustrated in Figure 2.2. The Koc values of the hydrophobic compounds were highly correlated with their respective water solubilities or octanol-water partition coefficients. log Koc = -0.686 log S(ug/ml) + 4.273 r 2 = 0.933 (2-3) log Koc = log Kow - 0.317 r 2 = 0.980 (2-4) The behavior of the aromatic hydrocarbons in both this and the Karickhoff et at. (15) study is not unexpected. These compounds are highly hydrophobic, lacking ring constituents or functional groups that could significantly modify their hydrophobic nature. However, the fact that the heterocyclic compounds studied as well as the compounds with reactive functional groups, with the exception of benzidine, were also hydrophobically sorbed strengthens the concept that the solute-solvent interaction is the dominant interaction controlling the sorption of these compounds. Equations 2-3 or 2-4 coupled with equation 2-1 represent a powerful tool that can be used to predict the sorptive behavior of hydrophobic compounds. Equation 2-3 or 2-4 can be used to predict the compound's Koc value based on the water solubility or octanol-water partition coefficient. The Koc value and the organic carbon content of the sorbent (soil or sediment) can then be used with equation 2-1 to predict the linear partition coefficient or molar Freundlich Kd value. The use of equations 2-1, 2-3, and 2-4 can be illustrated by calcu- lating the Koc value of parathion (0, 0-diethyl (9-p-nitrophenyl phosphoro- thioate). Parathion (CxqH^NOsPS) has a molecular weight of 291 and a reported water solubility of 12.9 yg/ml (21). log Koc = -0.686 log S + 4.273 log Koc = -0.686 log(12.9) + 4.273 log Koc = 3.511 Koc = 3244 In order to check the validity of the calculated Koc value, individual Kd values were calculated for the samples used by Wahid and Sethunathan (21) and predicted Kd values compared with their measured mass Kd values (Table 2.2) . Measured and predicted sorption constants were in good agreement and a regression of the predicted Kd values for the sorption of parathion against the measured values gave a correlation coefficient of 0.984. Regression of the measured Mass Kd values against %0C (Table 2.2) gave a Koc value of 3329 which is in excellent agreement with the predicted Kd value of 3244. 8 9 TABLE 2.2. PREDICTED AND MEASURED Kd VALUES FOR THE SORPTION OF PARATHION Mcis s XT J- CulL Leu Qamn 1 p o din p x c %or 1 /n JS.U. 10 0.44 7.67 1.04 14.27 8 0.94 12.30 1.05 30.49 11 1.67 38.02 1.11 54.17 13 3.21 125.90 1.05 104.13 15 4.77 213.80 1.03 154.70 14 14.31 457.10 1.02 464.21 Kd = Koc x %0C (decimal equivalent). TABLE 2.3. PREDICTED AND MEASURED Koc VALUES FOR THE SORPTION OF HALOGENATED HYDROCARBONS BY WILLAMETTE SILT LOAM (1.6% ORGANIC MATTER, 0.84% ORGANIC CARBON) Compound Measured Kd a Measured Koc B Solubility (yg/ml) Predicted Koc° 1, 2-Dichloroethane 0.30 36 8,450 38 1, 2-Dichloropropane 0.43 51 3,570 68 1, 2-Dibromoethane 0.58 69 3,520 69 1,1,2, 2-Tetrachloroethane 0.74 88 3,230 73 1,1, 1-Trichloreoethane 1.66 198 1,360 133 1, 2-Bromo-3-chloropropane 1.20 143 1,230 142 1, 2-Dichlorobenzene 2.88 343 148 608 Tetrach lor ethane 3.36 400 200 495 Kd = G x 0.016; Chiou et at. (22) Koc ■ Kd i 0.0084 log Koc = -0.686 log S + 4.273 10 A second example of the utility and versatility of these equations is given by comparing predicted Koc values with measured Koc values from a study by Chiou et at. (22) on the sorption of a series of halogenated hydrocarbons (Table 2.3). Correlation of the predicted Koc values with the measured values obtained by Chiou et at. gave a correlation coefficient of 0.931. The study of Chiou et at. illustrated the relationship of sorption and water solubility for one soil. The present study illustrates that this relationship can be extended to other classes of compounds and to other soils and sediments. Equations 2-1, 2-3 and/or 2-4 have tremendous potential value in calculating sorption constants for the great number of organic materials where sorption data do not exist. Where sorption constants exist for a given situation, these relationships allow extension to other soils and sediments. The predicted Koc and Kp values are good first-order approximations, but they should eventually be verified by actual sorption studies. The limits of the validity of the Koc, water solubility, octanol-water partition coefficient and organic carbon relationships (hydrophobic sorption) are not fully known. It has been demonstrated (see Section 4.2) that the limits are a function of both sorbate and sorbent properties. The concept of hydrophobic sorption may not be valid for compounds that contain hydrophilic functional groups. The presence of such groups may simultaneously increase solubility and decrease sorption, thus maintaining the established Koc-water solubility relationship (e.g., 6-aminochrysene or 2-aminoanthracene) , or they may increase solubility and dramatically increase sorption by forming a cation as was the case with benzidine. It also appears that the concept of hydro- phobic sorption may not be valid for sediments and soils that have low organic carbon contents in combination with medium to high swelling clay contents, particularly for compounds with low Koc values (see Section 4.2). 11 SECTION 3 RECOMMENDATIONS Additional research needs to be conducted to better define the adsor- bate and adsorbent properties where hydrophobic adsorption is the dominant adsorption process. If the Koc-Kow relationship (equation 2-3) and/or the Koc-S relationship (equation 2-4) are to provide as much possible information about the multitude of existing or new organic compounds, the limits of validity of the two equations must be defined. There appears to be no maximum Kow value nor minimum water solubility where the relationships are not valid other than the limits imposed by analytical sensitivity. In fact, for very insoluble organics where the limits of analytical detection are approached in sorption experiments and where impurities and degradation products present very difficult problems, the Koc- Kow and Koc-S relationships may provide more reliable numbers than the actual sorption experiments. The existing data suggest that there should exist a minimum Kow value or maximum solubility such that compounds with greater Kow values or lower water solubilities will be strictly hydrophobically sorbed. For compounds with lower Kow values or higher water solubilities the com- pounds may be strictly hydrophobically sorbed independent of soil properties other than organic carbon or the compounds may be hydrophobically sorbed only by selected soils and sediments. The acetophenone and 1-naphthol data suggest that the ratio of organic carbon to montmorillonite clay content may provide a method of predicting when hydrophobic sorption will dictate the behavior of a compound. These compounds were strictly hydrophobically sorbed above an organic carbon to montmorillonite ratio (%0C ' % montmorillonite) of 0.10 regardless of the nature of the soil or sediment. The relationship between this ratio and hydrophobic sorption needs further study to better define the critical ratio, if one exists. Much of the existing data in the published literature are in terms of mass Kd or Kp values. It is recommended before equations 2-3 and 2-4 are used that data be either in the form of linear Kp values or converted to molar Kd values by the relationship of Osgerby (19): w , „, Mass Kd x Mol. wt^^ n Molar Kd = Mol. wt (3-1) 12 SECTION 4 LABORATORY INVESTIGATIONS 4.1. ACETOPHENONE Acetophenone was chosen to represent one class of compounds, the aro- matic ketones, which have been shown to be present in coal gasification wastes (2). The physical properties of acetophenone are given in Table 4.1. TABLE 4.1. PHYSICAL PROPERTIES OF ACETOPHENONE Structure Molecular weight Melting point (°C) a Boiling point (°C) a a Density Flash point (°C) a Heat of vaporization, AHv (cal/gmol) a Aldrich catalog, 1979-80 b CRC Handbook, 1975-76 Water solubility was determined to be 5,440 ug/ml by the procedure given in Section 5.3. The octanol-water partition coefficient of acetophenone was determined to be 38.6 by the procedure given in Section 5.2. Batch equilibrium sorption isotherms were determined using li+ C- labeled acetophenone (>99 percent pure) obtained from ICN Pharmaceuticals, Inc., and ii 120.15 19-20 202 1.03 82 11,731.5 13 unlabeled acetophenone (>99 percent pure) from Eastman Kodak Co. Purity was verified using thin- layer chromatography (see Section 5.1.2). A stock solution (4,154 yg/ml) was prepared in ultrapure distilled water. The sorption iso- therms were determined in triplicate on a 2:5 sample to solution ratio (10 g soil/sediment sample and 25 ml acetophenone solution) with initial concentra- tions of 138, 277, 554, 831 and 1,108 yg/ml. Samples were equilibrated in stainless steel centrifuge tubes (aluminum foil-covered lids) in a tempera- ture-controlled shaking water bath at 25°C for 24 hours. Initial and final concentrations of acetophenone in the solution phase were determined by liquid scintillation counting. The concentration of acetophenone in the soil or sediment samples (yg/g) was determined by difference. A 11+ C mass balance was determined on selected samples to verify that there was no loss of compound from the system. Gas chromatography was used to determine if a significant quantity of the acetophenone had been degraded. Retention times for the parent compound and the compound present in the solutions equilibrated with soil or sediment samples were identical, and no unidentified peaks were observed. This was taken as evidence that no degradation of acetophenone had occurred. Sorption of acetophenone on the sediment and soil samples followed a linear trend over the entire concentration range studied. Representative isotherms are shown in Figure 4.1. Average values of the sorption isotherm data for each soil and sediment are given in Table 4.2. The sorption isotherms were described by the following equation: Cs = Kp-Cw (4-1) where Cs is the amount of compound sorbed by the soil or sediment in yg/g, Cw is the equilibrium solution concentration in yg/ml, and Kp is the linear partition coefficient. The linear partition coefficients and Koc values for the sorption of acetophenone are given in Table 4.3. The Kp values varied from a low of 0.07 for sample 8 to a high of 0.89 for sample 4. Simple correlations between Kp and selected soil properties were determined. These data indicated that the correlation between Kp and per- centage organic matter was highly significant at the 1 percent level of probability. The other factors tested, i.e., pH, CEC, % clay and % montmoril- lonite, were nonsignificant. Regression of Kp against organic carbon content (decimal equivalent) produced the following equation: Kp = 35.0(%OCJ r 2 = 0.898 (4-2) /6b where 35.0 is the Koc value for acetophenone. The Koc values obtained for the individual soils and sediments were closely grouped around the regression value of 35.0 with the exception of samples 6 and 9; these two samples have the smallest percent organic carbon to percent montmorillonite ratios. 14 FIGURE 4.1. REPRESENTATIVE ISOTHERMS FOR THE SORPTION OF ACETOPHENONE BY SOILS AND SEDIMENTS Numbers refer to soil or sediment samples. 15 TABLE 4.2. ACETOPHENONE SORPTION ISOTHERM DATA 3 Sample h Cw (yg/ml) Cs c (yg/g) Sample Cw (yg/ml) Cs (yg/g) Sample Cw (yg/ml) Cs (yg/g) B2 112 67 4 95 109 5 97 104 233 110 198 199 211 164 475 198 405 372 448 263 704 319 615 538 676 387 943 411 816 728 916 480 6 111 67 8 122 11 9 135 8 219 144 267 25 265 30 444 275 541 30 539 38 645 465 814 40 811 50 870 594 1070 94 1061 117 14 130 21 15 118 51 18 121 45 260 44 243 86 251 64 519 87 493 153 493 151 792 98 756 187 747 209 1062 115 1004 261 983 312 20 119 49 21 90 121 22 103 89 247 73 196 203 220 143 499 136 417 341 461 233 742 223 629 504 681 374 996 279 822 714 921 466 23 92 117 26 93 133 201 189 203 185 425 323 439 287 649 456 652 448 885 558 884 558 a Values are averages of triplicate determinations. Cw is the equilibrium aqueous concentration. Cs is the amount sorbed by the soil or sediment sample. 16 TABLE 4.3. LINEAR PARTITION COEFFICIENTS (Kp) AND Koc VALUES FOR THE SORPTION OF ACETOPHENONE BY SOILS AND SEDIMENTS Sample Organic carbon Kp Koc 4 2.07 0.89 43 5 2.28 0.56 24 6 0.72 0.68 95 8 0.15 0.07 48 — =c 0.11 0.09 82 « 14 0.48 0.12 25 15 0.95 0.27 28 18 0.66 0.30 46 1.30 0.29 22 21 1.88 0.85 45 22 SV/tC 1.67 0.53 31 23 2.38 0.68 29 26 1.48 0.66 45 B2 1.21 0.44 36 Comparison of samples 6 and 23 illustrates the organic carbon/clay interaction. Sample 6 gave an unexpectedly high Kp value (0.68) for its organic carbon content (0.72%). This apparently aberrant result can be partially explained by the high montmorillonite content (61%) of the sample sorbing acetophenone in excess of its organic carbon content. Sample 23, by contrast, had a similar Kp value (0.68) and yet was high in both montmoril- lonite (58%) and organic carbon (2.38%). Hence, for this sample the organic carbon appeared to mask the effect of the clay on sorbing acetophenone, and the relationship between Kp and organic carbon content remained valid. Stevenson (23) reported that the relative contributions of organic and inorganic surfaces to adsorption depends on the extent to which the clay is coated with organic substances. He considered clay-humus and clay alone as two major types of adsorbing surfaces normally available to pesticides. Hence clays should exhibit their maximum influence on sorption of nonpolar compounds in the absence of adequate organic materials for coating the clays. 17 4.2. 1-NAPHTHOL 1-Naphthol (a-naphthol) was chosen to represent the aromatic alcohols or phenolic materials found in coal conversion waste streams (2). The physi- cal properties of 1-naphthol are given in Table 4.4. TABLE 4.4. PHYSICAL PROPERTIES OF 1-NAPHTHOL Structure Molecular weight (Aldrich catalog, 1979-80) Melting point (°C) (Aldrich catalog, 1979-80) Boiling point (°C) (Aldrich catalog, 1979-80) Heat of vaporization, AHv (cal/gmol) (CRC Handbook, 1975-76) 144.17 95-96 278-280 14,205.6 Water solubility was determined to be 866 Ug/ml by the procedure given in Section 5.3. The octanol-water partition coefficient was determined to be 700 by the procedure given in Section 5.2. Batch equilibrium sorption isotherms were determined using 1 '*C-labeled 1-naphthol (>99% pure) from Aldrich Chemical Co. Purity of the original solu- tions and the equilibrium isotherm solutions was verified by thin- layer chro- matography. The sorption isotherms were determined in triplicate on a 1:10 soil to solution ratio (4.0 g soil/sediment and 40.0 ml solution) with initial concentrations ranging from 86 to 690 yg/ml. Samples were equilibrated in stainless steel centrifuge tubes (teflon-covered lids) in a temperature-con- trolled shaking water bath at 25°C for 24 hours. Initial and final aqueous phase concentrations of 1-naphthol were determined by liquid scintillation counting. The concentration of 1-naphthol sorbed by the soil/sediment phase was determined ^h^dj-fjej^n^^e^ A ^C mass balance was calculated to verify that there was no loss of the compound from the system. The C mass balance was determined by converting the compound sorbed on the soil/sediment to lt *C0 2 using a Packard Model 306 sample oxidizer. No significant loss of 1-naphthol was observed. 18 Soil thin-layer chromatography (TLC) was used to study the effect of solvent polarity on hydrophobic and nonhydrophobic sorption. Soil TLC plates were prepared by spreading a uniform layer of soil-water slurry 0.5 mm thick over 20x20-cm glass plates using the basic method of Helling and Turner (24) . The plates were dried in a desiccated chamber and spotted wi th ^C-labeled 1-naphthol using micropipette capillary tubes. Plates were developed in a chromatographic tank with the appropriate solvent and set with X-ray film in a darkroom. Rf values were determined by visual measurement of the developed film and verified by scraping and counting 1-cm segments from the developed plates in liquid scintillation vials containing 10 ml Aquasol (New England Nuclear) . The results of the sorption experiments with 1-naphthol are given in Tables 4.5 and 4.6. The isotherms were well represented by the Freundlich equation: Cs = Kd • Cw 1/n (4-3) where Cs is the amount of compound sorbed by the soil or sediment in umoles/g, Cw is the equilibrium solution concentration in umoles/ml, and Kd and 1/n are constants. The 1/n values ranged from 0.222 to 0.642, while Kd values varied from 2.60 to 30.23. Representative isotherms are given in Figure 4.2. The isotherms were non-linear, as expected from their 1/n values. Ten of the sixteen samples produced Koc values (Table 4.6) which were in good agreement with the Koc value (432) calculated from the octanol-water partition coefficient (700) of 1-naphthol. [N.B. The bulk of the present in- vestigation involved 14 soil/sediment samples. Two additional samples, #13 and //Bl, not referred to elsewhere in this report, were included in the 1-naphthol sorption isotherm determinations.] The remaining six samples gave Koc values that were much larger than the predicted value. This is further illustrated in Figure 4.3 which shows the relationship between Kd and the organic carbon content of the soils/sediments. The Kd values of ten samples were highly correlated (r 2 = 0.876) with organic carbon content, whereas the Kd values of the other six samples were not. The slope of the line in Figure 4.3 is equivalent to Koc and gave a value of 522 which is in good agreement with the predicted value. The justification for excluding the six samples with high Koc values can be seen in Figure 4.4. This figure presents Koc as a function of the organic carbon to montmorillonite ratio of the soil or sediment. For 1-naph- thol, there appeared to be two distinct families of data, One family of ten samples formed a line that was basically parallel to the %0C/%Mont axis and hence was independent of that variable. Those were the same soils and sedi- ments whose Koc values converged on the predicted value and thus appeared to hydrophobically sorb 1-naphthol. The other six samples produced a line that appeared to be a function of the organic carbon to montmorillonite ratio. The data suggest that clay may sorb 1-naphthol in sediments with low to medium organic carbon contents, but may not be a major factor controlling sorption in soils or sediments with high organic carbon contents. Comparison of samples 23 and 26 which have high montmorillonite contents, but in combination with high organic carbon contents, with the six samples that gave high Koc values 19 TABLE 4.5. 1-NAPHTHOL SORPTION ISOTHERM DATA 3 Sample Cw b Cs c Sample Cw Cs Sample Cw Cs (ymol/ml) (ymol/g) (ymol/ml) (ymol/g) (ymol/ral) (ymol/g) B2 0.06 0.37 1.20 1.95 2.90 5.40 8.32 11.93 16.43 18.82 0.03 0.13 0.58 1.33 2.00 5.72 10.65 18.16 22.59 27.82 0.21 0.57 1.45 2.32 3.17 3.95 6.33 9.55 12.83 16.89 0.010 0.023 0.234 0.875 1.52 5.88 11.73 21.59 27.15 32.62 ^ 8 0.44 0.94 1.99 3.10 4.15 1.55 2.60 3.99 4.91 6.38 0.01 0.04 0.62 1.31 2.38 5.84 11.52 17.74 22.75 24.04 14 0.24 0.56 1.18 1.86 2.53 1.32 1.89 3.18 3.92 4.75 15 0.002 0.009 0.038 0.221 0.430 2.91 5.79 11.29 15.34 19.21 18 0.05 0.22 1.00 1.92 2.98 5.56 9.83 14.06 16.49 18.31 20 0.004 0.010 0.104 0.264 0.617 2.90 5.78 10.71 14.99 17.35 21 0.19 0.58 1.41 2.29 3.26 3.73 6.15 9.87 13.03 15.28 ^22 0.23 0.60 1.43 2.34 2.59 3.71 5.97 9.66 12.45 21.94 J 23 0.08 0.32 0.92 1.96 2.68 5.25 8.81 14.88 16.45 21.26 26 0.15 0.49 1.29 2.07 3. 12 4.56 7. 10 11.04 15.20 16.62 'Values are averages of triplicate determinations. 'Cw is the equilibrium aqueous concentration. Cs is the amount sorbed by the soil or sediment sample. 20 TABLE 4.6. FREUNDLICH SORPTION CONSTANTS , r 2 VALUES AND Koc VALUES FOR THE SORPTION OF 1-NAPHTHOL BY SOILS AND SEDIMENTS Sample K (molar) 1/n (molar) r 2 Koc 4 15.99 0. 441 0.980 772 5 8. 17 0.550 0.992 358 6 30.23 0.310 0.941 4,198 8 2. 60 0.609 0.996 1,733 9 17.18 0.222 0.983 15,618 9.96 0.357 0.985 328 14 2. 81 0.563 0.991 585 15 25.65 0. 318 0.935 2,700 1j . o / u . /oj n QQ7 ? in? 20 21. 39 0.313 0.905 1,645 21 8.40 0.501 1.000 447 22 8.79 0.642 0.923 526 23 14.01 0.387 0.984 589 26 10.23 0.442 0.992 691 Bl 8.15 0.549 0.979 905 B2 9.96 0.357 0.985 823 illustrates the interaction of clay and humus. The higher organic carbon con- tents of samples 23 and 26 probably masked the effect of the clay, lowering the amount of sorption to a level controlled by the organic carbon contents of the samples. Shin (25) reported an increase in the sorption of DDT with partial removal of soil organic matter. Stevenson (23) reported that the relative contribution of organic and inorganic surfaces to sorption depends on the extent to which the clay is coated with organic substances. He con- sidered humus-clay and clay alone as two major types of sorbing surfaces nor- mally available to organic sorbates. In earlier work with the same sediment samples used in this study, acetophenone was shown to be hydrophobically sorbed by all but two of the samples, 6 and 9. The degree of departure of predicted and measured Koc values for acetophenone was much less than found with the six samples in the 1-naphthol experiments. Hydrophobic sorption was implied when there was a high degree of correlation between a compound's Kow value and sorption ex- pressed on an organic carbon basis (Koc) . The 1-naphthol and acetophenone 21 30 Cw (/i mole/ml) FIGURE 4.2. REPRESENTATIVE ISOTHERMS FOR THE SORPTION OF 1-NAPHTHOL BY SOILS AND SEDIMENTS Numbers refer to soil or sediment samples. 22 O 1 1 t 1 1 LU or t — \ o o >° CO t o > 1^- 00 ^: CD n Q LU Q — \ / O _J o O — ^ CM 00 CM LU LO 1 Q_ n < CO o Q LU Q _l O X LU CO LU < CO 1 o o O CM P H i-J CO o Pn o CO z. w H Pi o fn co Pi o w O H hJ o s PC H EH PC w Pn PC t-J > — ' M PC O T3 O U O ^ z o PC PC O EH CO w w PC i-l PC CO Z o M % •a- CO Z CO W W S3 i-h Ph Q S W < CO CO (HVIOIAI) P>\ 5=) O H 23 15,000 3,000 1,000 9,000 o 7,000 5,000 3,000 1,000 •-SAMPLES EXCLUDED FROM Koc CALCULATIONS o-SAMPLES INCLUDED IN Koc CALCULATIONS ^1 0. 0.2 0.3 0.4 0.5 0.6 %oc %MONT FIGURE 4.4. RELATIONSHIP BETWEEN Koc (FOR THE SORPTION OF 1-NAPHTHOL) AND THE %OC/MONTMORILLONITE RATIO OF SOILS AND SEDIMENTS FOR TEN SAMPLES (o) SHOWING HIGH CORRELATION BETWEEN Kd AND %0C AND FOR SIX SAMPLES (•) THAT GAVE HIGH Koc VALUES 24 data suggest that there is a lower limit (Kow value) where the correlation between Kow and Koc may not hold. The lower limit is not strictly a function of the compound's water solubility or its octanol-water partition coefficient, but is also a function of soil or sediment properties. Acetophenone has a water solubility of 5440 ug/ml and a Kow of 38.6 while 1-naphthol has a water solubility of 866 1ig/ml and a Kow of 700. Despite the more polar nature of acetophenone in aqueous solutions, acetophenone appears to be hydrophobically sorbed by soils and sediments that do not hydrophobically sorb 1-naphthol. For 1-naphthol the Koc-Kow relationship may or may not hold for samples with organic carbon to montmorillonite ratios below 0.10; for acetophenone the re- lationship appears valid until a ratio of 0.015 or less is reached. The role of the solute-solvent interaction in hydrophobic sorption is illustrated by a soil thin- layer chromatographic (24,26) study of 1-naphthol, dicamba (3 ,6-dichloro-2-methoxybenzoic acid) and 3-methylcholanthrene in dif- ferent solvent systems (Figure 4.5). In the pure water system 1-naphthol and 3-methylcholanthrene were strongly sorbed and hence showed little or no move- ment, while dicamba which is a more polar and hence more water-soluble com- pound had an Rf value close to 1.0. Increasing the percentage of ethanol in the mobile phase resulted in higher Rf values for 1-naphthol and 3-methyl- cholanthrene and decreased movement or lower Rf values for dicamba. As the percentage of ethanol increased, the mobile phase became a continually better solvent for 1-naphthol and 3-methylcholanthrene, and hence a stronger and stronger solute-solvent interaction occurred resulting in decreased sorption. The effect of the composition of the mobile phase on dicamba, a polar material, was the opposite, as expected. As the percentage of ethanol increased, the mobile phase became less polar and hence a poorer solvent for dicamba. Sorp- tion of dicamba increased and the Rf value dropped to less than 0.15 as the solute-solvent interaction weakened. The sorption of 1-naphthol on ten of the sixteen samples appeared to be an example of hydrophobic sorption. Sorption of this type is the result of a weak solute-solvent interaction and the subsequent sorption of the compound by humic materials. Factors that increase the affinity of the solute for the solvent result in decreased sorption and increased mobility of the solute. For compounds such as 1-naphthol and acetophenone the degree and type of sorption appears to be both a function of the water solubility or Kow value of the compound and a function of soil or sediment properties. One can specu- late that as the water solubilities of compounds decrease, a point is reached where only hydrophobic sorption functions and sorption can be accurately pre- dicted from Kow values. For compounds of higher water solubility they may be completely, partially or not at all hydrophobically sorbed. 25 % ETHANOLIN WATER FIGURE 4.5. EFFECT OF VARYING THE ETHANOL : WATER RATIO IN THE SOLVENT SYSTEM ON SOIL TLC R f VALUES FOR DICAMBA, a-NAPHTHOL AND 3-METHYLCHOLANTHRENE 26 4.3. BENZIDINE Benzidine (4,4'-diaminobiphenyl) and compounds of similar chemical and physiological properties are potential waste products of coal conversion plants ( 2 ) and other industrial activities. The introduction of benzidine, an aromatic amine, into the environment is reason for concern since benzidine has been identified as a potent carcinogen (27,28,29,30). The physical properties of benzidine are given in Table 4.7. TABLE 4.7. PHYSICAL PROPERTIES OF BENZIDINE Structure Molecular weight (CRC Handbook, 1975-76) 184.24 Melting point (°C) (CRC Handbook, 1975-76) 125 Boiling point (°C) (CRC Handbook, 19 75-76) 400 pKb x (Korenman and Nikolaev (31)) pKb 2 (Korenman and Nikolaev (31)) 4.3 3.3 The interaction of benzidine with clays, especially montmorillonite, has been extensively studied as benzidine forms a blue-colored complex with the clay upon sorption. The benzidine-clay complex is the result of the reversible oxidations of the benzidine by ferric iron or other electron donors with stabilization of the product by sorption to the clay. In the absence of clay or other suitable sorbents benzidine is irreversibly oxidized to a brown degradation product. The blue color of the benzidine-clay complex arises from the semiquinoidal radical cation of benzidine. A yellow quinoidal divalent cation is formed in acidic aqueous solutions of benzidine and clay. Benzidine may also form cations by protonation of the amino groups. The early research on benzidine-clay reactions has been reviewed by Solomon et at. ( 32 ) and Theng ( 33 ) . Tennakoon et dl. ( 34 , 35 , 36 ) discuss the proposed mechanism of benzidine-clay reactions. Batch equilibrium sorption isotherms were performed using llf C-labeled benzidine obtained from New England Nuclear. Unlabeled benzidine (RFR Corp. , Hope, RI) was used to adjust the activity of the labeled compound. A stock 27 aqueous solution of benzidine containing 270 ug/ml and 390Q dpm/ml was pre- pared using ultrapure water. The purity of the stock solution was determined to be at least 98% by thin-layer chromatography. The sorption isotherms were determined in triplicate using 4g:40 ml soil: solution ratio. Initial con- centrations ranged from 67 to 270 ug/ml. A few samples exhibited very strong sorption resulting in extremely low levels of benzidine left in solution. For these samples, isotherms were determined by varying the amount of soil or sediment from 0.25 g to 0.40 g and keeping the initial benzidine con- centration constant at 270 yg/ml (40 ml). The samples were equilibrated in stainless steel centrifuge tubes at 25°C for 20 hours in a temperature-controlled shaking water bath. After equilibration the phases were separated by centrif ugation and the aqueous phase was sampled for scintillation counting. The activity was deter- mined using a Packard model 3330 liquid scintillation spectrometer. The amount sorbed was calculated as the difference between initial and equilibrium concentrations. The extent of benzidine degradation in the equilibrium solutions was determined by gas chromatography. A Packard model 417 gas chromatograph with a flame ionization detector and a 2-m SE-30 column with a flow rate of 16ml/ min N2 carrier gas was used. The system was standardized using known solutions of benzidine and diphenylazine, a suspected degradation product. No extraneous peaks were observed for the equilibrium isotherm solutions. The sorption of benzidine by soils and sediments produced isotherms which were well represented by the Freundlich equation: Cs = Kd • Cw 1 ^ (4-3) where Cs is the concentration of benzidine in nmoles/g of soil or sediment, Cw is the equilibrium solution concentration in nmoles/ml and Kd and 1/n are constants. Representative isotherms are shown in Figure 4.6 and sorption constants are given Table 4.8. Average values for the sorption data are given in Table 4.9. Attempts to fit the data to the Langmuir equation gave poor fits both visually and statistically. Neither Freundlich constant was highly correlated with soil or sediment properties. This was probably due to the fact that variation between isotherms for different soils and sediments was expressed in two constants, Kd and 1/n. Small changes in the exponential constant 1/n can result in large changes in the Kd values and hence poor correlation with soil or sediment properties. To overcome this difficulty the sorption data were fit to a Freundlich equation (Kd2) where the exponential constant (1/n) had a value of 0.5. The mean of the 1/n values in Table 4.8 was 0.515. Cs = Kd 2 • Cw " 5 (4-4) The resulting Kd 2 values and correlation coefficients for the fit of the data are given in Table 4.10. The Kd 2 values were highly correlated (r 2 ■ 0.92) with hydrogen ion activities as calculated from pH measurements. 28 10,000 8,000 B2 _ 6,000 o E c O 4,000 □ 21 2,000 8 200 400 600 800 1000 Cw (nmoles/ml) FIGURE 4.6. REPRESENTATIVE ISOTHERMS FOR THE SORPTION OF BENZIDINE BY SOILS AND SEDIMENTS Numbers refer to soil or sediment samples. 29 TABLE 4.8, FREUNDLICH CONSTANTS. (Kd and 1/n) AND r 2 VALUES FOR THE SORPTION OF BENZIDINE BY SOILS AND SEDIMENTS (MOLAR BASIS) Sample Kd (molar) 1/n (molar) r 2 B2 500.4 0.423 0.91 4 570.5 0.513 0.96 5 589.4 0.468 0.97 6 1657.8 0.568 0.94 8 86.2 0.496 0.83 9 551.7 0.372 0.94 14 3941.3 0.664 0.93 15 1705.4 0.266 0.93 18 564.6 0.413 0.96 20 2332.9 0.426 0.91 21 49.6 0.694 0.97 22 73.9 0.640 0.95 23 1072.7 0.569 0.96 26 108. 2 0.656 0.89 The data suggest that the pH of the system controls the sorption of the benzidine by controlling the amount of benzidine in the ionized form in solution. The dependence of sorption on pH can further be illustrated by the results of experiments where sorption of benzidine by samples 6 and 14 was determined before and after adjustment of pH (Figure 4.7). For these experiments pH was adjusted with either concentrated NaOH or HC1, and the resultant sorption and pH values were measured after equilibration under the same sorption isotherm conditions described above. Sorption increased as pH decreased, that is, as a greater percentage of the total benzidine occurred as a charged (cationic) species. Benzidine, as already noted, can exist in solution as both an ionized (cationic) species and a neutral species. The distribution of aqueous benzidine between the two forms is a function of solution pH. Both species are subject to sorption, although the cationic form should be sorbed to a much greater extent. Karickhoff et al. ( 15 ) and Khan et al. ( 16 ) established a relationship between the octanol-water partition coefficient (Kow) of a 30 TABLE 4.9. BENZIDINE SORPTION ISOTHERM DATA 3 Sample Cw b Cs c Sample Cw Cs Sample Cw Cs (nmol/ml) (nmol/g) (nmol/m 1) (nmol/g) (nmol/ml) (nmol/g B2 17 1419 4 8 1504 5 8 1509 64 2533 20 2976 27 2901 187 4476 69 5662 138 4966 354 5984 166 7866 215 7371 467 8026 262 10084 360 9100 6 1 1577 8 87 718 9 8 1509 4 3140 210 1080 47 2705 11 6239 454 1793 218 4169 25 9272 734 2186 396 5564 47 12233 995 2746 620 6502 14 0.11 1589 15 0.63 1579 18 14 1448 1.67 3159 6 3110 39 2787 2.75 6323 78 5573 159 4764 3.88 9487 215 7377 340 6124 5.17 12649 412 8576 502 7680 20 1.42 1573 21 69 896 22 69 899 1.9 3153 149 1683 141 1764 5 6296 355 2801 364 2698 25 9276 539 4136 559 3938 111 11594 771 4987 754 5124 23 1.6 1569 26 41 1178 8 3099 102 2155 18 6171 261 3734 53 8994 410 5428 80 11902 545 7246 a Values are averages of triplicate determinations. Cw is the equilibrium aqueous concentration. Cs is the amount sorbed by the soil or sediment sample. 31 TABLE 4.10. MODIFIED FREUNDLICH PARTITION CONSTANTS (1/n = 0,5) FOR THE SORPTION OF BENZIDINE (Kd 2 ) AND IONIZED BENZIDINE (Kd 3 ) BY SOILS AND SEDIMENTS (MOLAR BASIS) VA Kci2 2 VA Kd 3 2 Samp le r r (molar) (molar) B2 339.1 0.99 3,552 0.99 4 609.8 0.99 33,542 0.99 5 498.0 0.99 18,206 0.99 6 2851.8 0.98 165,780 0.98 8 85.9 0.98 8,693 0.98 9 259.8 0.98 27,138 0.98 14 8037.2 0.98 13,063 0.98 15 473.8 0.95 26,132 0.95 18 344.5 0.99 18,352 0.99 20 1558.8 0.98 6,347 0.98 21 167.6 0.98 7,048 0.98 22 177.4 0.98 7,116 0.98 23 1790.3 0.98 28,208 0.98 26 276.7 0.96 14,335 0.96 nonpolar neutral compound and the linear partition coefficient for the sorption of the compound by soils and sediments: log (Kp/OC) = log Kow - 0.21 (4-5) where Kp is the linear sorption or partition coefficient of a compound on a particular soil or sediment and OC is the organic carbon content of the soil or sediment expressed on a fractional basis. The relationship expressed in equation 4-5 has been shown to apply to unionized hydrophobic organic compounds and is attributed to the hydrophobic or nonpolar bonding ( 14 ) of the compound to the soil or sediment organic matter (15,16). If the assumption is made that neutral benzidine will bind via hydrophobic bonding to soil organic matter, then equation 4-5 can be used to calculate the individual Kp values for the sorption of neutral benzidine by the soils and sediments. Calculations were based on the soil and sediments' organic carbon contents and the Kow (46.0) of benzidine. Three equations may 32 8 10 II pH FIGURE 4.7. EFFECT OF pH ON THE SORPTION OF BENZIDINE BY SOILS AND SEDIMENTS Numbers refer to soil or sediment samples. 33 then be written to model the sorption of benzidine by soils and sediments: bzn(sorbed) = Kp'bzn(aq) (4-6) bzn (aq) + 2H + = bzn H 2+ (aq) K = Kbi 'Kb 2 (4-7) bzn H 2+ (sorbed) = Kd -bzn H 2+ (aq) 0,5 (4-8) Using Kp values calculated from equation 4-5, H + activities calculated from measured pH values, and the total amount of benzidine added, equations 4-6, 4-7, and 4-8 can be solved to yield the Freundlich sorption constant Kd 3 (1/n = 0.5) for the sorption of the ionized benzidine by each soil or sedi- ment (Table 4. 10) . The Kd3 values thus obtained were significantly correlated with the surface area of the soils and sediments (r 2 = 0.72). Addition of organic carbon content and the interaction between surface area and organic carbon to the the regression equation improved the correlation (r 2 = 0.86). The regression coefficient for organic carbon content was non-significant while the regression coefficient of the interaction was significant (t = -2.379) and negative. This suggests that organic materials coat clay particles and lower the sorption capacity of the soil for ionized benzidine. The clay- organic matter interaction is illustrated by comparing the Kd3 values for samples 6 and 23. Both samples have comparable textures and clay mineralogy. Sample 23, however, has a higher organic carbon content which appears to mask the clay and consequently this sample sorbs substantially less of the ionized benzidine. The effect of surface area on benzidine sorption is illustrated in Figure 4.6. At any given pH, sample 6 which contains predominantly montmorillonite clays sorbs greater amounts of benzidine than sample 14 which contains predominantly kaolinite clays. The sorption of benzidine by "whole" soils and sediments was con- trolled primarily by the concentration of the ionized species. Sorption was highly correlated with pH since pH controlled the ratio of neutral to ionized benzidine in the aqueous phase. When the isotherms were corrected for sorption of the neutral species, sorption of the ionized benzidine was highly correlated with surface area and negatively correlated with organic carbon content. The organic matter appeared to coat and hence mask ionized benzidine sorption sites. These experiments suggest that extrapolation of sorption data from studies involving only clay minerals to situations involving "whole" soils or sediments may produce erroneous results. 34 4 . 4 PYRENE Pyrene, 3-methylcholanthrene, dibenz [a s h] anthracene and 7 ,12-dimethyl- benzfa] anthracene were chosen as representatives of the polynuclear aromatic hydrocarbons. The compounds represent different configurations of four and five-ring structures. The factors affecting the sorption of these compounds are discussed in Section 4.7. The physical properties of pyrene are given in Table 4.11. TABLE 4.11. PHYSICAL PROPERTIES OF PYRENE Structure Molecular weight (CRC Handbook, 1975-76) 202.26 Melting point (°C) (CRC Handbook, 1975-76) 156 Boiling point (°C) (CRC Handbook, 1975-76) 393 Density (CRC Handbook, 1975-76) 1.271 The octanol-water partition coefficient (Kow) of pyrene was determined over a range of aqueous concentrations using radiolabeled compound and the procedure outlined in Section 5.2. A Kow value of 124,000 was obtained for pyrene. The water solubility of pyrene was determined to be 0.135 yg/ml by the procedure described in Section 5.3. Batch equilibrium sorption isotherms were determined using H-labeled pyrene that had been tritiated by the method outlined in Section 5.1.3. The unlabeled pyrene used in the tritiation procedure was obtained from Aldrich Chemical Co. (>99% pure). The resulting generally-labeled 3 H-pyrene was puri- fied by microdistillation followed by preparative thin-layer chromatography. 35 An aqueous solution was prepared by evaporating (under a stream of N2 gas) an appropriate quantity of 3 H- labeled pyrene stock solution on the lower walls of a glass container, adding ultrapure water and stirring for 24 hours. The resulting solution was filtered through a 0.2u Nuclepore filter to remove any undissolved particles and diluted with ultrapure water to the desired con- centrations. The sorption isotherms were determined in triplicate on a 1:10 soil to solution ratio, with initial pyrene concentrations ranging from 10 to 80 ug/ml (the upper initial concentration representing ^59% of the maximum water solu- bility level) . Isotherm suspensions were shaken in stainless steel centrifuge tubes with teflon-covered lids in a temperature-controlled shaking water bath at 25°C for 24 hours. Initial and final aqueous phase concentrations of pyrene were deter- mined by liquid scintillation counting. The amount of pyrene sorbed by the soil/sediment phase was determined from the difference between the initial and final aqueous phase concentrations. No degradation products were found in either phase when analyzed by the procedure described in Section 5.4.1. The sorption of pyrene by the soils and sediments produced linear sorp- tion isotherms over the entire range of concentrations studied. Typical iso- therms are shown in Figure 4.8. Average Cw and Cs values for the sorption of pyrene by the soils and sediments are given in Table 4.12. The data gave good fits to the Freundlich sorption isotherm equation: Cs = Kd • Cw 1/n (4-3) Kd values (Table 4.13) ranged from 79 to 1191; the 1/n values were all close to unity. The data gave equally good fits (Table 4.1%) to the modified Freundlich equation where 1/n was forced to equal unity. Koc values calcu- lated from the linear partition coefficients (Kp) and the respective organic carbon contents of the sediments and soils are also included in Table 4.14. Regression of Kp against the organic carbon contents of the sediments and soils produced a Koc value of 63,400 (r 2 = 0.965). 36 Cw(ng/ml) FIGURE 4.8. REPRESENTATIVE ISOTHERMS FOR THE SORPTION OF PYRENE BY SOILS AND SEDIMENTS Numbers refer to soil or sediment samples. 37 TABLE 4.12. PYRENE SORPTION ISOTHERM DATA 3 Sample Cw Cs (ng/ml) (ng/g) Sample Cw Cs (ng/ml) (ng/g) Sample Cw Cs (ng/ml) (ng/g) B2 0.128 0.245 0.441 0.722 0.989 95.0 190.1 380.6 570.3 760.1 0.094 0.179 0.343 0.491 0.698 95.3 190.7 381.6 572.6 763.0 0.084 0.163 0.300 0.444 0.585 95.4 190.9 382.0 573.1 764.1 0.124 0.263 0.495 0.754 1.010 95.0 189.9 380.1 570.0 759.9 0.657 1.421 3.188 5.145 7.495 89.7 178.3 353.1 526.1 695.1 1.126 2.429 4.061 7.128 8.153 85.0 168.2 344.4 506.2 688.5 14 0.345 0.688 1.253 2.184 2.995 92.8 185.6 372.5 555.7 740.1 15 0.123 0.243 0.509 0.747 1.027 95.0 190.1 379.9 570.0 759.7 18 0.187 0.364 0.688 1.076 1.410 94.4 188.9 378.1 566.7 755.9 20 0.137 0.257 0.542 0.813 1.137 94.9 189.9 379.6 569.4 758.6 21 0.097 0.197 0.352 0.531 0.709 95.3 190.5 381.5 572.2 762.9 22 0.154 0.271 0.500 0.740 0.984 94.7 189.8 380.0 570.1 760.2 23 0.096 0.176 0.292 0.440 0.587 95.3 190.7 382.1 573.2 764.1 26 0.111 0.199 0.382 0.562 0.759 95.1 190.5 381.2 571.9 762.4 *Va lues are averages of triplicate determinations. ^Cw is the equilibrium aqueous concentration. C C8 is the amount sorbed by the soil or sediment sample. 38 TABLE 4.13. FREUNDLICH SORPTION CONSTANTS AND CORRELATION COEFFICIENTS (Kd, 1/n and r 2 ) AND THE MODIFIED FREUNDLICH PARTITION CONSTANTS (Kp, 1/n = 1) FOR THE SORPTION OF PYRENE BY SOILS AND SEDIMENTS Sample Kd 1/n r 2 Kp B2 774 1.022 0.988 760 4 1098 1.020 0.995 1065 5 1191 1.041 0.981 1155 6 633 0.943 0.964 614 8 125 0.876 0.989 101 9 79 0.953 0.979 71 14 285 0.967 0.978 277 15 783 0.977 0.994 783 18 509 0.989 0.995 504 20 747 0.987 0.986 723 21 1159 1.053 0.990 1119 22 811 1.026 0.977 806 23 1130 1.063 0.942 1043 26 1023 1.044 0.994 994 39 TABLE 4.14. LINEAR PARTITION COEFFICIENTS (Kp) AND Koc VALUES FOR THE SORPTION OF PYRENE BY SOILS AND SEDIMENTS c i & Organic Sample f Kp Koc carbon ft? .DZ 1 91 X . ZX 7 An A QQ 1 U J , J J X h H z , u / XUO_> ■J 9 9ft 1 1 R ^ ■^9 9 ^n C. o 79 A1 A R7 ftA7 o / , ot/ Q o U • i J lUi ft? O J j JJJ q nil / X 71 ftl ft / lj o xo 14 r\ to 0. 4o **> *7 "7 277 59 , 271 15 0.95 783 82,453 18 0.66 504 77,182 20 1.30 723 57,469 21 1.88 1119 61,628 22 1.67 806 48,557 23 2.38 1043 47,487 26 1.48 994 69,108 AO 4.5. 7, 12-DIMETHYLBENZ [a] ANTHRACENE 7 , 12-Dimethylbenz [a] anthracene (9 , 10-dimethyl-l , 2-benzanthracene , DMBA) was chosen as one of four compounds representing the polynuclear aromatic hydrocarbons. The physical properties of DMBA are given in Table 4,15, TABLE 4.15. PHYSICAL PROPERTIES OF 7 , 12-DIMETHYLBENZ [a ] ANTHRACENE Structure CH 3 Molecular weight 256.35 (CRC Handbook, 1975-76) Melting point (°C) 122-123 (CRC Handbook, 1975-76) The octanol-water partition coefficient of DMBA was determined over a range of aqueous concentrations using radiolabeled compound and the procedure described in Section 5.2. A Kow value of 953,000 was obtained. The water solubility of DMBA was determined to be 0.0244 yg/ml by the procedure given in Section 5.3. Batch equilibrium isotherms were determined using 1 '*C-labeled DMBA ob- tained from New England Nuclear. Purity of the radiolabeled compound was verified by thin-layer chromatography. Appropriate amounts of DMBA were plated from acetone solution onto the walls of stainless steel centrifuge tubes using the procedure of Karickhoff et at. (15) for hydrophobic compounds. These amounts represented the initial aqueous phase concentrations (2.54 to 12.27 ng/ml) that would have been present if the compound had been added as aqueous solution. Exact values for initial aqueous concentrations were deter- mined by pipetting the same amounts of stock solution into scintillation vials for counting as were pipetted into the tubes for the isotherm determi- nation. Sorption isotherms were determined in triplicate using a 4 g:40 ml soil to solution ratio. The suspensions were shaken in the centrifuge tubes with teflon-covered lids in a temperature-controlled shaking water bath at 25°C for 20 hours. The phases were separated by centrifugation . Initial and final aqueous phase concentrations of DMBA were determined by liquid scintillation counting. The amount of DMBA sorbed by the soil/sedi- 41 ment phase was calculated from the difference between initial and final radio- activity levels in the aqueous phase. Final solution concentrations were cor- rected for radiolabeled impurities and/or degradation products by the proce~- dure outlined in Section 5.4.1. The sorption of DMBA by the soils and sediments produced linear iso- therms over the entire concentration range studied. Representative isotherms are shown in Figure 4.9. Average values for the sorption isotherm data for each soil and sediment are given in Table 4.16. The sorption isotherms were described by the following equation: Cs = Kp • Cw (4-1) where Cs is the amount of DMBA sorbed by the soil or sediment in ng/g, Cw is the equilibrium solution concentration in ng/ml, and Kp is the linear parti- tion coefficient. The linear partition coefficients and Koc values for the sorption of DMBA are given in Table 4.17. The Kp values varied from a low of 562 to a high of 6777. Regression of Kp against the organic carbon content of the soils and sediments produced a Koc value of 225,308 (r 2 = 0.908). The sorp- tion of DMBA was not highly correlated with other soil or sediment properties such as pH, CEC, clay content or mineralogy. The sorption of DMBA is discussed in Section 4.7 along with sorption of the other three polynuclear aromatic hydrocarbons studied. 42 0.02 0.04 0.06 0.08 0.10 0.12 Cw(ng/ml) FIGURE 4.9. REPRESENTATIVE ISOTHERMS FOR THE SORPTION OF 7 , 12-DIMETHYL- BENZ [a ] ANTHRACENE BY SOILS AND SEDIMENTS Numbers refer to soil or sediment samples. 43 TABLE 4.16. 7, 12-DIMETHYLBENZ [a] ANTHRACENE SORPTION ISOTHERM DATA 3 Sample Cw^ Cs C Sample Cw Cs Sample Cw Cs (ng/ml) (ng/g) (ng/ml) (ng/g) (ng/ml) (ng/g) B2 0.0164 0.0225 0.0301 0.0411 0.0459 24.0 48.4 72.7 97.0 121.4 0.0145 0.0242 0.0270 0.0366 0.0403 24.0 48.2 72.7 96.9 121.4 0.0077 0.0122 0.0141 0.0182 0.0206 24.2 48.6 73.1 97.4 121.9 0.0298 0.0414 0.0604 0.0656 0.0785 23.4 47.6 71.4 95.8 119.8 0.0473 0.0770 0.1217 0.1499 0.9776 22.6 45.9 68.6 92.0 115.4 0.0246 0.0485 0.0684 0.0942 0.1055 23.0 46.0 69.3 92.2 116.0 14 0.0493 0.1016 0.1229 0.1707 0.1922 23.2 46.2 70.2 93.4 117.3 15 0.0089 0.0140 0.0194 0.0258 0.0313 24.3 48.7 73.1 97.5 121.9 18 0.0167 0.0310 0.0409 0.0518 0.0562 23.9 47.9 72.2 96.3 120.7 20 0.0169 0.0346 0.0429 0.0641 0.0673 24.0 47.8 72.1 95.9 120.3 21 0.0055 0.0095 0.0119 0.0170 0.0222 24.3 48.7 73.2 97.5 121.8 22 0.0117 0.0227 0.0282 0.0342 0.0427 24.1 48.2 72.5 96.9 121.1 23 0.0055 0.0092 0.0107 0.0129 0.0172 24.2 48.6 73.0 97.5 121.7 26 0.0089 0.0159 0.0186 0.0255 0.0313 24.3 48.6 73.1 97.4 121.8 a Value8 are averages of triplicate determinations. "Cw is the equilibrium aqueous concentration. c Cs is the amount sorbed by the soil or sediment sample. 44 TABLE 4.17. LINEAR PARTITION COEFFICIENTS (Kp) AND Koc VALUES FOR THE SORPTION OF 7, 12-DIMETHYLBENZ [a] ANTHRACENE BY SOILS AND SEDIMENTS Sample % Organic carbon Kp Koc B2 1.21 2371 195,998 4 2.07 2646 127,812 5 2.28 5210 228,499 6 0.72 1346 186,986 8 0.15 611 407,496 9 0.11 1028 934,225 14 0.48 562 117,161 15 0.95 3742 393,907 18 0.66 1895 287,196 20 1.30 1617 124,347 21 1.88 5576 296,580 22 1.67 2679 160,391 23 2.38 6777 284,743 26 1.48 3740 252,735 45 4.6. DIBENZ [a, ft] ANTHRACENE Dibenz [a 3 h] anthracene (1,2 ,5 , 6-dibenzanthracene) was chosen as a repre- sentative polynuclear aromatic hydrocarbon containing five aromatic rings. Dibenzanthracene is the least water-soluble and hence most hydrophobic of the compounds used in this study. The physical properties of dibenz [99% pure). The tritiated compound was purified by micro- distillation followed by preparative thin-layer chromatography. An aqueous stock solution was prepared by evaporating (under a stream of N2 gas) an appropriate quantity of purified 3 H-labeled acridine stock 62 solution onto the lower walls of a glass container, adding ultrapure water and stirring for 24 hours. To remove any undissolved material, the aqueous stock solution was filtered through a 0.2u Nuclepore filter. Ultrapure water was used to dilute the stock solution to the desired concentration levels for determining the isotherms. Batch equilibrium sorption isotherms were deter- mined in triplicate on a 1:10 soil to solution ratio with initial acridine concentrations ranging from 0.29 to 1.74 yg/ml. Isotherm suspensions were shaken in stainless steel centrifuge tubes with teflon-covered lids in a temperature-controlled shaking water bath at 25°C for 24 hours. Kinetic studies showed that sorption equilibrium was attained after approximately nine hours. Initial and final aqueous phase concentrations of acridine were deter- mined by liquid scintillation counting. The concentration of acridine sorbed by the soil/sediment phase was determined from the difference between the initial and final radioactivity levels in the aqueous phase. Final solution concentrations were corrected for radiolabeled impurities and/or degradation products by the procedure outlined in Section 5.4.1. Sorption of acridine by soils and sediments resulted in linear sorption isotherms over the entire concentration range studied. Average values for individual sorption isotherms are given in Table 4.28. Representative iso- therms are shown in Figure 4.13. Initially the data were fit to the Freundlich sorption isotherm equation: Cs = Kd-Cw 1/n (4-3) Since 1/n values were very close to unity for all samples, the data were fit to a modified Freundlich equation where 1/n was forced to be 1.0. A summary of these data is given in Table 4.29. Regression of Kp against the soil or sediment organic carbon content produced a Koc value of 12,910 (r 2 = 0.934). 63 TABLE 4.28. ACRIDINE SORPTION ISOTHERM DATA 3 Sample Cw^ Cs C Sample Cw Cs Sample Cw Cs (yg/ml) (yg/g) (yg/ml) (yg/g) (yg/ml) (yg/g) B2 0.027 0.065 0.086 0.126 0.156 4.6 11.9 16.4 23.4 28. 1 0.016 0.036 0.048 0.076 0.091 4.9 12.5 17.6 24. 8 30.0 0.018 0.042 0.065 0.086 0.106 4.8 12. 3 16.9 24.5 29.4 0.030 0.070 0.110 0.167 0.206 4.5 11.5 15.6 21.8 26.1 0.088 0.222 0.328 0.470 0.623 3.5 8.6 11.6 16.5 18.4 0.077 0.208 0.316 0.451 0.536 3.4 8.0 10.3 14.8 18. 1 14 0.034 0.069 0.102 0.164 0.209 4.6 12. 1 16.7 23.3 27.7 15 0.026 0.061 0.094 0.136 0.172 4.6 11.8 16.0 22.8 27.2 18 0.041 0.077 0.117 0.220 0.283 4.4 12.0 16.5 21.8 25.7 20 0.029 0.062 0.084 0.129 0.147 4.6 12.0 16.9 23.8 29.0 21 0.028 0.064 0.078 0.116 0.142 4.7 11.9 17.1 24.3 29.3 22 0.041 0.110 0.154 0.221 0.294 4.4 10.8 15.2 21.7 25.2 23 0.019 0.036 0.062 0.084 0.107 4.9 12.7 17.2 24.8 29.6 26 0.020 0.046 0.060 0.108 0.147 4.8 12.3 17.4 23.9 28.0 Values are averages of triplicate determinations. ^Cw is the equilibrium aqueous concentration. C C8 is the amount sorbed by the soil or sediment sample, 64 0.1 0.2 0.3 Cw(/i,g/ml) FIGURE 4.13. REPRESENTATIVE ISOTHERMS FOR THE SORPTION OF ACRIDINE BY SOILS AND SEDIMENTS Numbers refer to soil or sediment samples. 65 TABLE 4.29. MODIFIED FREUNDLICH PARTITION CONSTANTS (Kp, 1/n = 1) AND Koc VALUES FOR SORPTION OF ACRIDINE BY SOILS AND SEDIMENTS Sample % Organic carbon Kp Koc B2 1. 21 IOC 185 15 , 260 4 2. 07 334 16 , 160 5 2. 28 278 12 , 180 6 0. 72 132 18, 380 8 0. 15 33 21, 720 9 0. 11 34 30,560 14 0. 48 I/O 142 29,510 15 0. 95 165 17, 350 18 0. 66 101 15 , 270 20 1.30 193 14,820 21 1. 88 207 11,020 22 1.67 92 5,520 23 2.38 287 12,040 26 1.48 213 14,370 66 4.10. 2,2'-BIQUINOLINE The physical properties of 2, 2' -biquinoline (2,2'-biquinolyl) are given in Table 4.30. TABLE 4.30. PHYSICAL PROPERTIES OF 2,2' -BIQUINOLINE Structure Molecular weight 256.31 (Aldrich catalog, 1979-80) Melting point (°C) 193-196 (Aldrich catalog, 1979-80) The octanol-water partition coefficient (Kow) for biquinoline was determined over a range of aqueous concentrations using radiolabeled compound and the procedure outlined in Section 5.2. A Kow value of 20,200 was obtained. The water solubility was determined to be 1.02 yg/ml by the procedure described in Section 5.3. Batch equilibrium sorption isotherms were determined using 3 H-labeled biquinoline that had been tritiated by the method outlined in Section 5.1.3. The unlabeled biquinoline used in the tritiation procedure was obtained from Aldrich Chemical Co. (99% pure). The tritiated compound was purified by microdistillation followed by preparative thin-layer chromatography. Appro- priate amounts of biquinoline were plated from acetone solution (under a stream of N2 gas) onto the walls of stainless steel centrifuge tubes. These amounts represented the initial aqueous phase concentrations (0.38 to 1.91 yg/ml) that would have been present if the compound had been added as aqueous solution. Exact values for initial aqueous concentrations were determined by pipetting the same amounts of stock solution into scintillation vials for counting as were pipetted into the tubes for the isotherm determinations. Sorption isotherms were determined in triplicate using a 4 g :40 ml soil to solution ratio. Isotherm suspensions were shaken in the centrifuge tubes with teflon-covered lids in a temperature-controlled shaking water bath at 25 °C for several hours. Kinetic studies showed that sorption equilibrium was attained within about one hour. Initial and final aqueous phase concentrations of biquinoline were determined by liquid scintillation counting. The concentration of biquinoline 67 1 Cw(/xg/ml) FIGURE 4.14. REPRESENTATIVE ISOTHERMS FOR THE SORPTION OF 2,2'- BIQUINOLINE BY SOILS AND SEDIMENTS Numbers refer to soil or sediment samples. 68 TABLE 4.31. 2,2'-BIQUINOLINE SORPTION ISOTHERM DATA 3 Sample Cw b Cs c Sample Cw Cs Sample Cw Cs (yg/ml) (yg/g) (yg/ml) (yg/g) (yg/ml) (yg/g) B2 0.034 3.12 4 0.019 3.20 5 0.026 3.27 0.068 6.25 0.038 6.41 0.045 6.68 0.091 9.58 0.042 9.96 0.079 9.78 0.110 13.02 0.064 13.10 0.085 13.46 0. 153 15.95 0.058 16.90 0.076 17.49 6 0.028 3.30 8 0.061 2.65 9 0.065 2.78 0.042 6.93 0.112 5.53 0.115 5.88 0.060 10.46 0.161 8.43 0.160 9.08 0.079 13.96 0.204 11.48 0.197 12.42 0.099 17.43 0.233 14.80 0.236 15.74 14 0.039 2.92 15 0.020 2. 79 18 0.031 2.46 0.070 6.01 0.043 5.53 0.045 5. 26 0.103 9.07 0.062 8.34 0.078 7.68 0.138 12.09 0.078 11.23 0.096 10.38 0.170 15.18 0.101 13.95 0.115 13.08 20 0.039 3.47 21 0.019 3.54 22 0.020 3.56 0.080 6.89 0.037 7.11 0.042 7.09 0.116 10.40 0.042 11.05 0.059 10.72 0. 139 14. 16 0.068 14. 42 0.085 14.14 0.187 17.44 0.082 18.08 0.090 18.07 23 0.017 3.71 26 0.025 3.36 0.035 7.41 0.049 6.75 0.044 11.32 0.056 10.52 0.061 15.05 0.073 14.07 0.076 18.80 0.092 17.57 Values are averages of triplicate determinations. ^Cw is the equilibrium aqueous concentration, c Cs is the amount sorbed by the soil or sediment sample. 69 sorbed by the soil/sediment phase was determined from the difference between the initial and final radioactivity levels in the aqueous phase. Final solution concentrations were corrected for radiolabeled impurities and/or degradation products by the procedure outlined in Section 5.4.1. Sorption of biquinoline, like that of acridine, resulted in linear sorption isotherms. Representative isotherms are shown in Figure 4.14. Average values for individual sorption isotherms are given in Table 4.31. As in the case of acridine, the biquinoline sorption data were fit to a modified Freundlich equation. A summary of these data is given in Table 4.32. A regression of Kp against the organic carbon content of the soils and sediments produced a Koc value of 10,404 (r 2 = 0.922). As shown in Table 4.32, samples 8 and 9 had considerably higher calculated Koc values than the other samples studied. Both of these samples had organic carbon contents less than 0.2%. The possibility of error in the determination of such low organic carbon contents is large, and any such error is then magnified in those values (Koc) based on the low organic carbon content data. TABLE 4.32. MODIFIED FREUNDLICH PARTITION CONSTANTS (Kp, 1/n - 1) AND Koc VALUES FOR SORPTION OF 2,2 '-BIQUINOLINE BY SOILS AND SEDIMENTS Sample % Organic carb on Kp Koc B2 1.21 105 8,710 4 2.07 220 10,640 5 2.28 164 7,180 6 0.72 172 23,850 8 0.15 57 38,160 9 0.11 61 55,800 14 0.48 88 18,310 15. 0.95 135 14,234 18 0.66 106 16,110 20 1.30 93 7,140 21 1.88 216 11,500 22 1.67 181 10,830 23 2.38 243 10,210 26 1.48 180 12,180 70 4.11. 13#-DIBENZ0 [a, £]CARBAZOLE The third N-heterocyclic compound studied was 13#-dibenzo [a, i ] carbazole (1,2,7,8-dibenzocarbazole) . This compound is of interest because it is a known carcinogen (toxic substances list #H054250) . The physical properties of dibenzocarbazole are given in Table 4.33. TABLE 4.33. PHYSICAL PROPERTIES OF 13#-DIBENZ0 [a, i] CARBAZOLE Structure Molecular weight 267.33 (Aldrich catalog, 1979-80) Melting point (°C) 220-221 (Aldrich catalog, 1979-80) The octanol-water partition coefficient (Kow) for dibenzocarbazole was determined over a range of aqueous concentrations using radiolabeled compound and the procedure described in Section 5.2. A Kow value of 2,514,000 was obtained. The water solubility was determined to be 0.0104 yg/ml by the procedure outlined in Section 5.3. Batch equilibrium sorption isotherms were determined using 3 H-labeled dibenzocarbazole that had been tritiated by the method given in Section 5.1.3. The unlabeled dibenzocarbazole used in the tritiation procedure was obtained from Aldrich Chemical Co. (>99% pure). The tritiated compound was purified by microdistillation followed by preparative thin-layer chromatography. Appro- priate amounts of dibenzocarbazole were plated from acetone solution (under a stream of N2 gas) onto the walls of stainless steel centrifuge tubes. These amounts represented the initial aqueous phase concentrations (9.02 to 36.08 ug/ml) that would have been present if the compound had been added as aqueous solution. Exact values for initial aqueous concentrations were determined by pipetting the same amounts of stock solution into scintillation vials for counting as were pipetted into the tubes for the isotherm determination. Sorption isotherms were determined in triplicate using a 2 g:40 ml soil to solution ratio. Isotherm suspensions were shaken in the centrifuge tubes with 71 FIGURE 4.15. REPRESENTATIVE ISOTHERMS FOR THE SORPTION OF 13tf~DIBENZO- [a, i ] CARBAZOLE BY SOILS AND SEDIMENTS Numbers refer to soil or sediment samples. 72 TABLE 4.34. 13ff-DIBENZ0 [a, £ ] CARBAZOLE SORPTION ISOTHERM DATA 3 Sample Cw b (ng/ml) Cs c (ng/g) Sample Cw Cs (ng/ml) (ng/g) Sample Cw Cs (ng/ml) (ng/g) B2 0.0071 0.0116 0.0132 0.0176 0.0212 179 251 358 538 717 0.0082 0.0108 0.0166 0.0198 0.0261 179 251 359 539 718 0.0158 0.0206 0.0264 0.0372 0.0455 179 250 358 537 717 0.0182 0.0219 0.0312 0.0407 0.0545 179 250 358 537 716 0.0447 0.0613 0.0888 0.1768 0.2299 174 243 347 514 686 0.153 0.186 0.343 0.532 0.741 175 246 348 521 694 14 0.0192 0.0278 0.0497 0.0679 0.0801 177 248 352 529 708 15 0.0134 0.0158 0.0267 0.0402 0.0538 179 251 358 537 717 18 0.0114 0.0168 0.0242 0.0361 0.0537 179 251 358 537 715 20 0.0138 0.0189 0.0241 0.0350 0.0435 178 249 356 534 713 21 0.0108 0.0179 0.0188 0.0364 0.0367 179 251 359 537 718 22 0.0102 0.0131 0.0233 0.0267 0.0392 179 251 357 537 716 23 0.0135 0.0213 0.0236 0.0322 0.0539 179 249 357 537 714 26 0.0103 0.0110 0.0171 0.0233 0.0338 179 251 359 538 717 a Values are averages of triplicate determinations. b Cw is the equilibrium aqueous concentration. c Cs is the amount sorbed by the soil or sediment sample. 73 teflon-covered lids in a temperature-controlled shaking water bath at 25 °C for 24 hours. The phases were separated by centrif ugation. The initial and final aqueous phase concentrations of dibenzocarbazole were determined by liquid scintillation counting. The concentration of dibenzocarbazole sorbed by the soil/sediment phase was determined from the difference between the initial and final radioactivity levels in the aqueous phase. Final solution concentrations were corrected for radiolabeled impuri- ties and/or degradation products by the procedure outlined in Section 5.4.1. Sorption of dibenzocarbazole, like that of the other two N-heterocyclic compounds, resulted in linear sorption isotherms. Representative isotherms are shown in Figure 4.15. Average values for individual isotherms are given in Table 4.34. Modified Freundlich partition constants (Kp, where 1/n = 1) were calculated from individual isotherm values and presented in Table 4.35. A regression of Kp against soil or sediment organic carbon content produced a Koc value of 1,055,926 (r 2 = 0.830). TABLE 4.35. MODIFIED FREUNDLICH PARTITION CONSTANTS (Kp, 1/n =1) AND Koc VALUES FOR SORPTION OF 13#-DIBENZ0 [a, t]CARBAZOLE BY SOILS AND SEDIMENTS Sample % Organic Carbon Kp Koc B2 1.21 29,610 2,447,100 4 2.07 25,800 1,246,170 5 2.28 14,640 642,150 6 0.72 12,570 1,745,520 8 0.15 3,080 2,056,030 9 0.11 980 890,610 14 0.48 8,134 1,694,550 15 0.95 13,480 1,419,110 18 0.66 14,100 2,136,570 20 1.30 15,490 1,191,150 21 1.88 17,040 906,520 22 1.67 18,000 1,078,030 23 2. 38 13,900 583,930 26 1.48 21,620 1,460,730 74 The data in Table 4.35 show a 30-fold variation between the lowest and the highest Kp values; the extremes in Koc values differed by about four-fold. The dominant soil/sediment physical property accounting for sorption was the organic carbon content, a finding that was consistent for all three N-hetero- cyclic compounds studied. Coefficients for correlation of dibenzocarbazole Kp with selected soil/sediment properties are given in Table 4.36. TABLE 4.36. CORRELATION (r 2 ) OF 13//-DIBENZ0 [a, -£]CARBAZOLE Kp WITH SELECTED SOIL/ SEDIMENT PROPERTIES FOR THE 14 SOILS AND SEDIMENTS STUDIED Property r % Total clay 0. 14 % Expanding clay 0. 18 % Organic carbon 0. 66 a pH 0. 18 CEC 0. 20 Significant at the 1% level of probability. The sorption data for the representative N-heterocyclics chosen for this study gave good fits to the Freundlich sorption isotherm equation Cs = Kd-Cw 1/n (4-3) Individual Kd values varied considerably with both the soil or sediment and the specific compound. In all cases the 1/n values were quite close to unity, indicating linear sorption isotherm behavior; the data also gave good fits to the modified Freundlich equation where 1/n is forced to equal unity. Repre- sentative sorption isotherms illustrate graphically the linearity; data are given for plotting all isotherms, if desired. The dominant soil physical property affecting sorption of N-hetero- cyclics is the organic carbon content of the soil or sediment. A simple correlation of Kp with percent organic carbon content for all three compounds is significant at the 0.01% level of probability. None of the other physical properties measured (pH, CEC, % sand, % silt, % clay) were significantly correlated with Kp. The relationships between Koc and Kow and between Koc and water solubility (discussed in Section 2) appeared to be valid for all three N- heterocyclic compounds studied. The results of this study would imply that the factors controlling sorption of N-heterocyclics are not significantly different from those controlling sorption of polynuclear aromatic hydrocarbons. 75 4.12. 2-AMINOANTHRACENE Anthracene-9-carboxylic acid, 2-aminoanthracene and 6-aminochrysene were chosen as examples of substituted polynuclear aromatic hydrocarbons. The amine and carboxylic acid functional groups of these compounds greatly modify the hydrophobic nature of the parent compounds. These compounds were chosen near the completion of the contract to provide a test of the limitations of the hydrophobic sorption concept. The factors affecting the sorption of these compounds are discussed in Section 4.14 following the presentation of the sorption data for each of the substituted polynuclear aromatic hydrocarbons. The physical properties of 2-aminoanthracene (2-anthramine) are given in Table 4.37. TABLE 4.37. PHYSICAL PROPERTIES OF 2-AMINOANTHRACENE Structure Molecular weight (Aldrich catalog, 1979-80) Melting point (°C) (Aldrich catalog, 1979-80) The octano] -water partition coefficient (Kow) of 2-aminoanthracene was determined over a range of aqueous concentration using radiolabeled compound and the procedure given in Section 5.2. A value of 13,400 was obtained. The water solubility of 2-aminoanthracene was determined to be 1.30 ug/ml by the procedure outlined in Section 5.3. Batch equilibrium isotherms were determined using 3 H-labeled 2-amino- anthracene that had been tritiated by the procedure given in Section 5.1.3. The unlabeled 2-aminoanthracene used in the tritiation procedure was obtained from Aldrich Chemical Co. The tritiated compound was purified by microdistil- lation followed by preparative thin-layer chromatography. Appropriate amounts of 2-aminoanthracene were plated from acetone solution (under a stream of N2 gas) onto the walls of stainless steel centrifuge tubes. These amounts repre- sented the initial aqueous phase concentrations (0.64 to 13.65 ng/ml) that would have been present if the compound had been added as aqueous solution. Exact values for initial aqueous concentrations were determined by pipetting the same amounts of stock solution into scintillation vials for counting as were pipetted into the tubes for the isotherm determination. Sorption iso- therms were <]( t< rmined in triplicate using a 4 g:40 ml soil to solution ratio. 193.25 238-241 76 Cw(ng/ml) FIGURE 4.16. REPRESENTATIVE ISOTHERMS FOR THE SORPTION OF 2-AMINO- ANTHRACENE BY SOILS AND SEDIMENTS Numbers refer to soil or sediment samples. 77 TABLE 4.38. 2-AMINOANTHRACENE SORPTION ISOTHERM DATA 3 Sample Cw^ Cs c Sample Cw Cs Sample Cw Cs (ng/ml) (ng/g) (ng/ml) (ng/g) (ng/ml) (ng/g) B2 0.031 0.048 0.107 0.185 0.383 5.8 12.2 32.7 67.2 128.6 0.025 0.016 0.121 0.194 0.373 5.9 12.9 32.5 67.2 129.0 0.024 0.037 0.128 0.234 0.394 5.9 12.4 32.0 65.9 127.7 0.152 0.107 0.199 0.236 0.409 3.6 11.2 31.2 66.6 128.8 0.089 0.143 0.375 0.908 1.253 5.0 11.0 29.1 57.0 117.1 0.093 0.269 0.343 0.713 0.963 5.0 9.0 29.6 60.1 121.6 14 0.064 0.072 0.210 0.398 0.867 5.4 12.1 31.7 64.9 123.1 15 0.016 0.030 0.076 0.169 0.313 6.0 12.4 33.0 66.7 128.4 18 0.024 0.047 0.092 0.120 0.470 5.9 12.1 32.9 68.4 126.0 20 0.020 0.069 0.102 0. 146 0.247 5.8 11.0 31.7 66.5 128.8 21 0.023 0.058 0.077 0.123 0.230 5.8 11.7 33.0 68.0 130.7 22 0.031 0.016 0.071 0.158 0.241 5.8 12.9 33.4 67.9 131.6 23 0.012 0.031 0.051 0.088 0.131 6.0 12.2 33.3 68. 3 132.4 26 0.015 0.024 0.069 0.101 0.177 6.0 12.5 33.0 68.3 131.6 Values are averages of triplicate determinations. Cw is the equilibrium aqueous concentration. Cs is the amount sorbed by the soil or sediment sample. 78 Isotherm suspensions were shaken in the centrifuge tubes with teflon-covered lids in a temperature-controlled shaking water bath at 25°C for 24 hours. Initial and final aqueous phase concentrations of 2-aminoanthracene were determined by liquid scintillation counting. The concentration of 2- aminoanthracene sorbed by the soil/sediment phase was determined from the dif- ference between the initial and final radioactivity levels in the aqueous phase. Final solution concentrations were corrected for radiolabeled impuri- ties and/or degradation products by the procedure described in Section 5.4.1. Representative isotherms are given in Figure 4.16. The isotherms were linear and gave good fits to the following equation: Cs = Kp'Cw (4-1) The sorption data for 2-aminoanthracene are given in Table 4.38; the values are averages of three determinations. Linear partition coefficients and Koc values are given in Table 4.39. Regression of Kp against percent organic carbon in the soils or sediments gave a Koc value of 28,129 (r 2 = 0.871). TABLE 4.3 9. LINEAR PARTITION COEFFICIENTS (Kp) AND Koc VALUES FOR THE SORPTION OF 2-AMINOANTHRACENE BY SOILS AND SEDIMENTS Sample % Organic carbon Kp Koc B2 1.21 321.6 26,580 4 2.07 329.2 15,904 5 2.28 304.1 13,336 6 0.72 259.5 36,039 8 0.15 79.0 52,659 9 0.11 103.7 94,276 14 0.48 145.1 30,225 15 0.95 391.9 41,248 18 0.66 283.0 42,878 20 1.30 458.7 35,287 21 1.88 531.9 28,292 22 1.67 502.1 30,069 23 2.38 875.2 36,772 26 1.48 688.7 46,537 79 4.13. 6-AMINOCHRYSENE 6-Aminochrysene (6-chrysenamine) was chosen of one representative of substituted polynuclear aromatic hydrocarbons containing functional groups that could greatly modify their hydrophobic nature. The physical properties of 6-aminochrysene are given in Table 4.40. TABLE 4.40. PHYSICAL PROPERTIES OF 6-AMINOCHRYSENE Structure NH 2 Molecular weight 243.31 (Aldrich catalog, 1979-80) Melting point (°C) 209-211.5 (Aldrich catalog, 1979-80) The octanol-water partition coefficient (Kow) of 6-aminochrysene was determined over a range of aqueous concentrations using radiolabeled compound and the procedure outlined in Section 5.2. A value of 96,600 was obtained. The water solubility of 6-aminochrysene was determined to be 0.155 yg/ml by the procedure described in Section 5.3. Batch equilibrium isotherms were run using 3 H-labeled 6-aminochrysene that had been tritiated by the procedure given in Section 5.1.3. The un- labeled aminochrysene used in the tritiation procedure was obtained from Aldrich Chemical Co. The tritiated compound was purified by microdistillation followed by preparative thin-layer chromatography. Appropriate amounts of aminochrysene were plated from acetone solution (under a stream of N2 gas) onto the walls of stainless steel centrifuge tubes. These amounts represented the initial aqueous phase concentrations (1.81 to 36.3 ng/ml) that would have been present if the compound had been added as aqueous solution. Exact values for initial aqueous concentrations were determined by pipetting the same amounts of stock solution into scintillation vials for counting as were pi- petted into the tubes for the isotherm determination. Sorption isotherms were determined in triplicate using a 4 g:40 ml solid to solution ratio. Isotherm suspensions were shaken in the centrifuge tubes with teflon-covered lids in a 80 400 4 0.1 0.2 0.3 0.4 0.5 Cw(ng/ml) FIGURE A. 17. REPRESENTATIVE ISOTHERMS FOR THE SORPTION OF 6-AMINOCHRYSENE BY SOILS AND SEDIMENTS Numbers refer to soil or sediment samples. 81 TABLE 4.41. 6-AMINOCHRYSENE SORPTION ISOTHERM DATA a Sample Cw^ Cs c Samp le Cw Cs Sample Cw Cs (ng/ml) (ng/g) (ng/ml) (ng/g) (ng/ml) (ng/g) B2 0.015 0.029 0.055 0.062 0.193 17.6 35.3 89.2 177.6 356.6 0.008 0.025 0.032 0.064 0.104 17.8 35.5 89.9 178.9 359.5 0.017 0.012 0.032 0.038 0.084 17.5 35.9 89.9 179.7 360.2 0.040 0.039 0.083 0.144 0.323 16.6 35.0 88.2 176.2 352.2 0.051 0.073 0.150 0.296 0.579 16.4 33.9 86.0 171.1 343.7 0.049 0.080 0.137 0.274 0.466 16.5 33.6 86.4 171.9 347.4 14 0.059 0.075 0.152 0.193 0.312 16.1 33.8 85.9 174.6 352.6 15 0.020 0.055 0.072 0.124 0.268 17.4 34.5 88.6 176.9 354.1 18 0.019 0.030 0.063 0.125 0.244 17.5 35.3 88.9 176.8 354.9 20 0.033 0.076 0.103 0.187 0.394 17.0 33.8 87.6 174.8 349.9 21 0.026 0.030 0.041 0.069 0.120 17.2 35.3 89.6 178.7 359.0 22 0.026 0.082 0.104 0.111 0.191 17.2 33.6 87.5 177.3 356.6 23 0.011 0.031 0.046 0.079 17.7 35.3 179.4 360.4 26 0.035 0.057 0.066 0.120 0.169 16.9 34.4 88.8 177.0 357.4 *Va lues are averages of triplicate determinations. Cw is the equilibrium aqueous concentration. Cs is the amount sorbed by the soil or sediment sample. 82 temperature-controlled shaking water bath at 25°C for 20 hours. After equili- bration the phases were separated by centrifugation. Initial and final aqueous phase concentrations of 6-aminochrysene were determined by liquid scintillation counting. The concentration of 6-amino- chrysene sorbed by the soil/sediment phase was determined from the difference between the initial and final radioactivity levels in the aqueous phase. Final solution concentrations were corrected for radiolabeled impurities and degradation products by the procedure outlined in Section 5.4.1, Representative isotherms are given in Figure 4.17. The isotherms were linear and gave good fits to the linear partition equation (4-1). The sorp- tion data for 6-aminochrysene are given in Table 4.41; the values are averages of three determinations. Linear partition coefficients and Koc values are given in Table 4.42. Regression of Kp against soil or sediment organic carbon content gave a Koc value of 143,355 (r 2 = 0.944). TABLE 4.42. LINEAR PARTITION COEFFICIENTS (Kp) AND Koc VALUES FOR THE SORPTION OF 6-AMINOCHRYSENE BY SOILS AND SEDIMENTS Sample % Organic carbon Kp Koc B2 1.21 1,736 143,427 4 2.07 3,116 150,519 5 2.28 3,972 174,232 6 0.72 1,079 149,817 8 0.15 573 382,185 9 0.11 686 624,022 14 0.48 924 192,553 15 0.95 1,292 136,025 18 0.66 1,424 215,835 20 1.30 872 67,070 21 1.88 2,616 139,149 22 1.67 1,459 87,363 23 2.38 3,923 164,844 26 1.48 1,689 114,108 83 4.14. ANTHRACENE-9-CARBOXYLIC ACID Anthracene-9-carboxylic acid was the third compound chosen to represent the substituted polynuclear aromatic hydrocarbons. Anthracene-9-carboxylic acid forms an anion upon dissociation of the carboxyl group while 2-amino- anthracene and 6-aminochrysene form cations upon protonation of the amine groups. The physical properties of anthracene- 9-carboxy lie acid are given in Table 4.43. TABLE 4.43. PHYSICAL PROPERTIES OF ANTHRACENE-9-CARB0XYLIC ACID Structure Molecular weight (Aldrich catalog, 1979-80) Melting point (°C) (Aldrich catalog, 1979-80) C00H 222.24 214 (decomposes) The octanol-water partition coefficient (Kow) of anthracene-9-carboxy- lic acid was determined over a range of aqueous concentrations using radio- labeled compound and the procedure outlined in Section 5.2. A value of 1300 was obtained. The water solubility of anthracene-9-carboxylic acid was determined to be 85.0 ug/ml by the procedure described in Section 5.3. Batch equilibrium isotherms were run using 3 H-labeled anthracene-9- carboxylic acid that had been tritiated by the procedure outlined in Section 5.1.3. The unlabeled anthracene-9-carboxylic acid used in the tritiation pro- cedure was obtained from Aldrich Chemical Co. The tritiated compound was purified by microdistillation followed by preparative thin-layer chromatog- raphy. Appropriate amounts of anthracene-9-carboxylic acid were plated from acetone solution (under a stream of N 2 gas) onto the walls of stainless steel centrifuge tubes. These amounts represented the initial aqueous concentra- tions (3.85 to 79.94 ng/ml) that would have been present if the compound had been added as aqueous solution. Exact values for initial aqueous concentra- tions were determined by pipetting the same amounts of stock, solution into scintillation vials for counting as were pipetted into the tubes for the iso- therm determination. Sorption isotherms were determined in triplicate using a 4 g:40 ml solid to solution ratio. Isotherm suspensions were shaken in the centrifuge tubes with teflon-covered lids in a temperature-controlled shaking water bath at 25 J C for 20 hours. The phases were separated by centrif ugation. 84 0.4 20 Cw(/ig/ml) FIGURE 4.18. REPRESENTATIVE ISOTHERMS FOR THE SORPTION OF ANTHRACENE-9- CARBOXYLIC ACID BY SOILS AND SEDIMENTS Numbers refer to soil or sediment samples. 85 TABLE 4.44. ANTHRACENE-9-CARBOXYLIC ACID SORPTION ISOTHERM DATA 3 . b Sample Cw c Cw Sample Cw Cw Sample Cw Cs (yg/ml) (yg/g) (yg/ml) (yg/g) (yg/ml) (yg/g) B2 0.0014 0.0030 0.0096 0.0163 0.0366 0.0144 0.0289 0.0409 0.1362 0.1750 0.0016 0.0035 0.0095 0.0153 0.0385 0.0128 0.0233 0.0516 0.1684 0.1804 0.0014 0.0032 0.0090 0.0159 0.0299 0.0120 0.0207 0.0379 0.1233 0.2525 0.0006 0.0030 0.0089 0.0154 0.0347 0.0282 0.0261 0.0429 0.1359 0.1717 0.0009 0.0037 0.0094 0.0195 0.0380 0.0204 0.0093 0.0221 0.0385 0.0680 0.0098 0.0198 0.0378 0.0247 0.0555 0.1115 14 0.0017 0.0032 0.0087 0.0146 0.0304 0.0116 0.0292 0.0625 0.1776 0.3070 15 0.0020 0.0040 0.0116 0.0235 0.0409 0.0064 0.0146 0.0129 0.0286 0.1280 18 0.0020 0.0037 0.0094 0.0186 0.0436 0.0051 0.0162 0.0436 0.0987 0.0591 20 0.0028 0.0066 0.0144 0.0257 0.0328 0.0934 0.1721 0.3693 21 0.0008 0.0032 0.0086 0.0159 0.0342 0.0245 0.0244 0.0559 0.1402 0.2117 22 0.0012 0.0028 0.0090 0.0185 0.0368 0.0185 0.0333 0.0576 0.1142 0.2004 23 0.0018 0.0030 0.0076 0.0148 0.0303 0.0092 0.0305 0.0787 0.1700 0.2990 26 0.0003 0.0017 0.0065 0.0139 0.0419 0.0354 0.0581 0.1203 0.2352 0.2601 Values are averages of triplicate determinations. Cw is the equilibrium aqueous concentration. Cs is the amount sorbed by the soil or sediment sample. 86 Initial and final aqueous concentrations of anthracene-9-carboxylic acid were determined by liquid scintillation counting. The concentration of anthracene-9-carboxylic acid sorbed by the soil/sediment phase was determined from the difference between the initial and final radioactivity levels in the aqueous phase. Final solution concentrations were corrected for radiolabeled impurities and/or degradation products by the procedure described in Section 5.4.1. Representative isotherms are given in Figure 4.18. The isotherms were linear and gave good fits to the linear partition equation (4-1). Sorption data for anthracene-9-carboxylic acid are given in Table 4.44; the values are averages of three determinations. Linear partition coefficients and Koc values are given in Table 4.45. Regression of Kp against soil or sediment organic carbon content gave a Koc value of 422 (r 2 = 0.751). TABLE 4.45. LINEAR PARTITION COEFFICIENTS (Kp) AND Koc VALUES FOR THE SORPTION OF ANTHRACENE- 9- C ARB OXYLIC ACID BY SOILS AND SEDIMENTS Sample % Organic carbon Kp Koc B2 1.21 5.27 436 4 2.07 5.49 265 5 2.28 7.96 349 6 0.72 5.47 760 8 0.15 1.84 1,227 9 0.11 2.82 2,564 14 0.48 10.03 2,090 15 0.95 2.66 280 18 0.66 1.78 270 20 1.30 13.27 1,021 21 1.88 6.45 343 22 1.67 5.59 335 23 2.38 9.88 415 26 1.48 7.50 507 The addition of hydrophilic functional groups to a hydrophobic compound will greatly increase the water solubility of the compound. Anthracene has a water solubility of 0.073 yg/ml (15) while 2-aminoanthracene and anthracene-9- 87 carboxylic acid have respective water solubilities of 1.30 and 85.0 yg/ml. The addition of a hydrophobic group such as a methyl group will decrease water solubility (15). The hydrophilic functional group can affect sorption in different ways. If the functional group is involved in a specific reaction (e.g. , cation ex- change) with the soil or sediment, then the concept of the solute-solvent interaction dictating the degree of sorption will no longer be valid, and sorption will be much greater than predicted from the water solubility or Kow value. The sorption of benzidine is such an example. If the functional group does not enter into a specific reaction with the soil or sediment, then its effect will only be manifested in the solute-solvent interaction. In this case sorption will still be highly correlated with the compound's water solu- bility and /or Kow value. Anthracene-9- carboxylic acid, 2-aminoanthracene and 6 aminochrysene are examples of the latter case. 88 SECTION 5 EXPERIMENTAL METHODS 5.1. PREPARATION AND TRITIATION OF COMPOUNDS All of the compounds used in the present study were available com- mercially in relatively pure form. Those exhibiting a level of purity below 95% were successfully purified by the procedure outlined below. Because many of the compounds are light-sensitive, the various manipulative and analytical procedures were carried out in the presence of either darkness or subdued light. Furthermore, because many of the compounds are subject to atmospheric oxidation, all containers (e.g., tubes, flasks, developing tanks) were flooded with N2 to maintain an anaerobic atmosphere. 5.1.1. Chromatographic Analyses 5.1.1.1. Thin-Layer Chromatography The most frequently used procedure for identifying suspected compounds and determining their degree of purity in the present investigation was thin- layer chromatography (TLC). The procedure is rapid, relatively simple, reproducible, and generally reliable. The TLC developing tanks containing appropriate solvent systems (see Table 5.1) were flooded with N2 and allowed to equilibrate at room tempera- ture for about one hour. Commercially available 20x20-cm and 5x20-cm glass plates coated with 0.25 or 0.5-mm-thick layers of silica gel G (non-fluores- cent) were activated in a 100°C oven for one hour and stored in a desiccator box until used. A vertical line was scribed with a dissecting needle about one-eighth inch in from the side edges of each plate so that solvent migra- tion on the bulk of the plate would not be affected by the change in capillarity at the edge of the plate where the coating ceased. A horizontal solvent front line 15 cm above the origin (which was marked lightly with a pencil 1.5 cm from the bottom) and vertical lines demarking lanes were similarly scribed. A 5x20-cm plate accommodated up to three lanes easily, and a 20x20-cm plate a dozen or more lanes. In most cases, approximately 5 yg of compound in an appropriate sol- vent (e.g., ether) were spotted at the origin in a given lane. After development of the chromatogram in the appropriate solvent system (see Table 5.1), the plate was air-dried or gently blown dry in a stream of N2- Examination of the plate under short-wave ultraviolet (UV) light revealed fluorescent spots against a dark (UV-absorbing) background. (N.B. All compounds used in the present study fluoresced under UV light.) While the 89 CO CU S => •H i-l X CO O > U CX MH c 3 o a I U cn o o LT) co / — \ CO co CO CO m o O o o m o o o o ON 1—1 rH rH rH rl ri rl rl CU CU CU CU CU 43, 42 42 42 m CO 4-i 4-) 4-J 4J 4-J B aj cu CU » o 42 4= 42 (2 4J 4-1 4-t 4J CO 0) CU CU CU CU 42 ■u •H •H •rl •H •H 4-J C XI XI TJ TJ X) CU 0) e > rH 01 CU CU CU CU o a c c c (2 CU cO to cO cO CO (3 X X 8 CU CU CU CU N PC P2 P2 P3 C CU £: PQ CM CO C3> O 52 CO 42 4-1 r- CTi D2 cu ON * — > rH rH rH rH rH rH >>. CU o O O O O 42 (3 (2 52 (2 (2 (2 4J O CO CO cfl cO CO CU 4-1 42 42 42 42 42 ■H CU 4-1 4-) 4-1 4J 4-) X) a CU CU CU CU CU CO e s 6 CU {2 CU CU CU CU CU CU CO a 12 e f2 £2 (2 rJ CU CU cu CU CU CU CU N N N N N N P2 12 (2 (2 C2 (2 12 1 0) CU CU CU CU CU £ P3 PQ PQ PQ PQ PQ cu (2 cu CJ CO M CU 42 rH 4-1 o £2 N CO CU cO e cu 43 CU (2 U CJ CU CO N CO M a 12 M 42 CU CU CU 42 4-1 (2 C2 43 4-1 c CU •H •> rH (2 CO 42 rH cO rH Ph o CU 42 o o a o c 4-) 42 •H •H N o cu a 42 t2 C e rH 4-J CU XT CU CU ■H >S O q •rl 43 42 CU Q N 42 N •H PQ •H a C 1 c 4-1 (2 X) 1 O o CU CM CU CU CU •rH 4-J H fH 43 X 43 M CN CU •H •H CJ CO CJ P- Q cn Q <2 CM r-l < o 42 4-1 42 (X CO S3 I CU C2 •rl X) •H N C CU PQ CU £2 CU CJ CO U 42 4-» d co o (2 •H CU C2 CU CO >> r( 42 O O C2 •rl XI •H O CO CJ •rl rH ft O 43 r( cO CJ I cr« I CU (2 CU CJ cO 42 4-> I CM I M3 90 plate was still under the UV light, outlines of spots were marked with a pencil. The Rf value of a compound was calculated as the distance (in cm) from the origin to the center of the spot divided by the distance from the origin to the solvent front. In the indicated solvent systems, the Rf values listed in Table 5.1 were expected. 5.1.1.2. Gas-Liquid Chromatography The second most frequently used procedure for identifying suspected compounds, measuring the degree of purity of the compounds, and quantitating compounds in the present investigation was gas-liquid chromatography (GLC). Like TLC, this procedure is rapid, relatively uncomplicated, reproducible and generally reliable. A Varian Aerograph Series 1860 Gas Chromatograph with flame ionization detector was utilized for determinations involving a 12-foot OV-17 column. The hydrogen flow rate for this instrument was maintained at 32 ml/min, the air flow rate at 270 ml/min, and the nitrogen (carrier gas) flow rate at 25 ml/min. The injection port temperature was 225 °C, and the detector temperature 300°C. Column temperatures were either maintained isothermally or programmed at a rate of 8° per minute, depending upon the compound. A Varian Aerograph Series 2700 Gas Chromatograph with flame ionization detector was used when a Carbowax column or a 6-foot OV-17 column was pre- ferred. Both chromatographs were operated in conjunction with a Hewlett- Packard Model 3380A Recording Integrator. The hydrogen flow rate for the Series 2700 instrument was maintained at 25 ml/min, the air flow rate at 230 ml/min, and the nitrogen flow rate at 25 ml/min. The injection port temperature was 175°C, and the detector temperature 310°C. The temperatures of both columns were maintained isothermally. Silanized six and 12-foot glass columns were packed with 3% OV-17 on 80/100-mesh Supelcoport (Supelco, Inc., Belief onte, PA). These two columns were used for the detection and quantitation of most of the compounds in the present study. Under the indicated conditions, the retention times presented in Table 5.2 were expected. A silanized 6-foot stainless steel column was packed with 0.2% Carbowax 1500 on 60/80-mesh Carbopack C (Supelco). This column was used primarily for the detection and quantitation of acetophenone and small quantities of a-naphthol in water samples. Under the indicated conditions, the retention times of a-naphthol and acetophenone were ^24 and ^30 minutes, respectively. Because the retention times were rather long, triplicate injections could be made at 5-minute intervals, thus minimizing the time necessary for replicated results. 5.1.2. Purity and Purification Procedures 5.1.2.1. Determination of Specific Activity and Radiochemical Purity All radiolabeled compound preparations purchased from commercial sources or tritiated in the laboratory were analyzed to determine the exact 91 c • o X vH O 4-1 u 3 fX CD 3. j-i h <3 CD H B 0) n 3 4-1 ^— s cO C_> u o i CD H 3 B 3 I— I O U 03 CD CD Jh 4= M ►a CD 4J 00 •p Cu O O o o CO u CO M H 3 3 O a § CN CO m o o CO co O CO CO B ri CD X 4J O CO l-l CD 3 CD O CO U X 4J 3 c0 c CD rH 4J CD CD 3 CD O CO M X 4-1 3 CO CD C CD r» >s P-. 5 £ a i CM g CD X CD 3 CD Ih x 4-1 s CO iH O X O rH ^ X 4J CD X I co CD 3 CD X cx o •H X 4-1 o N 3 CD X CD 3 •H 13 •H M CJ < CD 3 •H iH O c •H 3 cr •H « I O N CO X M CO cj o N c CD X CO CD 3 O 3 CD X a o 4-1 CD CJ < O X 4-1 X a. cO SB I CD 3 •H T3 •H N 3 CD pa CD 3 CD cj CO U X u 3 3 a •H O CM CO B u CD X 4-1 o CO CD 3 CD CO ■8 o c •H CM O CD x 4-1 O CO ✓ — s — s ✓~\ /-— N /— \ * — ' r— «. /— s CO CM CM CM CM CM CM 3 iH rH i— 1 rH vO rH M rH X vD vO v£5 tun v ' V — ' * — / > / wa wa rH h» o o o rH rH rH iH rH rH rH rH x> rH X rH i-H rH rH 1 1 1 1 1 1 1 1 rH 1 rH 1 1 1 > > > > > CO > CO cS o O o o o u o U o •H O CO O •H rH fr O X rH CO O I CTv I CD 3 CD a co M X 4J r! CO CD f-J 1 1 o 00 ^ — ' vD 3 o o m rH rN CO »> s CI r— ' }j CO C_5 II II bd § ti cd r i 4J o o (J rH it) ni V-M 1 t »— ' rj n J- c CO O J^J 1 flj rv /•— \ w w CD CX5 (H CD C & o CO CD 4-1 CO U o O CO CJ 00 II CO O D- U O a, 92 amount of radioactivity (in disintegrations per minute, dpm) and the total weight of compound contained therein. In addition, each compound was analyzed for radiochemical purity by means of TLC (Table 5.3). The radiolabeled compound was dissolved or diluted in an appropriate solvent (e.g., acetone) in a 15-ml screw-capped centrifuge tube at the rate of 50 yCi/5 ml. In order to minimize evaporation and maximize compound stability, this stock solution was stored under N2 at 0°C between uses. Triplicate one-ul samples of the stock solution were subjected to liquid scintillation counting in order to determine the exact number of dpm present in the solution. In the case of radiolabeled compounds purchased from commercial sources, the total mg of compound received was calculated from the specific activity provided by the source and the dpm determined above. In the case of compounds tritiated in the laboratory, the total mg were determined by GLC using pure unlabeled compound as the standard. The weight was verified gravimetrically by evaporating a known quantity of solu- tion to dryness in a tared tube under a stream of N2 and reweighing the tube and contents in order to obtain the net weight. From the dpm and the weight values, the specific activity of the tritiated compound was calculated. A TLC tank containing the appropriate solvent system (see Table 5.1) was flooded with N2 and allowed to equilibrate at room temperature for about one hour. The stock solution was spotted at the rate of ^10,000 dpm/spot along with a small amount of unlabeled standard solution on a 5x20-cm 0.25- mm - thick silica gel G plate. The plate was developed to the solvent front and then air-dried or gently blown dry in a stream of N£. The location of the unlabeled standard was visualized under UV light and recorded. One-cm segments from the solvent front to the bottom of the plate were scraped into vials of Aquasol (New England Nuclear, Boston, MA) for liquid scintillation counting. The quantity of radioactivity in the segments corresponding to the unlabeled standard was calculated as a percentage of the total radioactivity in the sample lane. If the compound was >95% pure, no further purification was necessary. If the purity was <95%, the compound was subjected to the purification procedure outlined below. 5.1.2.2. Purification of Compounds The following modifications of the analytical thin- layer chromatography procedure for determining radiochemical purity (described above) were suf- ficient to convert it into a preparative procedure for purification of the compounds used in this study. The. compound in question was banded (without the presence of unlabeled standard) at the rate of about 15 yCi per 20x20-cm plate for radiolabeled compounds or 100 mg per 20x20-cm plate for unlabeled compounds on a 0.5-mm-thick silica gel G plate. After development of the chromatogram, the fluorescent band with the appropriate Rf value (see Table 5.1) was stipple-outlined by means of a dissecting needle. (This band was always the major band on the plate.) The band was immediately scraped into a 50-ml screw-capped centrifuge tube, and the compound eluted from the silica gel with three 20-ml volumes of appropriate solvent (e.g., acetone) under N2. 93 S>> 3 ■u u - •H CU 4H /»""> Jn 4-> «H &-S 3 4-1 )-i ^ P* CO 3 Ph 3 -H 4-1 •H X) Sh - 3 4J Pl, 0) — X CO rJ X) >, QJ 4-1 -H •H 4-1 U Ph OJ o M 4-i 3 •H O CJ CO 0) a CO X cu o u 3 O CO 3 o rH -H CU 4-1 X -H CO CO rJ O Ph 1 3 O Du 6 O CJ cu c cu J-l >» Pu X) vO CO m vO vD CN o £> C=> Ph oo ON 00 Ph Pm Ph ON CTi CTi ON ON Cu Ph ON ON ON vO vD co CN vX co m r»» CO o CN CN On O ON — ' •, <—i ^— > rH 1— 1 — ' II rH X l 33 33 pd 1 CJ cc Pd co ' 4-1 cu B CO no CO CO CO co rH , X 4-1 CU a i CN cu d cu CJ CO >H cu x 3 4-1 cu s CJ co CO — > >H rt. X •> 4-1 « c <0 N c cu •H Q CU c N X 4-1 CU Z I CO cu c CU X (X o •H -C 4J O N c cu rO cu c •H TJ •H M O •8 4-1 3 O X) 0) •H r4 r4 CO o C o •H 4-1 CO •H 4-1 •H Sh 4J > •H Q CU a o 4-> O CO •H O •H X) cO xl C CO cO O •H o 3 O >iH rJ 4-1 CO cO CU GO rH -H CJ 3 z X) 3 cO rH 4-1 oo 3 3 CU W CO cu cu z >H (X 00 3 •H >H 3 X) r4 CO CO CO CU CJ cu C 3 3 3 O •H 4-1 cO CJ •H 4H •H r4 3 a, >H cu X! 4J 3 C3 Ph Ph X X) QJ 4J Cfl •H 4J •H CU >H QJ ■? QJ Pi •H X) •H a co QJ X o rH 4-1 O CO \ o cj^ x^ 9A The eluates were combined and concentrated to a small volume under a stream of N2- Any residual silica gel present in the concentrated eluate was removed by centrif ugation. The eluate was then subjected to analytical TLC to check for purity. All compounds purified in this manner showed only a single spot on TLC and a single peak on GLC; all radiolabeled compounds were >95% pure. 5.1.3. Tritiation Five of the compounds selected for investigation were available commerciallly in 11+ C-labeled form and one in 3 H-labeled form. The remaining compounds were not available in radiochemical form except by custom synthesis which is expensive. Tritiation of the latter compounds by a modification of the procedures of Hilton and O'Brien (40) and Lu et at. (41) proved to be an effective and relatively inexpensive means of radiolabeling these compounds. Five millimoles (0.56 g) of P2O5 were placed in a 5-ml round-bottom flask that was fitted with a desiccating U-tube and chilled in a slurry of acetone-dry ice in a small crystallizing dish. By means of a disposable syringe and needle, 12 millimoles (0.21 ml) of 3 H20 were allowed to run down the side of the flask where the water froze. The flask was removed from the slurry bath so that the water gradually melted and reacted with the P2O5 to form tritiated phosphoric acid. When the reaction was complete, the reaction mixture was bubbled with BF3 gas for about 10 minutes. A magnetic stirring bar was added to the flask, followed by an appropriate quantity of tritiatable compound (50-500 mg) dissolved in a small quantity of a suitable organic solvent (about one ml). Characteristics of a suitable organic solvent include immiscibility with water, a minimum number of exchangeable hydrogens, and the capacity to dissolve the tritiatable compound at a level of 50-500 mg/ml. The reaction mixture was stirred overnight at room temperature with the drying tube in place; the hydrogen- tritium exchange occurred during this period. The reaction mixture was transferred to a 200-ml round-bottom flask; 50 ml of water and 50 ml of solvent, part of which had been used to rinse the reaction vessel, were added along with a magnetic stirring bar. The flask was loosely stoppered (small piece of aluminum foil inserted between stopper and flask) and the mixture stirred overnight at room temperature. The flask contents were transferred to a 500-ml separatory funnel; the aqueous phase was removed and saved. The solvent phase was extracted with two 100-ml vol- umes of water-solvent (2:1, v/v) ; the solvent phases were transferred to a 500-ml round-bottom flask. The combined aqueous phases were extracted with about 150 ml of solvent, the aqueous phase discarded, and the solvent phase added to the round-bottom flask. Sufficient methanol was added to provide a total volume of about 350 ml. The flask was placed in a heating mantle over a magnetic stirrer, a stirring bar added, and a microdistillation apparatus connected. Consecutive 25-ml fractions were collected and monitored for radioactivity. In cases where the initial solvent was cyclohexane, the distillate was biphasic; both phases were monitored for radioactivity. Distillation was continued until the radioactivity in the distillate reached a low plateau, at which point the initial solvent had been distilled off completely. During distillation, for 95 every 50 ml of distillate collected, 50 ml of methanol were added to the distillation flask. After the low radioactivity plateau was reached, the microdistillation apparatus was disconnected, a few grams of anhydrous Na2S0i + were added to the distillation flask, and the contents were allowed to stand for a few hours at room temperature to dry off residual water. The liquid phase was transferred to a fresh 500-ml flask, followed by several rinses of the previous flask, and evaporated to near-dryness on a rotary evaporator. The flask contents were transferred, with rinses, to a 15-ml screw-capped centrifuge tube. The solution was concentrated to a reasonable volume (e.g., 10 ml) under a stream of N£. Radiochemical purity of each tritiated compound was determined by the TLC procedure outlined earlier in this section. If not ^99% pure, the compound was purified by the preparative TLC procedure described earlier in this section. Aliquots of a solution of the pure compound were subjected both to liquid scintillation counting in order to determine radioactive content and to GLC along with standard solutions to determine concentration. A known volume of the solution was evaporated to dryness in a tared tube under a stream of N2 and the weight of the compound per given volume of solution verified. The specific activity of the tritiated compound was calculated. For all compounds tritiated during the present study, tritiation involved the exchangeable hydrogens on the aromatic rings. The rather strin- gent conditions inherent in the tritiation procedure were sufficient to remove any label from the substituent groups of 6-aminochrysene, 2-aminoanth- racene and anthracene-9-carboxylic acid. Nevertheless, in order to verify that no residual tritium remained on the substituent groups, the following procedure was utilized. Two small aliquots (e.g., 0.2 ml) of a methanolic solution of each tritiated amino-substituted compound were diluted to 5 ml with water (pH 6). One solution was adjusted to pH 10-11 with IN NaOH in order to ionize the substituent group, mixed well, allowed to stand for a few minutes, and then readjusted to pH 6 with IN HC1. Dilution of the carboxylic acid-substituted compound in water (at pH 6) was sufficient to ionize the substituent group, and therefore only a single solution of this compound was prepared. The solutions were extracted three times with equal volumes of methy- lene chloride. The volume of the extracted aqueous phases remained at 5 ml; the volumes of the extracts were recorded. Both the aqueous phases and the extracts were sampled in triplicate (1.0 ml for the former, 0.2 ml for the latter) for liquid scintillation counting. The percent radioactivity remaining in the aqueous phase after extraction was the same for both solu- tions (pH-adjusted and unadjusted) and amounted to less than 0.2% of the total radioactivity present in the extract and the extracted aqueous phase. If the substituent groups of the three compounds had been tritiated, the pH- adjusted aqueous phases would have contained 5-10% of the total radioactivity after extraction. 5.1.4. Radioactivity Measurements The radioactive content of samples, fluid or dry, was measured by 96 liquid scintillation counting in a Packard Model 2425 Tri-Carb Liquid Scintillation Spectrometer. Fluid samples included stock solutions, extracts, partition phases, eluates, filtrates and distillates. Scrapings from chromatogram plates were the only dry samples counted directly. Aquasol was used as the counting cocktail for all direct samples, and Permafluor V (Packard Instrument Co., Inc., Downers Grove, IL) or Monophase-AO (Packard) for samples processed in a Packard Model 306 Sample Oxidizer. The latter samples were primarily soil and sediment phases from sorption isotherm determinations. ll+ C from oxidized samples was collected as ll+ C02 in Carbo- Sorb (Packard) and added to Permafluor V for counting; 3 H from oxidized samples was collected as 3 H20 directly in Monophase-40. All samples were counted for 10-minute periods. All counts were corrected for background radioactivity and counting efficiency. 5.2. OCTANOL-WATER PARTITIONING PROCEDURE The partitioning of a solute between two immiscible solvent phases has become a handy reference characteristic. In addition, when the partitioning of the solute into one phase (especially a volatile phase) is very much greater than into the other phase, the technique becomes a useful means of solute extraction. The literature contains numerous articles describing partitioning methodology and interpretations. Portions of the methods of Leo et at. (42), Chiou et at. (43), Karickhoff et at. (15) and Karickhoff (44) were utilized in the present investigation to develop a procedure (Figure 5.1) that gives reproducible results with organic compounds possessing limited water solubility. For compounds with known or predicted low water solubility, such as the polynuclear aromatic hydrocarbons (PAH) , the pure radiolabeled compound was used without the addition of unlabeled compound in order to provide significant levels of radioactivity in the aqueous phases. For compounds with moderate- to-high water solubility, such as acetophenone or a-naphthol, a sufficient quantity of pure unlabeled compound was added to the radiolabeled compound to avoid excessively high levels of radioactivity in the octanol phases . Using a predicted Kow of the compound to calculate the quantity of compound necessary to provide a statistically significant level of radio- activity in the aqueous phase, a sufficient quantity of radiolabeled compound was dissolved or diluted in a suitable solvent (e.g., acetone). An appropri- ate amount of solution was shell-evaporated under a stream of N2 onto the lower walls of each of triplicate foil-covered screw-capped Erlenmeyer flasks; 50-ml flasks were used for volumes ranging from 10 to 15 ml, and 500-ml flasks for volumes ranging from 100 to 300 ml. Five ml of 1-octanol that had been purified by the method of Karickhoff et at. (15) were added to each flask. The flasks were shaken at moderate speed on a Burrell Wrist-Action shaker at room temperature (24°C) for one to two hours to ensure full dissolution of the compound in the octanol. Octanol-saturated ultrapure (>10 megohm/cm) water was added to each flask at the rate of 5-10 ml for moderately water-soluble compounds and 100-300 ml for the less water-soluble compounds. The flasks were flooded with N2 and shaken at moderate speed for 15 minutes. 97 Pure radiolabeled compound i Dissolve in suitable solvent I Shell-evaporate (under ^) on lower walls of triplicate foil-covered screw-capped 500-ml Erlenmeyer flasks i Add 5 ml purified 1-octanol to each flask i Shake (wrist-action shaker), moderate speed, 1-2 hours, room temperature I Add 300 ml octanol-saturated ultrapure (>10 megohm/cm) water to each flask; flood flasks with N£ I Shake at moderate speed, 15 minutes, room temperature I Transfer upper portion of mixture to glass (Corex) centrifuge tubes; centrifuge at 30,000xg, 20 minutes, 24°C; discard lower (uncentrif uged) portion of mixture i Repeat centrif ugation if interface is not clear I Sample octanol phase (10 yl in triplicate) for liquid scintillation counting; remove and return octanol phase to original Erlenmeyer flask I Remove and discard all residual octanol at interface (including some aqueous phase, if necessary) I Carefully sample aqueous phase (1.0 ml in triplicate) for liquid scintillation counting (make sure that pipette is wiped free of any residual octanol) 1 Calculate Kow (Kow = dpm/ml octanol phase t dpm/ml aqueous phase) I Repeat partitioning procedure with fresh octanol-saturated ultrapure water until Kow value remains constant (indicative of compound purity) 1 By means of HPLC, determine % of radioactivity in aqueous phase actually representing parent compound; correct Kow value accordingly FIGURE 5.1. PROTOCOL FOR DETERMINING THE OCTANOL-WATER PARTITION COEFFICIENT OF A HYDROPHOBIC ORGANIC COMPOUND 98 The full contents of the small flasks were transferred to 15-ml Corex centrifuge tubes for centrif ugation. The contents of the large flasks were transferred to appropriate-size Erlenmeyer flasks. When the bulk of the octanol had risen to the surface, the upper portion of the mixture (including all of the octanol phase and about 25 ml of the aqueous phase) was divided between two 15-ml Corex tubes; the remaining aqueous phase was discarded. Tubes and contents were centrifuged at 30,000xg for 20 minutes at 24°C. If the interface between the two phases was not clear, centrif ugation was re- peated. The octanol phases were sampled (10 ul in triplicate) for liquid scintillation counting, and then carefully removed and returned to their original Erlenmeyer flasks. All residual octanol, with some water phase, if necessary, was re- moved from the centrifuge tubes and discarded. The aqueous phases were carefully sampled (1.0 ml in triplicate) for liquid scintillation counting; during this sampling, pipettes were wiped free of any residual octanol. The remaining portions of the aqueous phases were carefully transferred to clean vials in preparation for degradation checks, if necessary. The octanol-water partitioning coefficient (Kow) for each partitioning was calculated by dividing the dpm/ml of octanol phase by the dpm/ml of aqueous phase. The partitioning procedure was repeated with fresh octanol- saturated water until the Kow values remained constant (indicative of com- pound purity in the octanol phase) . Once the Kow values for the partitioning of a compound had become constant, the octanol phase was analyzed for the presence of degradation products by the procedure outlined in Section 5.4.2. In all cases purity of the compound dissolved in the octanol phase was >99%. The aqueous phase saved from the last partitioning was analyzed (by the degradation procedure described in Section 5.4.2) for the percent of radioactivity actually representing the parent compound. The level of radioactivity in the aqueous phase was adjusted accordingly and the Kow value corrected. The octanol- water partition coefficients determined by the above procedure for the 14 compounds in the present study are presented in Table 5.4. 5.3. WATER SOLUBILITY DETERMINATION Many procedures for determining the water solubilities of organic compounds have been recorded in the literature. Portions of the methods of Haque and Schmedding (45) and Mackay and Shiu (46) were utilized in the present investigation to develop a procedure (Figure 5.2) that gives repro- ducible results with organic compounds varying from relatively soluble to almost insoluble in water. A quantity of pure radiolabeled compound representing at least a 10- fold excess (per 900 ml of water) over the predicted solubility level was dissolved or diluted in an appropriate solvent (e.g., acetone) or measured without dilution if already liquid (e.g., acetophenone) . For compounds with a predicted low water solubility, such as PAH, the radioactive stock solution was used without the addition of unlabeled compound. For compounds with moderate- to-high water solubility, such as acetophenone or a-naphthol, a 99 TABLE 5.4. OCTANOL-WATER PARTITION COEFFICIENTS (Kow) OF ENERGY-RELATED ORGANIC POLLUTANTS Compound Kow Pyrene 124,000 + 11,000 7, 12-Dimethylbenz [a]anthracene 953,000 + 59,000 Dibenz [a, h] anthracene 3,170,000 + 883,000 3-Methylcholanthrene 2,632,000 + 701,000 Dibenzothiophene 24,000 + 2,200 Acridine 4,200 + 940 2,2' -Biquinoline 20,200 + 2,200 13#-Dibenzo [a 3 i ]carbazole 2,514,000 + 761,000 Acetophenone 38.6 + 1.2 1-jNapntnol inn / uu bz Benzidine 46.0 + 2.2 2-Aminoanthracene 13,400 + 930 6-Aminochrysene 96,600 + 4,200 Anthracene-9-carboxylic acid 1,300 + 180 sufficient quantity of pure unlabeled compound was added to the radioactive stock to avoid producing an aqueous solution with an excessively high (wasteful) level of radioactivity. The dissolved or liquid compound was divided equally between nine foil- covered screw-capped> 250-ml Erlenmeyer flasks, thus providing at least a 10- fold excess per flask over the predicted solubility level. The solution was shell-evaporated (under N2) on the lower walls of the flasks. One hundred ml of degassed (boiled) ultrapure (>10 megohm/cm) water were added to each flask. Flasks were flooded with N2 , capped, and shaken at moderate speed on a Burrell Wrist-Action shaker at room temperature (24°C). Triplicate flasks were removed at 24, 48 and 72 hours; for all of the compounds in the present study, saturation was achieved within 24 hours. In order to ensure that any increase in solubility was not due to degradation of the parent compound, a 100 Pure radiolabeled compound (allow at least 10-fold excess/flask over predicted solubility level) I Dissolve in suitable solvent i Shell-evaporate (under N2) on lower walls of several (e.g., 9) foil-covered screw-capped 250-ml Erlenmeyer flasks I Add 100 ml degassed ultrapure (>10 megohm/cm) water to each flask I Flood flasks with N2 1 Shake (wrist-action shaker) , moderate speed, room temperature 1 Remove triplicate flasks at 24, 48 and 72 hours (or later, if saturation level not reached yet) I Filter several successive 5-ml samples from flask through same 0.2y Nuclepore filter; save filtrates I Sample each filtrate (1.0 ml in triplicate) for liquid scintillation counting 1 Calculate solute concentration in each filtrate in terms of ppm or ppb; average the values of successive filtrates containing similar solute levels (disregard any early filtrates reflecting adsorption of solute onto filter membrane) I Perform degradation check on 10-ml sample from at least one flask each day to ensure that any increase in solubility is not due to compound degradation FIGURE 5.2. PROTOCOL FOR DETERMINING THE WATER SOLUBILITY OF A HYDROPHOBIC ORGANIC COMPOUND 101 10-ml sample from at least one flask each day was analyzed for degradation products by the procedure outlined in Section 5.4.3. The flask contents were sampled (1.0 ml in triplicate) for liquid scintillation counting. Several successive 5-ml samples from a flask were filtered through the same 0.2y Nuclepore filter. (For all compounds in this study, 0.2y and 0.08y Nuclepore filters gave identical results.) Each filtrate was saved and sampled (1.0 ml in triplicate) for liquid scintillation counting. Sequential filtrations were continued until the level of radio- activity in the later filtrates remained constant. The first two or three filtrates of the series usually contained less radioactivity than the sub- sequent filtrates, probably because the compound was being adsorbed onto the surface of the filter. For most compounds, the level of radioactivity in the filtrates was less than in the unfiltered solution, probably reflecting the retention of aggregates by the filter. From the specific activity of the compound added to each flask and the level of radioactivity present in the filtrates, the solute concentration in each filtrate was calculated in terms of ug/ml (parts per million, ppm) . For all nine flasks, the values of successive filtrates containing similar solute levels were averaged. The water solubilities determined by the above pro- cedure for the 14 compounds in the present study are presented in Table 5.5. 5.4. DEGRADATION STUDIES It was noted early in the investigation that most of the compounds were reasonably stable to atmospheric and photooxidation when present in high concentration either in solution or on a sorbing surface. However, these same compounds in dilute solution (e.g., aqueous solution) were much more susceptible to degradation. In addition, even after purification, many of the compounds tended to retain a small percentage of impurity (<3%) that was usually a polar oxidation product and thus tended to stay in the more polar or aqueous phases during any procedure. Because liquid scintillation counting of a radioactive sample does not distinguish between parent compound and con- tamination or degradation products present in the sample, it was necessary to determine what percentage of the radioactivity actually represented the parent compound. This was especially important for the very slightly soluble compounds where the level of radioactivity in the aqueous phase of either a sorption isotherm determination or an octanol-water partitioning was normally very low in comparison to the level in the soil/sediment or octanol phase. Even if only half of the radioactivity in the aqueous phase represented parent compound, the resulting Kp or Kow value would be double that obtained without correction for degradation. Thus the following procedures were developed . 5.4.1. Sorption Isotherms 5.4.1.1. Soil/Sediment Phases The soil/sediment phases were monitored for degradation products by means of TLC. A small sample (^0.5 g) of soil/sediment phase, after centri- fugation and removal of the aqueous phase, was suspended in 5 ml of acetone, 102 TABLE 5.5. WATER SOLUBILITIES OF ENERGY-RELATED ORGANIC POLLUTANTS Compound Water Solubility (yg/mi) Pyrene 7, 12-Dime thy lbenz [a] anthracene Dibenz [a 3 h] anthracene 3-Methylcholanthrene Dibenzothiophene Acridine 2, 2 '-Biquinoline 13#-Dibenzo [a, i]carbazole Acetophenone 1- Naphthol Benzidine 2- Aminoanthracene 6-Aminochrysene Anthracene-9-carboxylic acid 0.135 ± 0.013 0.0244 ± 0.0042 0.00249 ± 0.00081 0.00323 ± 0.00017 1.47 38.4 1.02 866 360 1.30 0.155 85.0 0.14 4.5 0.12 0.0104 ± 0.0041 5440 ± 71 ± 31 ± 8.0 ± 0.159 ± 0.018 ± 1.9 shaken well, allowed to stand several minutes, reshaken, and then banded ('vO.l ml) across the origin of half of a 20x20-cm 0.5-mm- thick silica gel G plate. The suspension of another soil/sediment was banded on the other half of the plate. A small amount of unlabeled standard solution was spotted at the origin between the two sample bands, and the plate was developed in the appropriate solvent system (see Table 5.1). The location of the unlabeled standard was visualized under UV light and recorded. One-cm segments from the solvent front to the bottom of the plate were scraped from each sample lane into vials of Aquasol for liquid scintillation counting. The quantity of radioactivity in the segments corresponding to the unlabeled standard was calculated as a percentage of the total radioactivity in the sample lane (half-p late-width) . For all compounds in the present study, >99% of the radioactivity present in the soil/sediment phases represented parent compound. 103 5.4.1.2. Aqueous Phases The aqueous phases that had been removed from the isotherm tubes after centrif ugation and sampling were re-centrif uged at 30,000xg for 20 minutes in preparation for high pressure liquid chromatographic (HPLC) analysis. A Waters HPLC system consisting of a Model 660 Solvent Programmer and two Model 6000A Solvent Delivery Systems was operated in conjunction with a Schoeffel Model FS 970 L.C. Fluorometer for monitoring the eluate from the column. The chromatograph was fitted with a 5-ml injection loop and a yBondapak C\q column, and operated at a flow rate of 2 ml/min. Suitable gradient and elution conditions for each compound were established using unlabeled standard solution to locate the appropriate fluorescent peak and a solution of pure radioactive compound to locate the fractions containing the parent compound. A 5-ml aqueous phase sample was injected; the pumps were programmed to deliver 20 ml of water to flush all sample components not adhering to the column from the system. This was followed by the appropriate water-to- methanol gradient to elute the parent compound and any contaminants and degradation products in a reasonable length of time. The water flush was collected as three 6.3-ml fractions and the gradient eluate as multiple 1-ml fractions in vials containing Aquasol for liquid scintillation counting. The quantity of radioactivity in the fractions corresponding to the fluorescent parent peak was calculated as a percentage of the total radioactivity in all of the fractions. The concentration of radioactive components in the isotherm aqueous phase was then corrected to reflect parent compound concentration. During a period when the HPLC system was non-functional, an alternate procedure for determining the percent degradation in the aqueous phases of acridine isotherms was developed utilizing preparative TLC. Approximately eight 0.5-ml aliquots of the aqueous phase were banded across the origin of a 20x20-cm 0.5-mm-thick silica gel G plate, with gentle but complete drying (heat gun) between and after applications. Unlabeled standard solution was spotted at a few locations across the origin. The plate was developed in benzene: methanol (95:5, v/v) to the scribed solvent front mark. Standard spots were located under UV light and their positions recorded. The plate was scored into horizontal 1-cm-wide segments from the solvent front to the bottom of the plate; each segment was scraped into a separate vial of Aquasol for liquid scintillation counting. The quantity of radioactivity in the segments corresponding to the unlabeled standard spots was calculated as a percentage of the total radioactivity on the plate. The concentration of radioactive components in the aqueous phase was then corrected to reflect parent compound concentration. The degradation study on acridine isotherm aqueous phases was repeated using the HPLC system when it was again functional. The two procedures gave comparable results, 5.4.2. Octanol-Water Partitionings 5.4.2.1. Octanol Phases The octanol phase of the final partitioning of each compound was monitored for degradation products by means of TLC. A small quantity (V> yl) of the octanol phase was spotted along with a small amount of unlabeled standard solution at the origin of a 5x20-cm 0. 25-mm- thick silica gel G plate. 104 The plate was developed in the appropriate solvent system, scraped, and analyzed as indicated above. For all compounds in the present study, >99% of the radioactivity present in the octanol phase represented parent compound. 5.4.2.2. Aqueous Phases The aqueous phase of the final partitioning of each compound was analyzed for degradation products by HPLC in a manner similar to that used for the analysis of isotherm aqueous phases described earlier in this section. The level of radioactivity in the aqueous phase was adjusted accordingly to reflect parent compound radioactivity before final calculation of the Kow. 5.4.3. Water Solubilities A IC-ml sample from a solubility flask was extracted with three 5-ml volumes of an appropriate solvent (e.g., ether for PAH, methylene chloride for N-heterocyclic compounds). The volume of the extracted aqueous phase remained at 10 ml; the volume of the extract was recorded. Both the ex- tracted aqueous phase and the extract were sampled in triplicate (1.0 ml for the former, 0.2 ml for the latter) for liquid scintillation counting. The extract was concentrated under N2 to a small volume (^1 ml) ; a small quantity of extract (^0.02 ml) was spotted along with a small amount of unlabeled standard solution at the origin of a 5x20-cm 0. 25-mm- thick silica gel G plate. The plate was developed in the appropriate solvent system, scraped, and analyzed as indicated above. For each compound in the present study, the degradation checks of all flasks yielding similar solubility levels after several filtrations revealed that >98% of the total radioactivity was solvent- extractable, and >98% of the radioactivity in the extract represented parent compound. 105 SECTION 6 SAMPLE SELECTION AND CHARACTERIZATION 6.1. CRITERIA FOR SAMPLE SELECTION Sampling sites were chosen by two major criteria: (1) to be in close proximity to and downstream from potential coal gasification sites, (2) to provide a wide range in soil and sediment properties that have been shown to affect the degree of adsorption of organic compounds. Eight sites in the United States have been identified as areas of high potential for gasification development (Map 6.1). These include Jefferson, Harrison and Belmont counties in Ohio; Washington and Greene counties in Pennsylvania; Marshall, Marion and Monongalia counties in West Virginia; Hopkins, Muhlenberg, Webster, Union and Henderson counties in Kentucky; St. Clair, Washington, Saline, Gallatin, Hamilton, Williamson, Perry, Madison, Sangamon, Christian, Macoupin, Mont- gomery, Bond, Vermilion, Edgar, Knox, Fulton and Peoria counties in Illinois; San Juan county in New Mexico; Big Horn, Rosebud, Powder River and Custer counties in Montana; Campbell and Johnson counties in Wyoming; and Dunn and Mercer counties in North Dakota. These areas with the exception of sites in New Mexico, Montana and Wyoming are in the Missouri, Mississippi, Illinois, Wabash and Ohio River watersheds. Hence, sample collection was centered on these rivers and their watersheds (Map 6.2). In addition to being in close proximity to or downstream from potential coal gasification sites, these sites also provided a wide variation in properties which have been shown to be related to the capacity of soils and sediments to adsorb various materials. At several sites an effort was made to collect sediment samples as well as soil samples represen- tative of major soils in the corresponding watershed. Locations of each sampling site and some pertinent field notes are given in Table 6.1. TABLE 6.1. FIELD NOTES a Sample Notes EPA-4 Sediment sample from Missouri River, north side of Sakakawea Park. Stanton, North Dakota. Dark-grayish silty material with some fine sand. EPA-5 Sediment sample from Beaver Creek public use area of Lake Oahe on the Missouri River eighteen miles west of Linton, North Dakota. Appears to be an inundated soil, as sample had noticeable structural development . 106 TABLE 6.1. FIELD NOTES - Continued , a Sample Notes EPA- •6 Sediment sample from Antelope Creek public use area in lake behind Big Bend Dam on Missouri River, southwest of Pierre, South Dakota. Sample is a grayish clayey material with a fair silt content. EPA- 8 Sediment sample taken from Missouri River near bridge across river, Onawa exit 1-29, Iowa side of river. Sample taken next to main channel; high velocity of flow in channel; sample extremely sandy. EPA- •9 Loess sample taken _fro_m_ bluff iust north of Turin, Iowa. EPA- •14 Soil taken from a hiehlv eroded red clav hillside southeast of mouth of Big Sandy River near Ohio River. Point Park, Ceredo, West Virginia. EPA- 15 Sediment taken from Ohio River three miles south of Leavenworth, Indiana, at base of steep bluffs in Harrison Crawford State Forest. Silty material. EPA- ■18 Sediment taken from half-submerged clay lens in the Mississippi River, next to ferry on Kentucky Route 80, near Columbus, Kentucky. EPA- ■20 Soil sample taken from an old field succession, south of scenic overlook exit, Feme Clyffe State Park, Illinois. Soil series undetermined. EPA- ■21 Sediment taken from a small river (creek) feeding Illinois River. Sample taken two miles east of Lorenzo, Illinois, where highway crosses the creek. EPA- •22 Sediment taken from large shallow bay of the Illinois River south of bridge across river at Lacon, Illinois. Silty material. EPA- 23 Sediment taken from Crane Lake north of Blind number 63. Saneanois Wildlife Refuge, confluence of Sangamon and Illinois Rivers. EPA- ■26 Sediment from wide side channel of the Mississippi River by private ferry across channel near McClure, Illinois. EPA- ■B2 Sediment sample taken from a small stream roughly one-quarter mile VipIolj fhp P — 1 tjp i~ prch prl nn 1~Vip ^nnfliPTn PipHmnnt - rnn^PTVPtl nn U C ± U w LUC J- 1 W CX LC J. oil CU U1L LUC lJUU LllCl LI i- J.CUH1UUL UUllOCl Vu L1U11 Research Center (USDA) farm in Oconee County near Watkinsville, Georgia. Stream essentially originates at the terminus of the watershed and drains the watershed. Samples EPA-4 through EPA-26 collected by W. L. Banwart and J. J. Hassett; sample EPA-B2 collected by D. S. Brown. 107 108 MAP 6.2. SAMPLING SITES ON THE MISSOURI, OHIO, WABASH, ILLINOIS AND MISSISSIPPI RIVER SYSTEMS AND THEIR WATERSHEDS 109 6.2. SAMPLE CHARACTERIZATION Table 6.2 gives the pH, cation exchange capacity, percent total nitro- gen and organic carbon as well as textural information for the samples col- lected. Soil reaction, pH, was determined on 1:1 and 1:2 soil-water mixtures by the method of Peech ( 47 ) . Cat ion exchang e__c_apaf j ty_w as determined by t he ammoni um ac eX ato mctho 4-~as_ mo dif ied by Banwart and Hassett ( 48 ) . Total ni- trogen was determined by the method given by B remner ( 49 ) with the exception that the entire digestion sample was distilled rather than an aliquot. Or- ganic carbon was determined by the Walkley-Black method as given by Allison (50). Particle size analysis was determined by the hydrometer method of Day ( 51 ) using hydrogen peroxide to destroy the organic matter ( 52 ) . TABLE 6.2. CHARACTERISTICS OF SOILS AND SEDIMENTS pH (1: 1) pH (1: 2) CEC (me/lOOe ) Total N (%) Organic carbon (%) Sand (%) Clay (%) Silt (%) EPA- 4 7. 79 8. 22 23.72 0.190 2.07 3.0 55.2 41.8 EPA- 5 7.44 7.20 19.00 0.192 2.28 33.6 31.0 35.4 EPA- 6 7. 83 8.23 33.01 0.097 0.72 0.2 68.6 31. 2 EPA- 8 8.32 8.56 3.72 0.010 0.15 82.4 6.8 10.7 EPA-9 8. 34 8.55 12.40 0.015 0.11 7.1 17.4 75.6 EPA- 14 4.54 4.30 18.86 0.064 0.48 2.1 63.6 34.4 EPA- 15 7. 79 7.80 11. 30 0.092 0.95 15.6 35.7 48.7 EPA- 18 7.76 7. 79 15.43 0.062 0.66 34.6 39.5 25.8 EPA-20 5.50 5.16 8.50 0.126 1.30 28.6 71.4 EPA- 21 7.60 7.95 8. 33 0.157 1.88 50.2 7.1 42.7 EPA- 2 2 7.55 7.90 8.53 0.128 1.67 26.1 21.2 52.7 EPA- 2 3 6. 70 7. 10 31. 15 0.195 2.38 17. 3 69.1 13.6 EPA- 2 6 7. 75 8. 10 20.86 0.152 1.48 1.6 42.9 55.4 EPA-B2 6.35 6.50 3. 72 0.073 1.21 67.5 18.6 13.9 m 110 6.2.1. Instrumental Neutron Activation Analysis (INAA) The basic concept of INAA is that a small fraction of the stable nuclei present in a sample becomes radioactive when bombarded with neutrons in the 1.5 MW TRIGA reactor. In many of the subsequent radioactive decay processes, one or more high energy photons or gamma rays are emitted. A high resolution semiconductor detector interacts with a gamma ray and yields a pulse with the maximum voltage of the pulse being proportional to the gamma ray energy. This pulse is amplified and shaped and then sorted by a pulse height analyzer so that various energy gamma rays result in counts in dif- ferent locations in a computer-like memory. The energy of a gamma ray is unique to a particular isotope of a specific element so that a qualitative analysis can be made by observing the spectrum of gamma ray energies emitted by the activated samples. A quantitative analysis can be made by relating the number of gamma rays emitted by the sample relative to a standard con- taining a known amount of that element. Approximately 100 mg of the solid sample are weighed out into a pre- cleaned polyethylene container. A standard is prepared by pipetting a small known quantity of an aqueous solution containing a known concentration of the element of interest onto a Whatman No. 41 filter in a clean polyethylene con- tainer. The standard is irradiated simultaneously with the sample. The gamma ray spectra from a series of samples and standards are recorded on mag- netic tape. A fully automatic computer code then reduces the counting data to elemental concentrations. A list of elements and their interference-free limits of detection is given in Table 6.3. Table 6.4 gives the results of INAA analysis of the soil and sediment samples. 6.2.2. Clay Mineral Analysis Twenty-gram samples of the soils and sediments were pretreated with 200 ml of NaOCl (pH 9.5) to remove organic matter according to the method of Anderson (53) . Samples were placed in a nearly boiling water bath and con- stantly stirred until the reaction was complete. The suspension was centri- fuged at 5,000 rpm for 5 minutes and the supernatant decanted. The solid residue was then treated with 200 ml of IN NaOAc (pH 5) and placed in a nearly boiling water bath for one hour to facilitate removal of the carbonate cementing agents. The solid residue after centrif ugation and decantation of the supernatant was treated by the method of Mehra and Jackson (54) using 160 ml of 0.3M sodium citrate, 20 ml of 1M NaHC03 and 2 grams of solid sodium dithionate to remove the amorphous cementing agents. A portion of the residue remaining after removal of the organic matter, carbonates and amorphous materials was saturated with Mg 2+ ions to facilitate uniform inter layer water adsorption of the expandable layer-silicates. A second portion of the residue was saturated with K ions, thus restricting water adsorption by the normally expandable layer-silicate vermiculite. These K + and Mg 2+ -treated samples were then washed twice with methanol and twice with acetone to remove excess salts. Ill TABLE 6.3. ELEMENTAL LIMITS OF DETECTION BY INSTRUMENTAL NEUTRON ACTIVATION ELEMENT INAA (yg) Aluminum 0.004 Antimony 0.007 Arsenic 0.005 Barium 0.02 Bismuth Bromine 0.003 Cadmium 0.005 Calcium 4. Cerium 0.2 Cesium 0.001 Chlorine 0.05 Cobalt 0.01 Copper — _ Dysprosium 0.00003 Europium 0.0001 Gallium 0.002 Gold 0.0005 Hafnium 0.0006 Indium 0.00006 Iodine 0.002 Iridium 0.0003 Iron 2. Lanthanum 0.005 Lead Magnesium 0.5 Manganese 0.0001 Mercury 0.003 Molybdenum 0.1 Nickel 0.7 Niob ium 3. Potassium 0.2 Rubidium 0.02 Samarium 0.001 Scandium 0.001 Selenium 0.01 Sodium 0.004 Strontium 0.005 Tantalum 0.1 Terbium 0.03 Thallium Thorium 0.2 Titanium 0.1 Tungsten 0.004 Uranium 0.003 Vanadium 0.002 Zinc 0.1 Zirconium 0.8 112 TABLE 6.4. INSTRUMENTAL NEUTRON ACTIVATION ANALYSIS OF SOILS AND SEDIMENTS EPA-B2 EPA-4 EPA-5 EPA-6 EPA-8 EPA-9 EPA- 14 (yg/g) (yg/g) (yg/g) (yg/g) (yg/g) (yg/g) (yg/g) As <17 5.2 11.8 12.9 9.8 4.6 9.9 Ba 383 671 788 1183 830 872 450 Br <2.2 <2.6 <3.4 <3.0 <1.9 <2.3 - Ce 84 59.8 61.3 55.1 41.1 65.9 86.9 Co 8.2 12.45 8.89 19. 11 5.96 10.88 11.00 Cs 1.5 4.55 3.24 4.62 1.27 2.98 7.97 Dy 4.1 3.45 3.83 <0.8 2.10 3.17 4.49 Eu 1.22 0.948 1.026 1.020 0.304 1.296 0.943 Ga 12 6.7 10.8 5.7 7.2 8.7 22.9 Hf 10.4 5.22 7.49 6.13 7.81 12.25 8.03 Hg — — — — — — — La 40.4 31.2 25.4 33.8 21.6 30.6 46.1 Lu 0.37 0.62 0.38 Mn 1140 818. 8 392. 1 4300 347.8 764.8 216 . 1 Nd 27 24 36 19 .0 31 53 43 Ni Rb 72 80.0 66.7 105 62.4 97 200 Sb 0. 18 1.31 6.78 2.14 0.51 1.25 6.05 Sc 7.58 11.52 7.662 9.349 3.739 6.895 16.15 Se <0.4 0.44 1.53 1.19 <0.19 0.89 1.88 Sr <78 43 171 338 260 299 <80 Ta 0.66 0.83 0.540 0.48 0.314 0.54 1.08 Tb 1.0 0.68 0.64 0.70 0.460 0.368 0.959 Th 10.9 9.03 8.65 8.25 5.74 9.47 14.61 U 2.6 Yb 2.54 1.841 2.110 1.850 1.424 2.64 2.93 Zr 271 40 102 174 410 37 249 %%%%%%% Ca 0.84 2.02 1.41 1.44 1.39 3.08 0.51 Fe 2.18 3.100 2.508 3.039 1.257 2.300 4.888 K 1.00 1.778 1.657 1.331 1.324 1.455 2.453 Na 0.771 0.7556 1.081 0.6590 1.051 0.922 0.152S 113 TABLE 6.4. Continued T7P A i i; trA- 1 j (yg/g) A— lo (yg/g) vd a on trA- zU (yg/g) Lr A— z 1 (yg/g) hF A— z z (yg/g) r>T) A O O EPA- 2 3 (yg/g) EPA- 2 6 (yg/g) A ^ AS /. 4 y . 5 4. / < 16 <16 8 . 1 <26 ba 451 / za 4oo Ilk 34y 415 544 595 d r <9 A ^»Z . h <"9 ft **Z . o J.U J . H 90 1 7 S f\ 1 J . u 76 4 / U • H 7 Q AD 9 4u . Z a ^ n / J . J Cn 70 9 1 1 A A ^ 14 • O J 14 8"} -LH • OJ in 7 8 9 o • z in r 11 1 11 .1 3 QS J . -7 J 1 8 1 9 9 7 z • z / ^ 7 9 A Z ■ 4 J . u uy S 81 J • O 1 A Q9 H • 7t t.OJ o . z 7 J . y J . H IlU 1 Q 1 1 A 1 1 . ID J 1 1:7c: 1 • j / J U . / / n qa u . yo 1 1 /. 1 .ZD "7 1 / . 1 13.0 <3 . 1 9 . 4 <4. 2 16 <4 .8 "I 15 . o j 9 . 50 1 / 1 o 14 . 15 / "7 4. 7 6 . 9 2.0 7.3 Hg Sorption and Transport Processes in Soils, pp. 63-78. Monograph 37. ^ London: Society of Chemical Industry. f ^/ 20. Gillingham, D. , and D. Heien. 1971. Regression thru the origin. Am. Stat. 25:54-55. 21. Wahid, P. A., and N. Sethunathan. 1978. Sorption-desorption of parathion in soils. J. Agric. Food Chem. 26:101-105. 128 22. Chiou, C. T. , L. J. Peters and V. H. Freed. 1979. A physical concept of soil-water equilibria for nonionic organic compounds. Science 206: 831-832. 23. Stevenson, F. J. 1972. Organic matter reactions involving herbicides in soil. J. Environ. Qual. 1:333-343. 24. Helling, C. S., and B. C. Turner. 1968. Pesticide mobility: Deter- mination by soil thin- layer chromatography. Science 162:562-563. 25. Shin, Y.-O. , J. J. Chodan and A. R. Wolcott. 1970. Adsorption of DDT by soils, soil fractions, and biological methods. J. Agric. Food Chem. 13:1129-1133. 26. Helling, C. S. 1971. Pesticide mobility in soils. I. Parameters of soil thin-layer chromatography. Soil Sci. Soo. Am. Proa. 35:732-737. 27. U.S. Environmental Protection Agency (Washington, DC). 1973. List of toxic [water] pollutants. Fed. Register 38:24342-24344. 28. U.S. Department of Labor (Washington, DC). 1974. Carcinogen. Fed. Register 39:3756-3797. 29. Clayson, D. B. 1976. Occupational bladder cancer. Frev. Med. 5:228- 244. 30. Haley, T. J. 1975. Benzidine revisited. Review of the literature and problems with the use of benzidine and its congeners. Clin. Toxicol. 8:13-42. 31. Korenman, I. M. , and B. A. Nikolaev. 1974. Determination of the protonation constants of weak diacidic bases by an extraction method. Zh. Fiz. Khim. 48:2545-2549 (Russ.). 32. Solomon, D. H. , B. C. Loft and J. D. Swift. 1968. Reactions catalyzed by minerals. IV. The mechanism of the benzidine blue reaction on silicate minerals. Clay Miner. 7:389-397. 33. Theng, B. K. G. 1971. Mechanisms of formation of colored clay-organic complexes. A review. Clays Clay Miner. 19:383-390. 34. Tennakoon, D. T. B., J. M. Thomas, M. J. Tricker and J. 0. Williams. 1974. Surface and inter elate chemistry of layered silicates. Part I. General introduction and the uptake of benzidine and related organic molecules by montmorillonite. J. Chem. Soc. 3 Dalton Trans. (20) : 2207-2211. 35. Tennakoon, D. T. B. , J. M. Thomas and M. J. Tricker. 1974. Surface and intercalate chemistry of layered silicates. Part II. An iron-57 MBssbauer study of the role of lattice-substituted iron in the benzidine blue reaction of montmorillonite. J. Chem. Soc. s Dalton Trans. (20): 2211- 2215. 129 36. Tennakoon, D. T. B. , and M. J. Tricker. 1975. Surface and intercalate chemistry of layered silicates. Part V. Infrared, ultraviolet, and visible spectroscopic studies of benzidine-montmorillonite and related systems. J. Chem. Soo. 3 Valton Trans. (18) : 1802-1806 . 37. Blumer, M. 1976. Polycyclic aromatic compounds in nature. Sci. Am. 234:35-45. 38. MacKenzie, M. J., and J. V. Hunter. 1979. Sources and fates of aro- matic compounds in urban stormwater runoff. Env. Sci. Teohnol. 13: 179-183. 39. Briggs, G. G. 1973. A simple relationship between soil adsorption of organic chemicals and their octanol/water partition coefficients. Proa. 7th British Insecticide Fimgicide Conf. } pp. 475-478. 40. Hilton, B. D. , and R. D. O'Brien. 1964. A simple technique for triti- ation of aromatic insecticides. J. Agric. Food Chem. 12:236-239. 41. Lu, P.-Y., R. L. Metcalf amd E. M. Carlson. 1978. Environmental fate of five radiolabeled coal conversion by-products evaluated in a labora- tory model ecosystem. Environ. Health Perspect. 24:201-208. 42. Leo, A., C. Hansch and D. Elkins. 1971. Partition coefficients and their uses. Chem. Rev. 71:525-616. 43. Chiou, C. T., V. H. Freed, D. W. Schmedding and R. L. Kohnert. 1977. Partition coefficient and bioaccumulation of selected organic chemicals. Environ. Soi. Teohnol. 11:475-478. 44. Karickhoff, S. W., and D. S. Brown. 1979. Determination of octanol/ water distribution coefficients, water solubilities, and sediment /water partition coefficients for hydrophobic organic pollutants. Report No. EPA 600/4-79-032. Athens, GA: Office of Research and Development, U.S. Environmental Protection Agency. 45. Haque, R. , and D. Schmedding. 1975. A method of measuring the water solubility of hydrophobic chemicals: Solubility of five polychlorinated biphenyls. Bull. Environ. Contam. Toxicol. 14:13-18. 46. Mackay, D. , and W. Y. Shiu. 1977. Aqueous solubility of polynuclear aromatic hydrocarbons. J. Chem. Engin. Data 22:399-402. 47. Peech, M. 1965. Hydrogen-ion activity. In Methods of Soil Analysis^ Agronomy 9(2), ed. C. A. Black, pp. 914-926. Madison, WI : American Society of Agronomy. 48. Banwart, W. L. , and J. J. Hassett. 1976. Laboratory Introduction to Soil Science. Champaign, IL: Stipes Publishing Co. 118 p. 130 49. Bremner, J. M. 1965. Total nitrogen. In Methods of Soil Analysis, Agronomy 9(2), ed. C. A. Black, pp. 1149-1178. Madison, WI: American Society of Agronomy. 50. Allison, L. E. 1965. Organic carbon. In Methods of Soil Analysis, Agronomy 9(2), ed. C. A. Black, pp. 1367-1378. Madison, WI: American Society of Agronomy. 51. Day, P. R. 1965. Particle fractionation and particle size analysis. In Methods of Soil Analysis, Agronomy 9(1), ed. C. A. Black, pp. 545-567. Madison, WI: American Society of Agronomy. 52. Kilmer, V. J., and L. T. Alexander. 1949. Methods of making mechanical analyses of soils. Soil Sci. 68:15-24, 53. Anderson, J. U. 1961. An improved pretreatment for mineralogical anal- ysis of samples containing organic matter. Clays Clay Miner. 10:380-388. 54. Mehra, 0. P. , and M. L. Jackson. 1958. Iron oxide removal from soils and clays by a dithionate-citrate system buffered with sodium bicar- bonate. Clays Clay Miner. 5:317-327. 55. Theisen, A. A., and M. E. Harward. 1962. A paste method for preparation of slides for clay mineral identification by x-ray diffraction. Soil Soi. Soa. Am. Proa. 26:90-91. 56. Brown, C. Q. 1959. Clay mineral relations in two tributary basins within the York River tributary basin. Southeastern Geol. 1:95-104. 57. Lucas, J. 1962. The transformation of clay minerals during sedimenta- tion: Studies of Triassic clays. Mem. Serv. Carte Geol. Alsace Lorraine 23:1-202 (Fr.). Washington, DC: (Transl. and publ. for) U.S. Department of Agriculture and National Science Foundation. 21 p. 58. Jackson, M. L. 1965. Free oxides, hydroxides, and amorphous alumino- silicates. In Methods of Soil Analysis, Agronomy 9(1), ed. C. A. Black, pp. 578-603. Madison, WI: American Society of Agronomy. 59. Khan, A., and W. L. Banwart. 1979. Effect of incubation and microbial inhibition at field moisture capacity on changes in DTPA-extractable Fe, Zn, and Cu in soils of varying pH. Comm. Soil Sci. Plant Anal. 10: 613-622. 60. Khan, A., and P. N. Soltanpour. 1978. Effect of wetting and drying on DTPA-extractable Fe, Zn, Mn, and Cu in soils. Comm. Soil Sci. Plant Anal. 9:193-202. 61. Molloy, M. G., and R. B. Lockman. 1979. Soil analysis as affected by drying temperatures. Comm. Soil Sci. Plant Anal. 10:545-550. 131 62. Sertsu, S. M. , and P. A. Sanchez. 1978. Effects of heating on some changes in soil properties in relation to an Ethiopian land management practice. Soil Sci. Soc. Am. J. 42:940-944. 63. Thien, S. J., D. A. Whitney and D. A. Karlen. 1978. Effect of micro- wave radiation drying on soil chemical and mineralogical analysis. Comm. Soil Sci. Plant Anal. 9:231-241. 64. Soltanpour, P. N. , A. Khan and A. P. Schwab. 1979. Effect of grinding variables on the NH^HCC^-DTPA soil test values for Fe, Zn, Mn, Cu, P, and K. Comm. Soil Sci. plant Anal. 10:903-909. 65. Soltanpour, P. N. , A. Khan and W. L. Lindsay. 1976. Factors affecting DTPA-extr actable Zn, Fe, Mn, and Cu from soils. Comm. Soil Sci. Plant Anal. 7:797-821. 66. Deli, J., and G. F. Warren. 1971. Adsorption, desorption, and leaching of diphenamid [N,N-dimethyl-2 ,2-diphenylacetamide] in soils. Weed Sci. 19:67-69. 67. Doherty, P. J., and G. F. Warren. 1969. Adsorption of four herbicides by different types of organic matter and a bentonite clay. Weed Res. 9: 20-26. 68. Grover, R. 1971. Adsorption of picloram by soil colloids and various other adsorbents. Weed Sci. 19:417-418. 69. Hance, R. J. 1965. The adsorption of urea and some of its derivatives by a variety of soils. Weed Res. 5:98-107. 70. Hance, R. J. 1965. Relation between the adsorption of diuron and the nature of the adsorbent. Weed Res. 5:108-114. 71. Upchurch, R. P., and W. C. Pierce. 1958. The leaching of monuron from Lakeland sand soil. II. The effect of soil temperature, organic matter, soil moisture, and amount of herbicide. Weeds 6:24-33. 72. Coats, G. E. , H. H. Funderburk, Jr., J. M. Lawrence and D. E. Davis. 1966. Factors affecting persistence and inactivation of diquat and paraquat. Weed Res. 6:58-66. 73. Weber, J. B. , and D. C. Scott. 1966. Availability of a cationic herbi- cide adsorbed on clay minerals to cucumber seedlings. Science 152: 1400-1402. 74. Weed, S. B. , and J. B. Weber. 1968. The effect of adsorbent charge on the competitive adsorption of divalent organic cations by layer-silicate minerals. Am. Mineral. 53:478-489. 132 TECHNICAL REPORT DATA (Please read Instructions on the reverse before completing) 1. REPORT NO. 2. EPA-600/3-80-041 3. RECIPIENT'S ACCESSI ON> NO. 4. TITLE AND SUBTITLE Sorption Properties of Sediments and Energy-Related Pollutants 5. REPORT DATE April 1980 issuing date 6. PERFORMING ORGANIZATION CODE 7. AUTHOR(S) John J. Hassett, Jay C. Means, Wayne L. Banwart, and Susanne G. Wood 8. PERFORMING ORGANIZATION REPORT NO. 9. PERFORMING ORGANIZATION NAME AND ADDRESS Department of Agronomy University of Illinois at Urbana-Champaign Urbana, Illinois 61801 10. PROGRAM ELEMENT NO. A1MH1E 11. CONTRACT/GRANT NO. 68-03-2555 12. SPONSORING AGENCY NAME AND ADDRESS Environmental Research Laboratory — Athens GA uiiicc or tiesearcn ana ueveiopmenc U.S. Environmental Protection Agency Athens, Georgia 30605 13. TYPE OF REPORT AND PERIOD COVERED Final, 7/77-12/79 IH, orUIMOUnl IMU HOCIMV/ I LUUl EPA/600/01 15. SUPPLEMENTARY NOTES 16. ABSTRACT This report describes the factors that determine the extent of sorption of organ- ic compounds that are representative of coal conversion waste streams. The compounds, all radiolabeled, were acetophenone; 1-naphthol; pyrene; 7, 12-dimethylbenz (a) anthra- cene; 3-methylcholanthrene; dibenz(a,h) anthracene; acridine; 2 ,2 '-biquinoline; 13H- dibenzo (a, i) carbazole; dibenzothiophene; benzidine; 2-aminoanthracene ; 6-aminochrysene ; and anthracene-9-carboxylic acid. Batch equilibrium isotherms were determined for each compound on 14 sediments and soils that had been collected from the Missouri, Illinois, Mississippi, and Ohio rivers and their watersheds. Laboratory procedures for determining octanol-water partition coefficients and water solubilities were developed and then performed on the compounds. The sorption constants were correlated with soil and sediment properties and witl the water solubilities and octanol-water partition coefficients of the compounds. Re- gression equations were developed that allow prediction of a hydrophobic compound's linear partition coefficient from knowledge of the compound's octanol-water partition coefficient or its water solubility and the organic carbon content of the sediment or soil. 17. KEY WORDS AND DOCUMENT ANALYSIS a. DESCRIPTORS b. IDENTIFIERS/OPEN ENDED TERMS c. COSATI Field/Group Adsorption Chemical analysis Coal Energy Organic compounds Sediments 68C 68D 99A 99D 13. DISTRIBUTION STATEMENT RELEASE TO PUBLIC 19. SECURITY CLASSJThis Report) UNCLASSIFIED 21. NO. OF PAGES 147 20. SECURITY CLASS (This page) UNCLASSIFIED 22. PRICE EPA Form 2220-1 (9-73) 133 OUS GOVERNMENT PRINTING OFFICE 1980 - 6 5 7-146 / 56 34 3 0112 112943300 > i ON o o UJ I 00 o I o -O- c a ID a ■o X) 9 •< s re TJ It u o £ 01 n c 3~ o VI a 0J 3 It it o 3 a O 3 3 ai ha o O o g o o ?r a. r> ■o < ar S VI 3" 3 a ■a ■< HI all n o 3 It o 3 vi O a n It (D 00 o o o CD C o 3 O IT *- • o 01 «i (/) 0) > m c (O 3 3 !t> < - o a 3 £ 2 • Q) (/> "0 o 55" o 5 3 o 5' n 3' 3 ai o 0) 3 X = o O I cn O) 00 m 3 < o 3 3 3 3 V) a> 0) — 1 o 3" "0 (O > ID CO O 00 < tn > -0 m Ti "D (0 o n a. qi 3 a