BARR for Decontamination of Soil and Groundwater Volume 2 This is the TECHNICAL APPENDIX Copyright 1993 by Larry Dieterich All Rights Reserved Duplication and distribution is permitted, but credit must be given to the author and the document must be distributed in its entire form. No fragmentation, editing or deletions are permitted on copies for distribution. This document is part of a 3 volume set. Volume 1 BARR report Volume 2 (this document) BARR Technical Appendix Volume 3 BARR Bibliography The BARR process was developed by Larry Dieterich Larry Dieterich 405 E 7th Street Davis, California 95616 USA voice/fax (916) 758-9260 Internet email- water@well.sf.ca.us This edition of the BARR reports is made available in ASCII format for distribution over the Internet. Paper, disk, or formatted Macintosh files with graphics are available upon request. Contact the author at the above address. BARR: For Decontamination of Soil and Groundwater B Bio- A Anaerobic R Reduction & R Re-oxidation Bio-Anaerobic Reduction & Re-oxidation BARR: For Decontamination of Soil and Groundwater TECHNICAL APPENDIX BARR TECHNICAL APPENDIX CONTENTS TRANSFORMATION PROCESSES / CHEMODYNAMICS BIOTIC DEGRADATION ABIOTIC DEGRADATION SORPTION BONDING FORCES FREE ENERGY INTERFACIAL TENSION SALT EFFECTS COSOLVENCY pH REDOX MICELLES PRECIPITATION COLLOIDS CLAY FRACTION ORGANIC MATTER PROPERTIES ORGANIC MATTER AND ORGANIC POLLUTANT INTERACTIONS HUMIC/MINERAL ASSOCIATIONS COLLOID STABILITY BIOCOLLOIDS COLLOIDS IN GROUNDWATER COLLOIDAL METAL BEHAVIOR COLLOIDAL BEHAVIOR OF RADIONUCLIDES COLLOID SUMMARY TRANSFORMATION PROCESSES / CHEMODYNAMICS Transformation occurs when a thermodynamically favorable reaction occurs. Transformation denotes a change in the target chemical. Whether by adding or removing a substituent group, re-arranging, breaking or forming bonds. Transformation is not necessarily the same as degradation (although it may be), but transformation is often a step toward degradation as it represents a modification of the target chemical as a step toward its ultimate mineralization. BIOTIC DEGRADATION Biological Transformations Basically 5 processes are involved in microbial transformation: 1. Biodegradation- The contaminant serves as a substrate for growth. 2. Cometabolism- Material is transformed by metabolic reactions but does not serve as an energy source. 3. Conjugation- The contaminant molecule is joined to another molecule in the system. 4. Accumulation- In which the material partitions (is incorporated into) into the organism 5. Secondary effects of microbial activities-In which the pollutant is transformed because of changes in pH, redox conditions, reactive products, etc., brought about by microorganisms. Enzymes in soil can be separated into a number of categories according to their location within the soil microenvironment. Indeed, the measured activity of a particular enzyme is usually a composite of activities belonging to two or more categories: 1. Enzymes associated with living, metabolically active cells. 2. Enzymes associated with viable but nonproliferating cells such as resting vegetative cells, bacterial endospores, fungal spores, and protozoan cysts. 3. Enzymes which are associated with, at least briefly, with their substrates in enzyme-substrate complexes. 4. Enzymes attached to entire dead cells, cell debris or having diffused away from dead and lysed cells. Many enzymes in this category may have had an original functional location on or within a cell yet may survive for a short period when released into the soil aqueous phase. 5. Enzymes which are more or less permanently immobilized on the soil clay and humic colloids. Expandable clays have a high affinity for enzymes although this is not always synonymous with the retention of catalytic ability. Enzymes associated with humates retain their activity for long periods. The ubiquitous presence of biological processes and products acts to influence the overall system. Enzymes, both intact and extracellular, will catalyze many reactions. Biological lipid membranes will provide a nonaqueous, nonpolar interface in the aqueous system. This has special significance for nonpolar pollutants, such as common petrochemicals. The ability of microorganisms to catalyze transformation of organic chemicals is considered prerequisite information. There are many types of microorganisms: fungi, bacteria, actinomycetes, viruses, viroids, mycoplasmsa-like organisms (MLO's), slime molds protozoans, rotifers, algae, ... the list is actually endless, both because of the widespread disagreement between workers in the field as to the definitions and distinctions between organisms and the fact that the organisms themselves are continuously c hanging and evolving and mutating. Many microbes have not been discovered and described and the organisms that have been described are continuously subject to reclassification as natural processes of growth and evolution occur. Regarding the BARR process, the most important microbes appear to be the bacteria. Bacteria are small, (0.5 - 3 micrometer), unicellular organisms. Bacteria are the smallest free-living microbes. They are small compared to the pore sizes in most nonindura ted geological materials and they are large in relation to the size of hydrated inorganic ions and molecules. Bacteria have been shown to effect any number of chemical transformations in both mineral and organic materials. They multiply by binary fission. (1,2,4,8,16,32,64,128... 2n, n being the number of divisions) and are capable of extremely large populations in very short time periods under favorable conditions. Bacteria catalyze nearly all the important redox reactions that occur in groundwater. This means that although the reactions are spontaneous thermodynamically, they require the catalyzing effect of microorganisms such as bacteria in order to proceed at a significant rate. The catalytic capability of bacteria is produced by the activity of enzymes that normally occur within the bacteria. Enzymes are protein substances formed by living organisms that have the power to increase the rate of redox reactions by decreasing the activation energies of the reactions. They accomplish this by interacting strongly with complex molecules representing molecular structures halfway between the reactant and the product. The local molecular environment of many enzyme reactions is very different from the bulk environment of the aqueous system. Bacteria and their enzymes are involved in redox processes in order to acquire energy for synthesis of new cells and maintenance o f old cells. Some of the energy obtained from the redox reactions is maintenance energy required by bacterial cells for such things as mobility, to prevent an undesirable flow of solutes either into or out of the cell, or for resysnthesis of proteins that are constantly degrading. For bacteria to be able to make use of an energy yield from a redox reaction, a minimum free energy change of approximately 60 kJ/mol between the reactants and the products is required. The main source of energy for bacteria in the groundwater zone is the oxidation of organic matter. Bacteria that can thrive only in the presence of oxygen are called aerobic bacteria. Anaerobic bacteria require an absence of dissolved oxygen. Facultative bacteria can thrive with or without oxygen. The lower limit of dissolved oxygen for the existence of most aerobic bacteria is considered to be about 0.05 mg/l, but some aerobic species can persist at lower levels. Since most methods commonly used for measuring dissolved O2 have a lower detection limit of about 0.1 mg/l, it is possible that aerobic bac teria can mediate redox reactions in situations that might appear to be anaerobic based on the lack of detectable oxygen. Bacteria of different varieties can withstand fluid pressures of many hundreds of bars, pH conditions from 1 to 10, temperatures from near 0 C to greater than 75 C, and salinities much higher than that of seawater. They can migrate through porous geological materials and in unfavorable environments can evolve into resistant bodies that may be activated at a later time. In spite of these apparent characteristics of hardiness, there are many subsurface environments in which organic matter is not being oxidized at an appreciable rate. As a result, the redox conditions have not declined to low levels even though hundreds or thousands of years or more have been available for the reactions to proceed. Acclimation Period Biodegradation of organic compounds in the environment is often preceded by an extended acclimation period. The acclimation period is taken to mean the time interval during which biodegradation is not detected, but the pollutant has been introduced. This period may be significant because a compound may pass through treatment systems and deeper into the environment during this time. There are several reasons offered for this lag time: (1) The time is needed for small populations to become sufficiently lar ge to give detectable loss of chemical. (2) Adverse environmental conditions at the site where the chemical is discharged. (3) Rarity of microbes able to degrade specific chemicals. (4) Presence of compounds inhibiting microbes. The inhibitor need not be another compound, but the substrate itself, because the concentrations of many chemicals introduced into the environment are probably high enough to suppress the microbes having the ability to degrade those compounds. Other evidence exists that one compou nd may shorten the acclimation period needed before another is degraded. (5) An acclimation period may result from the time required for the appearance of a new genotype after a mutation or genetic exchange, which presumably occurs during the time microor ganisms are exposed to the compound. The new organism then grows and mineralizes the compound. Mutations and gene transfers are rare events, however and appear in only a small and random percentage of samples. The acclimation periods prior to detectable dehalogenation of halogenated benzoates in anaerobic lake sediments ranged from 3 weeks to 6 months. These acclimation periods are reproducible over time and among sampling sites and were characteristic of the chemical tested. (LINKFIELD et al. 1989). The lengthy acclimation period appears to represent an induction phase in which little or no aryl dehalogenation is observed, followed by an exponential increase in activity typical of an enrichment response. Con tinuous growth from the time of the first exposure to the chemical is inconsistent with the extremely low per-cell activities estimated for the early days of the acclimation period and the fact that the dehalogenation yields no carbon to support microbial growth. The finding of a characteristic acclimation time for each chemical argues against nutritional deficiency, inhibition or predation as an explanation for this phase of metabolism, while the reproducibility of the findings with time and space and among replicates argues against genetic changes as the explanation. The acclimation times correlate with the eventual dehalogenation rates. This may reflect the general energy limitations in the anaerobic communities and suggests that those chemicals with faster deh alogenation rates provide more energy for the induction and growth phases of the active population. In aerobic microbial communities the acclimation periods for xenobiotic compounds typically range from several hours to several days, but for anaerobic communities these periods are much longer, often from 2 weeks to 6 months or longer. The concentration and structure of the xenobiotic compound itself probably influences the acclimation period. In particular, low concentrations and structure-biodegradability relationships are known to affect aerobic biodegradation rates, and in some cases they have been shown to affect the acclimation period as well. The hypotheses posed for aerobic acclimation periods should be equally valid for the anaerobic ones Acclimation periods are a function of chemical structure; halogen type and other ring substituents alter the acclimation period. In the case of the bromobenzoates, the effect of ring position is clearly demonstrated: the para isomer exhibits greater acclimation periods than either the ortho or the meta form. Increased acclimation periods as a function of high xenobiotic concentration have been reported for the anaerobic microbial metabolism of one and two carbon halogenated solvents. The acclimation period observed in literature reports may be due to either a "true" acclimation i.e., a period of no biodegradation followed by initiation and acceleration of degradation, or an "apparent" acclimation, in which biodegradation would proceed from the time of addition but at rates so slow as to be nondetectable. A true acclimation could be caused by the time required for genetic changes, induction of new protein synthesis, or exhaustion of preferential substrates. An apparent acclimation would probably be the case if the initial population was very small but grew continuously, until, at some point, it was large enough to result in detectable biodegradation. Lag periods are reportedly longer for samples collected from field sites that were low in dissolved inorganic nutrients (N,P). Lag periods decreased with fertilization with N & P. (LEWIS et al., 1986) If the redox reactions that require bacterial catalysis are not occurring at significant rates, a lack of one or more essential nutrients for bacterial growth is the likely cause. There are various types of nutrients. Some are required for the incorporation into the cellular mass of the bacteria. Carbon, nitrogen, sulfur, and phosphorous compounds and many metals are in this category. Other nutrients are substances that function as electron donors and energy sources, such as water, ammonia, glucose, and H2S, and substances that function as electron acceptors, such as oxygen, nitrate, and sulfate. Aerobic Metabolism Aerobic transformation is generally considered to occur in the presence of oxygen. Oxidizing agents are the chemicals that provide the terminal electron acceptors in the reaction. These can be any material capable of donating electrons. Oxygen and hydrogen peroxide are commonly used. The tendency for reduced hydrocarbons to become oxidized provides the basis for most of the treatment measures applied to cleanup of deep ground contamination. There are a number of limitations and difficulties in the creation and maintenance of oxidizing conditions for the degradation of chemicals in the subsurface. Lowering the water table or other measures to maintain oxygen supply to a contaminated formation is commonly reported as a remedial measure. There are reports of more or less success with the oxidative biodegradation of subsurface contamination. It is frequently found that the limiting factor in such treatments is the availability of a reduced carbon source to sustain the levels of microbes av ailable to degrade the pollutant of concern. The degradation rate declines markedly at about 60% recovery (BOETHING and ALEXANDER, 1979) Concentration of the compound may be a significant factor affecting its susceptibility to microbial attack, and organic chemicals may persist in some environments as a result of low prevailing concentration or low solubility in water. Both of these limitations are removed by the BARR process. Anaerobic Transformations Under certain hydrologic conditions, such as infiltration of precipitation, anaerobic zones in shallow aquifers may become aerobic as a result of the rise or decline of the water table, as influenced by the capillary fringe. Because of possible alternating oxidizing and reducing conditions, it is reasonable to assume that both aerobic and anaerobic microbes in these zones of contaminated aquifers have adapted enzyme systems capable of metabolizing organic compounds, thus influencing their fate and transport in the subsurface. Under aerobic conditions, molecular oxygen is required as a terminal electron acceptor during respiration and also for hydroxylation of aromatic compounds before ring cleavage. These reactions are mediated by oxygenases. Under anaero bic conditions, microbial activities tend to attack complex aromatic structures at substituent groups and convert the parent substrates to hydroxylated, carboxylated, or amino derivatives. (KUHN et al. 1989). However, in the absence of molecular oxygen, aromatic compounds may be degraded by methanogenic consortia in the contaminated aquifer or in the presence of other electron acceptors such as nitrate of sulfate. It is reasonable that water can be the source of oxygen for the hydroxylation of petroleum-derived groundwater contaminants under sufficiently reducing conditions. In the absence of molecular oxygen, the catabolism of organic matter by micro-organisms can be achieved either by fermentation or by anaerobic respiration. The formal definition of fermentation is that organic substrates act both as electron donors and acceptors and may concern large polymers such as starch and cellulose. Fermentative processes do not require exogenous electron acceptors. The organic substrate plays the role of an internal electron donor (reducing agent) and one of the metabolites produced acts as an electron acceptor (oxidizing agent). The organic substrate is not fully oxidized to CO2 in fermentative processes which accounts for the build-up of partially reduced compounds (alcohols, organic acids, etc.) The metabolism of organic substrates by anaerobic respiration involves inorganic electron acceptors, such as nitrate, sulfate and CO2 . Such metabolism generally uses small molecules produced by fermentative metabolism or by the aerobic biodegradation which takes place in the nearby oxic environment. Bacteria which use nitrate as an electron acceptor are called nitrate-reducing bacteria and the process is known as nitrate respiration. When nitrate is completely reduced to gaseous products, the process is called denitrification. Many bacteria can grow aerobically with oxygen and anaerobically with nitrate. These organisms are commonly called "facultative" organisms. The strictly anaerobic bacteria which use sulfate as an electron acceptor form the ecophysiological group of sulfate reducing bacteria. They produce sulfide as the end product of the dissimilative sulfate reduction. Carbon dioxide serves as electron acceptor for the strictly anaerobic methanogenic bacteria which reduce it to methane. Under some circumstances, methane production may arise from the reduction of some organic compounds. All of the above bacterial groups (denitrifiers, sulfate reducers, methanogens) are able to mineralize organic matter to CO2 . Thus they are considered as the terminal mineralizers of organic matter in anoxic environments. During methanogenesis in anaerobic environments, degradable organic matter is converted to CH4 and CO2 . During this anaerobic process, about 90% of the available energy is retained in the methane produced, with a relatively low yield of microbial cells. (MACKIE and BRYANT 1990). The relatively small release of energy must be divided and distributed among the different bacteria involved in the sequential process of anaerobic degradation. Most of the information on anaerobic degradation comes from studies on ruminal fermentation, bacterial enrichment cultures and fermentation of sewage sludge. The basic difference between aerobic and anaerobic oxidation is that in the aerobic system, oxygen is the ultimate hydrogen acceptor with a large release of energy, but in anaerobic systems the ultimate hydrogen acceptor may be nitrate, sulfate or an organic compound with a much lower release of energy. The process of anaerobic decomposition of organic material involves discrete stages. Briefly, insoluble organics (via hydrolytic bacteria) give soluble organics (via acid-forming bacteria) give volatile acids (via methanogenic bacteria) give gases. Although bacteria are the major group of microbes involved in anaerobic metabolism, fermentative ciliate and flagellate protozoa and some anaerobic fungi also occur. It is convenient to think of these stages as different trophic levels, and although all three stages are normally occurring simultaneously within an active system, the micro-organisms involved at each stage are metabolically dependent on each other for survival. For example, the methanogenic bacteria require the catabolized end products of the acid forming bacteria. The first stage is the hydrolysis of high molecular weight carbohydrates, fats and proteins that are often insoluble, by enzymatic action into soluble polymers. In the first stage, the major substrates in the organic material are hydrolyzed to basic components; proteins to amino acids, fats to glycerol and long-chain fatty acids, and polysaccharides to mono and disaccharides. Proteins are hydrolyzed to smaller units such as polypeptides, oligopeptides, or amino acids by extracellular enzymes called proteases, which are produced by only a small proportion of the bacteria. The majority of bacteria are able to utilize these smaller peptides or the amino acids, which pass through the cell wall and are broken down intracellularly. Little is known about the lipolytic bacteria even though they have been shown to be highly effective in anaerobic digesters. They are present in densities of up to 700000/ml and the addition of vegetable oil to digesters to enhance gas production is commonly practiced in some countries (GRAY 1989). The second stage involves the acid-forming bacteria which convert the soluble polymers into a range of organic acids (acetic, butyric and propionic acids), alcohols, hydrogen and carbon dioxide. Acetic acid, hydrogen and carbon dioxide are the only end-products of the acid production that can be converted directly into methane by methanogenic bacteria. The heterogeneous group of facultative and anaerobic bacteria, which are responsible for hydrolysis are also responsible for acid formation. In this second stage, the hydrolyzed substrate is converted to organic acids and alcohols, with new cells also being produced. Various biochemical pathways are utilized, including fermentation and beta-oxidation. There is very little stabilization of the substrate in terms of BOD or COD removal, with the products of acid fermentation being large organic molecules. Mono and disaccharides, long-chain fatty acids, glycerol, amino acids, and short-chain peptides provide the main carbon source for growth, with saturated fatty acids, carbon dioxide and ammonia being the main end-products. Alcohols, aldehydes and ketones are also produced. The third stage is when the organic acids and alcohols are converted to acetic acid by acetogenic bacteria. It is in the fourth and final stage, when methanogenic bacteria convert the acetic acid to methane. Methanogens are unusual in that they are composed of many species with very different cell morphology. They require a strict anaerobic environment for growth with a redox potential below -300 mV (GRAY 1989). They have simple nutritional requirements; CO2 , NH3 and sulfide. Ammonia is the essential nitrogen source for growth and no methanogen species are known to utilize amino acids or peptides. In the overall anaerobic fermentation of carbohydrate to CO2 and CH4, equal volumes are produced. The carbon dioxide evolved partially escapes as a gas because it is soluble in water. It will react with any OH- ions to form bicarbonate. During methanogenesis in anaerobic environments, degradable organic matter is converted to CH4 and CO2 . During this anaerobic process, about 90% of the available energy is retained in the methane produced, with a relatively low yield of microbial cells. The relatively small release of energy must be divided and distributed among the different bacteria involved in the sequential process of anaerobic degradation. The bacteria responsible for methanogenesis are similar in different environments, although little is known about them because of the problems with isolating and maintaining cultures of bacteria under anaerobic conditions. Most of the information on anaerobic degradation comes from studies on ruminal fermentation, bacterial enrichment cultures and fermentation of sewage sludge. Anaerobic digestion and methane production are not unique to anaerobic digesters, they occur in natural environments including the digestive tract of most animals, in the sediments of lakes and rivers, and in estuaries, swamps, marshes and bogs. In anaerobic subsurface sediments and aquifers, chlorinated alkenes may be converted by reductive dehalogenation to dichloroethylene and ultimately to more potent carcinogens such as vinyl chloride. Transformations may be the result of the microbes actually using the contaminant as a substrate for growth (direct metabolism) or the transformation may be the result of some metabolic process which yields a product that cannot be utilized as a substrate for growth (cometabolism) There may be degradation of a contaminant due to an enzyme of broad activity that is able to make some cleavage of a contaminant that results on a non-food end product. The anaerobic consortium is credited with degradation of a wide variety of organic chemicals. Under anaerobic conditions, microbial activities tend to attack complex aromatic structures at substituent groups and convert the parent substrates to hydroxylated, carboxylated, or amino derivatives. (KUHN et al 1989). Reductive Dehalogenation Chlorinated compounds are the most extensively studied because of the highly publicized problems associated with DDT, other pesticides and numerous chlorinated solvents. Most of the information available on the biodegradation of chlorinated compounds is on oxidative degradation, since aerobic culture techniques are relatively simple compared with anaerobic culture methods. For convenience, the chlorinated hydrocarbons degraded by microorganisms (bacteria and fungi) are grouped in to three classes (i) aliphatic, (ii) polycyclic, and (iii) aromatic. Pathways have been elucidated for a number of degradation processes of these classes of compounds. A number of organisms have been isolated and, in many cases, plasmids linked to the metabolic capability have been identified. The objective being to introduce degradative ability into organisms for use in bioremedial schemes. To establish the potential applications of the recombinant strains in the environment, the strains must be stable members of the indigeno us microflora and the recruitment of catabolic enzymes and gene regulators with appropriate effector specificities (by natural gene transfer or by laboratory manipulation) to create new hybrid pathways for chlorinated compounds must not significantly alte r the host or the natural ecosystem. Although abiotic transformations can be significant within the time scales commonly associated with groundwater movement, the biotic processes typically proceed much faster, provided that there are sufficient substrates, nutrients and microbial population s to mediate such transformations. It is widely reported that strongly reducing environments are associated with dehalogenation reactions, although the mechanism appears to be unknown. There is evidence for the dehalogenation being associated with the transition from aerobic to anaerobic conditions (KAESTNER 1991). The reductive dechlorination of PCE (perchloroethylene) via TCE (trichloroethylene) depended on specific transition conditions after consumption of the electron acceptor oxygen or nitrate. Transformation required an additional decrease in the redox potential caused by sulfide. The decrease must be considered the driving force for the onset of reductive dechlorination. Repeated feeding of TCE or PCE to cultures after the change to anaerobic conditions yielded no further dechlorination. In cultures that were anaerobic from the beginning of incubation, no transformation of PCE was observed. The transformation may be catalyzed by the aerobic or facultatively anaerobic organisms if the redox potential drops to nonphysiological values. Transformation in the dechlorinating cultures occurred when the redox potential in the medium decreased to values between -50 and -150 mV and when carbon sources (electron donors) were present in excess. After consumption of the electron acceptor oxygen or nitrate by growth of the aerobic bacteria, the redox potential in the medium reached only to 0 mV. To reach the low redox potential required for dechlorination, a further decrease in the redox potential caused by sulfide was necessary. However dechlorination was also stimulated by the release of sulfide from the degradation of organic sulfur compounds without the growth of sulfidogenic bacteria. The number of cells in dechlorinating culture was already decreasing during dechlorination. This result implies that the release of cell compounds from the dying cells may also be involved in dechlorination.(KAESTNER 1991). The influence of supplemental organic substrates on the degradation of xenobiotics in anaerobic conditions was reported by GIBSON and SUFLITA (1990). They noted that degradation of 2,4,5-T by anaerobic bacteria had a shorter acclimation period at times of the year when the groundwater from the methanogenic site had dissolved humic material contained in it. The conclusion was that reductive dehalogenation reactions are limited by the availability of suitable electron donors. The hypothesis was confirmed by the inclusion of common fermentation products in their anaerobic mixtures. Butyrate, propionate, or ethanol additions had the greatest stimulatory effect on anaerobic 2,4,5-T metabolism and reduced the acclimation time to less than one month (from 3 months without the added substrate). Acetate and methanol were less stimulatory and a 2-3 month acclimation was observed with these treatments. Not only did the carbon amendments stimulate the onset of dehalogenation, they also increased the extent of 2,4,5-T metabolism. Biotransformation under methanogenic conditions of several other chlorinated aliphatics in widespread use (e.g. tetrachloroethylene and trichloroethylene) occurs primarily by reductive dechlorination. This process requires the supply of an external electron donor.(FREEDMAN and GOSSETT, 1991). Anaerobic removal of the Cl substituents proceeds via reductive dechlorination before the aromatic ring is cleaved. This dehalogenation results in the formation of less toxic, more soluble, and less recalcitrant compounds. (HENDRIKSEN et al. 1991) In general, reductive dehalogenation reactions are favored under highly reducing methanogenic conditions. (KUHN et al. 1990) Recent research has shown that natural gas may stimulate TCE degradation in aerobic sediment samples. These results suggest that methane-utilizing bacteria may be able to biodegrade chlorinated alkenes such as TCE. Methanotrophic bacteria are ubiquitous organisms that posses a unique methane monooxygenase enzyme system which enables them to utilize methane as a sole carbon and energy source. The methane monooxygenase enzyme complex has a low substrate specificity and is able to oxidize or dechlorinate a wide variety of economically and environmentally important compounds. The bacteria used was isolated from a waste well where chlorinated organic solvents were disposed of directly to the groundwater. Contamination ranges from low to very high (100 mg/l). The inability of this organism to degrade TCE in the absence of methane or methanol suggests that TCE biodegradation by methanotrophs is a cometabolic process which provides little or no metabolic benefit to the organism. The fact that most aquifers are not well supplied with oxygen will favor anaerobic reactions. Recharge from stripping towers used to remove volatile organic compounds from contaminated aquifers will introduce oxygen into the recharge zone (as well as spores from the surface) and create aerobic conditions in the immediate area of the recharge well. (FREEDMAN and GOSSETT, 1991). Nitrosomonas europea is an obligate chemolithotrophic nitrifying bacterium which derives its energy for growth exclusively from the oxidation of ammonia to nitrite. Evidence indicates that ammonia monooxygenase in cells of this bacteria is also capable of cooxidizing hydrocarbons, including DBCP and TCE (RASCHE et al, 1991). ALLARD, et al, (1991) studied degradation of chlorocatechols by metabolically stable anaerobic cultures. A high degree of specificity in dechlorination was observed, and some chlorocatechols were appreciably more resistant to dechlorination than others: only 3,5-dichlorocatechol, 4,5-dichlorocatechol, 3,4,5-trichlorocatechol, and tetrachlorocatechol were dechlorinated, and not all of them were dechlorinated by the same consortium. 3,5-dichlorocatechol produced 3-chlorocatechol, 4,5-dichlorocatechol; tetrachlorocatechol produced only 3,4,6-trichlorocatechol. Incubation of uncontaminated sediments without additional carbon sources brought about dechlorination of 3,4,5-trichlorocatechol to 3,5-dichlorocatechol. O-demethylation of chloroguaiacols was generally accomplished by enrichment cultures, except that catechin enrichment was unable to O-demethylate tetrachloroguaiacol. None of the enrichments dechlorinated any of the polychlorinated phenols examined. The results suggested that dechlorination was not dependent on enrichment with or growth at the expense of chlorinated compounds. They concluded that the transformations described were mediated by bacterial reactions. CRIDDLE, et al (1990) suggest that the formation of radicals from carbon tetrachloride may explain the product distribution resulting from its transformation. Radicals formed by reduction of tetrachloromethane (carbon tetrachloride) presumably react with constituents of the surrounding milieu to give the observed product distribution. Use of oxygen and nitrate as electron acceptors generally prevented carbon tetrachloride metabolism. At low oxygen levels (about 1%) transformation of carbon tetrachloride t o CO2 and attachment to bacterial cells material did occur. Under fumarate respiring conditions, the carbon tetrachloride was recovered as CO2 , chloroform and a nonvolatile fraction. In contrast, fermenting conditions resulted in more chloroform, more cell-bound radiolabeled C (from carbon tetrachloride ) and almost no CO2 . Rates of transformation were faster under fermenting conditions than under fumarate respiring conditions. It is also reported that transformation rates decreased over time. (CRIDDLE et al. 1990) A plausible hypothesis for the observation is that many halogenated xenobiotics undergo reductive dehalogenation is that these transformation are fortuitous, resulting from the inherent activity of reducing agents created by microorganisms; i.e. cometabolism. If this hypothesis is correct, then many common and familiar microorganisms may bring about unexpected and possibly unpredictable transformations when confronted with chemicals that are foreign to their evolutionary history. The agents of transformation may be generated by biological activity or they could result from a change in solution chemistry brought about by microbial activity. Trihalomethyl radicals undergo addition reactions at the double bonds of lipids and unsaturated acids. One explanation for the lack of transformation under fully aerobic reactions is a paucity of sufficiently powerful reducing agents. (CRIDDLE, et al. 199 0) Pentachlorophenol Pentachlorophenol (PCP), the chlorinated phenols used as wood preservatives, herbicides, fungicides and general biocides are a large group of toxic xenobiotics that are serious environmental pollutants. Reductive dechlorination of PCP has been observed in flooded soils. Actinomycetes and fungal organisms have also been found to metabolize PCP, However, little is known about the microorganisms responsible for the anaerobic degradation of PCP. MADSEN and AAMAND (1991) report degradation of PCP under methanogenic and sulfate-reducing conditions with an anaerobic mixed culture derived from sewage sludge. The consortium degraded PCP via 2,3,4,5-tetrachlorophenol, 3,4,5-trichlorophenol and 3,5-dichlorophenol and eventually accumulated 3-chlorophenol. Dechlorination of PCP and metabolites was inhibited in the presence of sulfate, thiosulfate, and sulfite. A decrease in the rate of PCP transformation was noted when the endogenous dissolved H2 was depleted below 0.11 micromoles per liter in sulfate reducing cultures. The effect on dechlorination observed with sulfate could be relieved by addition of molybdate, a competitive inhibitor of sulfate reduction. Addition of H2 reduced the inhibition observed with sulfoxy anions. The inhibitory effect of sulfuoxy anions may be due to a competition for H2 between sulfate reduction and dechlorination. When cultured under methanogenic conditions, the consortium degraded several chlorinated and brominated phenols. Halogenated phenols are reductively dehalogenated in sewage sludge, aquatic sediments and soils. Some of these studies indicate that reductive dehalogenation reactions may be favored in methanogenic environments. Experiments with anaerobic groundwater sediment showed that a dehalogenating potential was present in the methanogenic part of he aquifer, but this potential was at least partially inhibited by the high concentration of sulfate at the nearby sulfate-reducing site. Sulfate is commonly present in anaerobic habitats, such as aquatic sediments, soil and wastewater sludge. Bromacil Dehalogenation The metabolic fate of bromacil in anaerobic aquifer slurries held under denitrifying, sulfate reducing and methanogenic conditions revealed that bromacil was debrominated under methanogenic conditions but was not degraded under other incubation conditions . (ADRIAN and SUFLITA, 1990). This finding extends the range of aryl reductive dehalogenation reactions to include nitrogen heterocyclic compounds. Creosote Breakdown Unlined surface impoundments at a wood preserving plant in Florida in direct hydraulic contact with the aquifer resulted in two distinct phases when the creosote and water mixed. (GODSY et al. 1992). A denser than water hydrocarbon phase that moved vertically downward, and an organic-rich aqueous phase that moved laterally with the groundwater flow. The aqueous phase was enriched in organic acids, phenolic compounds, and single- and double ring aromatic hydrocarbons. The ground water was devoid of dissolved O2, was 60-70% saturated with CH4 and contained H2S. Field analyses documented a greater decrease in concentration of organic fatty acids, benzioc acid, phenol, 2-,3-,4-methylphenol, quinoline, isoquinoline, 1(2H)- quinolinone, and 2(1H)-isoquinolinone during downgradient movement in the aquifer than could be explained by dilution and/or dispersion. Laboratory studies showed that within the study region, this effect could be attributed to microbial degradation to CH4 and CO2 . A small but active methanogenic population was found on sediment materials taken from highly contaminated parts of the aquifer. Creosote is a complex mixture of more than 200 major individual compounds with differing molecular weights, polarities, and functionalities, along with dispersed solids and products of polymerization. The major classes of compounds previously identified in creosote show that it consists of approximately 85% by weight polynuclear aromatic hydrocarbons (PAH), 12% phenolic compounds and 3% heterocyclic nitrogen, sulfur and oxygen containing compounds (NSO). During the first 50 days of residence in the microcosm, C3-C6 volatile fatty acids were rapidly converted to acetic acid and ultimately to CH4 and CO2 . Benzoic acid, quinoline, and isoquinoline are also biodegraded. Phenol degradation occurs between days 50 and 99 in the microcosms and phenol also disappears from the ground water during transit. After 100 days in the microcosm, 2-,3-,4- methylphenol, 2(1H)-quinolinone and 1(2H)-isoquinolinone are biodegraded and are removed from the system after about 20 0 days. A similar pattern of disappearance is observed for 2-, 3-, and 4-methylphenol. The degradation of 2-methylphenol appears to be somewhat slower than the other methylphenols in the groundwater. This compound, which is widely held to be recalcitrant, was readily degraded during downgradient transport in the aquifer and the microcosm used in the cited study. (GODSY et al. 1992). Substituted Indoles Aromatic N-heterocyclic compounds, including substituted indoles are often found in aqueous waste effluents associated with oil shale and coal mining operations. Not surprisingly, the ability of sediment and sewage sludge microcosms to degrade indole is dependent upon several factors, including incubation temperature and the amount of sediment or sludge inoculum used. Degradation of indole by an indole-degrading methanogenic consortium enriched from sewage sludge proceeded through a two-step hydroxylation pathway yielding oxindole and isatin. (GU and BERRY, 1991). Under anaerobic conditions, the source of oxygen for th e hydroxylation reaction is water (a nucleophile). It is found that an indole degrading methanogenic consortium is capable of transforming 3-methylindole and 3-indolyl acetate. In neither case were the aromatic ring structures catabolized. Cyanide breakdown Upflow, anaerobic, fixed-bed activated charcoal biotreatment columns capable of operating at free cyanide concentrations of > 100 mg/l with a hydraulic retention time of <48h were developed. (FALLON et al. 1991). Methanogenesis was maintained under a vari ety of feed medium conditions which included ethanol, phenol or methanol as the primary reduced carbon source. Under optimal conditions, >70% of the inflow free cyanide was removed in the first 30% of the column height. Strongly complexed cyanides were resistant to removal. Ammonia was the nitrogen end product of cyanide transformation . In cell material removed from the charcoal columns, bicarbonate was the major carbon end product . Cyanide spontaneously hydrolyzes to formic acid at a rate positively correlated with pH, temperature, and trace metal concentration. Therefore, under the right conditions, with a long enough retention time, the cyanide spontaneously disappears. ABIOTIC DEGRADATION Abiotic Transformation In real systems, it is generally not possible to distinguish between biotic and abiotic; or even aerobic and anaerobic transformation events. Numerous transformations occur in the homogeneous phases, especially in the liquid phase. Other transformations occur in the interface between phases. These include reactions that are heterogeneously catalyzed and those that occur in solution under the influence of the electric field of charged surfaces. Biologically produced enzyme and other biochemical compounds in the soil can be involved in transformations of pollutants. Thus, sterilization, which destroys living organisms, will also affect the abiotic chemical reactions that are dependent on substances generated by biological processes. Sterilization processes, furthermore, can alter nonbiological constituents of the treated systems. Heat and radiation, for example, can affect the free-radical content of the soil. Finally, many degradative pathways include both biologically and chemically controlled steps. Abiotic degradation occurs when a chemical undergoes a thermodynamically favorable change without the catalyst of biological enzymes. This includes a large number of potential reactions, including those catalyzed by hydrogen ions. Hydrogen ion activity affects transformation kinetics by two major mechanisms; acid-base mediated hydrolysis reactions and dissociation of acidic or basic compounds. In addition to pH induced reactions, other abiotic processes may involve reactions such as those involving dissolved organics and suspended particles, metal ions, redox reactions. Natural sorbents may be biotic and abiotic; may be organic, inorganic, or chemical composites thereof and may range in size from macromolecules to gravel. The surface charge and chemical composition may change with ambient solution pH and redox potential.These changing conditions may affect the colloidal size, and molecular configuration. Pollutants vary in water solubility from complete miscibility to virtual insolubility. As a function of ambient conditions, any given material can undergo chemical reaction. If a thermodynamic gradient exists and sufficient activation energy, or a suitable catalyst is provided, a reaction is possible. The actual reaction event, or on a larger scale, the rate of the reaction, may be kinetically limited. Degradation processes always flow in the direction of the least energy. Oxidation is a favored reaction because of the stability of the products in an oxidizing environment. Reduction is a favored reaction in a reducing environment because of the stability of the products in a reducing environment. Sometimes activation energies prevent thermodynamically favorable reactions from happening. Catalysts, such as enzymes and surface interfaces, help to allow a favorable reaction to occur by; (1) lowering the activation energy so the reaction proceeds under ambient conditions of temperature, pH, pE, etc. ; or (2) by increasing the relative concentrations of the reactive species at the reactive surfaces. SORPTION Phase interfaces are important. Most chemical reactions that occur in water takes place at phase discontinuities, such as the air-water or solid-water interfaces (STUMM and MORGAN, 1981). Sorption onto a surface can alter the configuration or energy status of a molecule in such a way as to enable a reaction. The physical process of adsorption onto a surface causes changes in the conformation or arrangement of the bonds in the adsorbed species (BARROW, 1973). Such changes may increase the rate of a reaction and thus be considered a catalytic effect. The catalysis of chemical reactions by certain surfaces is an important process. Mineral clays are reported to catalyze some reactions involving organic chemicals (MORTLAND, 1985). In addition to catalysis, concentration of material at a surface can increase the effective concentration of reactants and thus enable reactions that might not be possible in dilute systems. Adsorption is the concentration of a component on the external surface at an interface while absorption is usually meant to describe the movement of something into the interior of a matrix. Because of the difficulties in discerning the boundaries of the solid-water interfaces, the more general term "sorption" has frequently been adopted to describe both adsorption and absorption. Sorption is a more generally applicable term which encompasses both processes and simply relates interfacial flux. In practice, sorption is usually meant to indicate the movement from the free, or mobile phase (gas or liquid) into or onto the fixed phase. Desorption is used to denote the movement from the fixed phase back into the mobile phase. In actuality, sorption involves two processes; the movement from one phase to another involves changes in both phases and the overall systems of both phases will reflect the event. In the case of adsorption from aqueous solution, it is usually considered that the process is competitive and that something must be removed (desorbed) or re-arranged to accommodate the newly sorbed species. Absorption, on the other hand, can involve the movement from one liquid into another with out the necessity of removing a sorbed species or competing for a site on the surface (MINGELGRIN and GERSTL, 1983). BONDING FORCES A number of different attractive forces may be involved in sorption and different mechanisms have been proposed to explain the process by which a given species can partition between phases. The mechanisms act between all of the system constituents; i.e., species in solution can interact with each other, they can also react with the water molecules or with the solid phase to influence the overall system behavior. Covalent bonds are thought to involve the sharing of electrons between adjacent atoms in a molecule. The formation of such bonds between species is a possible form of partition interaction. In such an arrangement, the bonding electrons of the reactants share orbitals and form relatively strong bonds. These types of bonds are typically exothermic, largely irreversible, and frequently require an activation energy or catalyst to form. Uneven distribution of electrons due to differences in electronegativity between molecular substituents can result in the formation of a dipole such as water. Molecules with a net dipole moment can undergo mutual bonding interactions by virtue of localized areas of charge resulting from uneven concentrations of electrons relative to the positively charged nuclei of the constituent atoms. Because water is polar, polar materials tend to be more or less soluble in water. In much the same way as the dipoles in the water milieu can form positive associations with polar contaminants, so can polar molecules other than water form polar associations with polar contaminants. Hence, polar interactions can serve to partition contaminants into or out of the aqueous phase. Molecules in the solid phase can, by virtue of unequal distribution of electrons, affect molecular orientation and distribution by polar interaction. Hydrogen bonding is a very important polar interaction in water. The importance of hydrogen bonding in aqueous systems is such that there is often a tendency to consider hydrogen bonding as some sort of special or unique bond type. It may well be considered as an extreme manifestation of dipole-dipole interactions which typically arises when hydrogen is attached to very electronegative atoms. Hydrogen bonding also occurs in some other polar liquids such as alcohols. Molecular excess or deficiency of electrons relative to protons can produce anions or cations which posses a static charge. Electrostatic charge can influence the distribution or orientation of charged or polar species in solution. Species which have net charge can be dissolved by the water matrix in a manner somewhat akin to the polar mechanism of solution described above. Water can act to dissolve ions in a crystal, when the force of attraction of the water dipoles is greater than the force of attraction of the bonds in the crystal lattice of the solid phase. The presence of an ion in solution will cause a certain amount of "ordering" or nonrandom arrangement of the water molecules as the dipoles orient their polar ends in response to the standing charge on the central ion, coordinated as ligands in a sphere of hydration around the ion. The formation of neutral ion pairs in solution can occur when oppositely charged ions become attached to each other and the attraction or motion of the water molecules is sufficient to maintain the ion pair in the dissolved state against the forces of gravity or attraction to the crystalline solid phase. Ions in solution can affect the distribution and status of other dissolved ions as well. An ion in solution may be surrounded by neutral or oppositely charged ligands to form a complex. Chelation is an especially strong sort of complex whereby a ligand forms two or more bonds with the central coordinating ion. A single ligand capable of forming more than one association is termed multidentate or polydentate. In addition to the interaction of charged species in solution, sorption onto the solid phase is also possible due to the interaction of a charged surface with the dissolved ions in solution. A charged surface will attract oppositely charged counterions and result in a relative concentration of counterions in the interfacial region. The interfacial region between a charged surface and a solution containing ions is considered to be "ordered", relative to the ionic distribution in the bulk phase, as counterions in solution are distributed in response to the static charge on the surface. This has been called an electrical double layer. There have been several descriptive models advanced to relate the distribution of ions in solution in response to the existence of a charged interface (STUMM and MORGAN, 1981). The thickness of the double layer will diminish with the ionic strength of the solution. As the concentration of charge in solution increases (i.e., the ionic strength of the solution increases), the total concentration of countercharges adjacent to the charged surface increases, and the thickness of the layer of counterions decreases or collapses in proportion to the ionic strength of the bulk solution. An increase in solution ionic strength may have an impact on the processes of adsorption at the surface, not only by decreasing the thickness of the electrical field surrounding the charged surface, but by increasing the competition for adsorption by other solution components. Ionic species can induce a dipole in a nonpolar molecule over a short range. London forces exist between instantaneous and induced dipoles, and are operative between all bodies when they are close together. They are also commonly called van der Waals attr active forces after the Dutch physicist (J.D. van der Waals) who described these forces as being active in crystals (PAULING, 1957). The London/van der Waals force is also frequently referred to as the dispersion force. This force is very important in solution phase as well. The nature of the London force is that it is proportional to the molecular volume and the number of polarizable electrons of the species experiencing the force. Even nonpolar neutral species undergo momentary imbalances in electron distribution. The forces which exist between instantaneous dipoles are responsible for much of the interactive cohesion in solutions of nonpolar liquids. The impact of the London force on sorption from solution tends to become pronounced when large molecules are involved; larger molecules have a larger molecular volume and more electrons. It is thought that the essence of the van der Waals force is the attraction of electrons of one molecule for the atomic nuclei of another (PAULING, 1957). The ability of species to engage in van der Waals bonding is related to the number of electrons and to the ability of those electrons to accommodate the close approach of the bonding partner's electrons. This latter ability is called polarizability and may be thought of as the ease of inducing a dipole moment in a species. As a result of the nature of the intermolecular interaction which gives rise to van der Waals force, this force is only active at very close range. The molecules must approach one another closely before the attraction which results in sorption can exert itself. It is generally believed that the force of the van der Waals attraction between two molecules is proportional (a) to the square of the polarizability and varies inversely with the sixth power of the distance between the molecules. Q (proportional) n2/ r6 (1) where Q = force of the attraction between molecules n = polarizability r = distance between the molecules The variation of the energy of attraction attributed to van der Waals force as a function of distance between sorbate and sorbent may be described graphically with a hypothetical plot of potential energy vs distance. (ascii format does not support this graphic- see original report for graph) Figure 1. Hypothetical plot of van der Waals attraction as a function of distance between interacting molecules At distances greater than a few molecular diameters, the energy of attraction is negligible. As the molecules approach, the force of attraction increases (the potential energy decreases) as natural or induced dipoles begin to interact. As the molecules grow even closer, steric factors come into play and the potential increases dramatically. The point of minimum potential energy, then, is the point of maximum attraction and relates to the point of closest approach. These bonding interactions described are frequently considered to be representative of the major types of forces which exist between species, although there is some disagreement about the nature and magnitude of the forces involved. It is probable that combination or hybrid forces come into play in real material interactions. It is also probable that multiple types of attractive and repulsive forces act simultaneously in many complex systems (e.g., WERSHAW and PINKNEY, 1973). FREE ENERGY Partitioning is governed by free energy change. The net free energy describes the overall tendency of the system to make a specific change. The concept is in accord with the laws of thermodynamics and assumes that it is the natural tendency of a system to spontaneously seek a condition of minimum energy and maximum disorder. The most common form of the equation is delta G = delta H - T delta S (2) where delta G is the change in free energy associated with the event delta H is the change in enthalpy T is the absolute temperature delta S is the change in entropy which accompanies the event. The consideration of net free energy is associated with a specified change and demands clear definitions of the system under consideration, both before and after the change. The value of the free energy relation is that spontaneous reactions must always be associated with a negative change in free energy i.e., delta G < 0. If delta G is greater than zero, the reverse reaction is thermodynamically favored. The free energy of a sorption process can, in principle, be determined from K, the slope of the linear isothermal plot according to the equation delta G = RT ln K (3) where R = the gas constant T = the absolute temperature Quantitative application of free energy data requires rigorous definitions of the system. Since the equilibrium constant for the distribution between the bulk and surface phases (K) is not well defined due to the uncertainty in the thickness (volume) of the adsorption layer, the values of delta G are only approximate (MINGELGRIN and GERSTL, 1983). The tendency of the system to minimize its energy is accounted for by considering the energy (enthalpy, H) contained in the bonds or forces of association between the system components before and after the specified change. If the net energy of bonds is lower in the system after the change, the change is considered to be favorable from the aspect of net enthalpy. The free energy concept accounts for the tendency of the system to maximize disorder through the entropy term; S. The entropy of the system is directly related to the numbers of system components and the freedom of random motion of the system before and after the specified change. It is incorrect to assume that adsorption always represents a decrease in system entropy. Adsorption at the surface by a solute component may require the removal of another species which is adsorbed to the surface, hence the increased order or disorder of the system accompanying competitive adsorption from solution is not so clear cut as might be the case of adsorption of a gas molecule from a near-vacuum. The transfer of a hydrophobic solute from aqueous solution across a phase boundary into an immiscible liquid phase is reported to represent an increase in entropy. Two major sources of entropy increase have been suggested. One is that hydrophobic solutes lead to increased structuring of water. Decreased structuring when the solute leaves the aqueous phase would increase randomness in water and therefore increase entropy. Another cause of increase in entropy is greater conformational freedom of hydrophobic molecules in non-aqueous media than in water. The increase in structural conformations leads to an increase in randomness and an increase in entropy (HASSETT and ANDERSON, 1982). Entropy changes in complex systems may be difficult to enumerate. In fact, spontaneous events (i.e. those with delta G < 0) are observed to display variations in both magnitude and sign for enthalpy (H) and entropy (S) changes (MINGELGRIN and GERSTL, 1983; OPPERHUIZEN et al., 1988). It is the combination of these two parameters, along with the consideration of the temperature (T), which describes the net free energy, and hence the opportunity for a spontaneous event. In any case it is well to remember that the existence of a favorable free energy gradient (delta G < 0) does not guarantee that an event will occur within any time frame. Kinetics are not considered in the free energy determination, nor is the existence of an activation energy. An event may have a favorable free energy gradient and yet be limited by the kinetics or activation energy requirements. INTERFACIAL TENSION Water molecules at the air-water interface experience unbalanced attraction for the water and the air. This is a manifestation of the polar nature of water in contact with a nonpolar phase (the air). The water molecules are drawn together, resulting in a phenomenon called "surface tension". The contact area between the water and the nonpolar phase is a region of relatively high interfacial tension and the system will naturally tend to minimize such contact. This polar structure of water will tend to make the aqueous medium relatively inhospitable to nonpolar neutral (uncharged) molecules as well (HORVATH, et al.,1976). A nonpolar neutral species in a polar medium such as water experiences interfacial tension. Solvophobic theory is a general statement of hydrophobic theory which has been developed to explain the tendency of neutral organic species to flee the water phase. It has been reported that the solution of nonelectrolytes in water is attended by a net decrease in entropy (EGANHOUSE and CALDER, 1976). This has been attributed to an increased structuring of water molecules in the vicinity of the solute. The process may be conceptually rationalized by considering that a solute must occupy space in a cohesive medium. The solute must create a "cavity" in the water milieu and then occupy that cavity. (e.g., McDEVIT and LONG, 1952; AMIDON et al., 1974; YALKOWSKY et al.,197 5; BRIGGS, 1981). The very high cohesive density of water creates considerable interfacial tension in the region of contact with a nonpolar solute and is responsible for the magnitude of the hydrophobic effect. This interfacial tension has also been called the internal pressure (GORDON and THORNE, 1967) and it creates a driving force for the nonelectrolyte to flee the solution as the system tries to minimize the area of contact between the water and the nonpolar solute. This is a more rigorous way of saying that oil and water don't mix. The hydrophobic concept has been of great utility in explaining the behavior of organic chemicals in water. Hydrophobic forces can drive nonpolar neutral solutes across an interfacial boundary into an adjacent immiscible nonpolar liquid. A substantial part of the driving force of this reaction may be a positive entropy change which was described above. What is sometimes called "hydrophobic bonding" is largely the extension of solvophobic behavior to create a partitioning event such as adsorption onto a solid material. The so-called hydrophobic bond is not so much a special type of bond as it is a way for the system to minimize the area of the polar and nonpolar interface (HORVATH et al.,1976). If the site of sorption is itself hydrophobic, sorption of a nonelectrolyte onto such a site will be attended by a proportionally greater reduction in the overall system interfacial tension and the driving force will be that much greater. Upon sorption, London forces are certainly involved and so bonding per se is occurring, but the solvophobic tendency is providing a considerable gradient for the sorption event. A direct consequence of hydrophobic theory is manifest in Traube's rule, which states that the water solubilities of an homologous organic series decrease as the length of the carbon chain increases. As the length of the nonpolar carbon chain increases, so does the nonpolar surface area of the molecule. While a functional group may be relatively polar, the nonpolar surface area creates the interfacial tension in aqueous solution and thus the water solubility will decrease as the chain length increases. Traube's rule accommodates the balance between hydrophobicity and hydrophilicity. Traube's rule has been somewhat extended and formalized with the development of a quantitative methods to estimate the surface area of molecules based on their structures (e.g., YALKOWSKY et al.,1972,1975; AMIDON et al.,1974; HORVATH et al.,1976; BRIGGS 1981). The molecular surface area approach suggests that the number of water molecules that can be packed around the solute molecule plays an important role in the theoretical calculation of the thermodynamic properties of the solution. Hence, the molecular surface area of the solute is an important parameter in the theory. In compounds other than simple alkanes, the functional groups will tend to be more or less polar and thus relatively compatible with the polar water matrix. Hence, the total surface area of the molecule can be subdivided into "functional group surface area" and "hydrocarbonaceous surface area". These quantities may be determined for simple compounds as an additive function of constituent groups with subtractions made for the areas wh ere intramolecular contact is made and thus no external surface is presented (AMIDON et al.,1975). It is found that for molecules with longer carbon chains, the ability to predict solubility based on calculated molecular surface area is diminished. It has been suggested that, as molecular size increases, coiling and self-association of the flexible chains, or perhaps multimolecular aggregate formation spontaneously occurs to minimize the nonpolar surface area (AMIDON et al.,1974). The solubility of organic chemicals in water ranges from complete miscibility to near insolubility. Many natural aqueous systems contain a nonpolar phase such as soil organic matter or lipids in organisms. The octanol/water partition coefficient has arisen as a measure of the water/nonpolar partitioning behavior. This parameter has gained wide use as a key indicator of the environmental fate of organic chemicals. The octanol/water partition coefficient is the ratio of a chemical's concentration in octanol phase to its concentration in the aqueous phase of a two phase octanol/water system. The octanol/water partition coefficient is not the same as the ratio of a chemical's solubility in octanol to that in wa ter, because the organic and aqueous phases of the binary octanol/water system are not pure octanol and pure water (CHIOU and FREED, 1977b; CHIOU et al.,1982). The octanol water partition coefficient is a useful indicator of the hydrophobicity of an orga nic chemical. SALT EFFECTS Natural water is not pure and it has been observed that solution modifications affect the behavior of solutes. The presence of dissolved salts in solution has been observed to both increase and decrease the solubility of neutral nonelectrolytes. "Salting out" is the term that has been used to describe the decreased solubility of nonelectrolytes as a function of concentration of simple salts in solution (McDEVIT and LONG, 1952; AQUAN-YEUN et al.,1979). Such behavior is consistent with the hydrophobic theory discussed above. The addition of simple salt to water increases the interfacial tension between water and nonpolar phases and will thus tend to force nonelectrolytes out of solution. The tendency to "salt-out" is apparently countered when salts containing large complex ions are used for the electrolyte. In such cases, the solubility of the nonelectrolyte is frequently seen to increase, in a phenomenon that has been logically called "salting-in". The tendency to salt-in is seen to increase with the size of the ions involved, whether they are cations or anions. This seems to indicate that additional interaction terms of the van der Waals type must be considered (LONG and McDEVIT, 1952) . Such interactive forces involve the polarizability of salt ions, solvent molecules and non-electrolyte solute molecules as well as forces between any component dipoles that may be present (SAYLOR et al.,1952; STEIGMAN et al.,1965). The effect of salts on the solubility of nonelectrolytes may be described by the Setschenow equation; log f = ksCs (4) where f = So/S = molar activity coefficient of the nonelectrolyte So = molar solubility of the nonelectrolyte in pure water S = molar solubility of the nonelectrolyte in salt solution ks = parameter dependent on the particular salt Cs = the molarity of the salt In any case the effectiveness of different electrolytes to create these effects in solution is considerable. (ascii format does not support this graphic- see original report for graph) Figure 2. Effect of salts (approximate) on the activity coefficient of benzene (after McDevit and Long, 1952) It is important to note that the effect of salt on nonelectrolyte solubility (benzene in this case) is significant only at salt concentrations that exceed those typical in fresh waters. COSOLVENCY Polar neutral organics can be very miscible in water by virtue of compatibility with the polar water molecules; for example dipole-dipole interactions such as those interactions between short-chain alcohols and water give rise to essentially complete miscibility. In contrast to the increase in surface tension accompanying solutions of salts, miscible organics in solution tend to decrease the surface tension of the aqueous medium. (ascii format does not support this graphic- see original report for graph) Figure 3. Approximate surface tension as a function of the composition in mixed solvents and salt solutions (after Horvath et al.,1976). Miscible organic solutes modify the solvent properties of the solution to decrease the interfacial tension and give rise to an enhanced solubility of organic chemicals in a phenomenon often called "cosolvency". According to theory, a miscible organic chemical such as a short chain alcohol, will have the effect of modifying the structure of the water in which it is dissolved. On the macroscopic scale, this will manifest as a decrease in the surface tension of the solution. It is noted that surface tension is a gross parameter and experimentally determined interfacial tensions (that which exist between the solute and solvent species) are generally less than would be predicted based on surface tension measurement ( YALKOWSKY et al.,1976). It has been generally considered that there is an exponential increase in the solubility of a solute as the fraction of the cosolvent increases linearly. The only requirement for the log linear relationship seems to be that the solute must be less polar than the mixed solvent (YALKOWSKY et al.,1976). The validity of the log-linear nature of the cosolvent process has been well validated in the literature (e.g., NKEDI-KIZZA et al.,1985, 1987; RAO et al.,1985; WOODBURN et al.,1986,1989; WALTERS and GUISEPPI- ELIE, 1988; WALTERS et al.,1989). The effect of a cosolvent on solubility has been figured according to the following equation: ln (Sm) = fc ln (Sc) + (1 - fc) ln (Sw) (5) where Sm = molar solubility of a nonpolar solute fc = nominal cosolvent volume fraction Sc = molar solubility in pure cosolvent Sw = molar solubility in pure water This model assumes the absence of specific solute-solvent interactions and is based upon a linear relationship between the free energy of solution and solute surface area. It assumes that the overall solubility is simply the sum of the solubilities in the individual solvent components. This model treats the cosolvent and the water as distinct entities and neglects any interaction between them. More recent work with cosolvency in dilute systems seems to indicate that the magnitude of the solubility enhancement is linear up to some 10-20% cosolvent fraction (BANERJEE, 1985; BANERJEE and CASTROGIVANNI, 1987; BANERJEE and YALKOWSKY, 1988; ZACHARA et al.,1988). At very low concentrations of cosolvent, the assumption of non-interaction between the cosolvent and water cannot hold. In dilute solutions the individual cosolvent molecules will be fully hydrated, and as a result, will disrupt the water network structure. If the total volume disrupted is regarded as the extended hydration shell, and if Sc* is the average solubility within this shell, then the overall solubility Sm in the water-cosolvent mixture will be approximated by Sm = fcVH Sc* + (1- fcVH) Sw ; fcVH < 1 (6) where VH is the ratio of the hydration shell volume to the volume of the cosolvent. In dilute solutions, the solute will, on average, contact only one hydrated cosolvent molecule at a time, and the degree of solubilization should be a linear rather than a logarithmic function of cosolvent content. Thus, it is expected that the log-linear relationship between Sm and fc that applies at high cosolvent concentrations will become linear at low cosolvent levels due to a change in the mechanism of solubilization. If S+ is defined as solubility enhancement , (Sm - Sw), then the relative solubility enhancement at low cosolvent concentration will be given by S+/Sw = fc VH(Sc*/ Sw -1) (7) While cosolvency has been applied to environmental water chemistry discussions, it is important to point out that the principle was originally described by pharmaceutical chemists interested in solubilizing nonpolar drugs. The presence of 20% miscible cosolvent in a therapeutic treatment for a mammal is somewhat different from 20% cosolvent in a natural water system. In a natural water system where cosolvent was present at sufficient levels to influence contaminant solubility, the cosolvent itself would probably constitute a contaminant. In a contaminated groundwater, however, such a cosolvent concentration may be realistic to create, thereby significantly enhancing the degradation of the target pollutant. If the cosolvent were itself biodegradable, the resulting effect would be the removal of the pollutant without adverse long-term effects on the resource. The log-linear solubility enhancement by cosolutes may be important in characterizing concentrated leachate plumes or chemical spills, but will be of little importance in characterizations of the dilute aqueous systems that predominate in nature (NKEDI-KI ZZA et al.,1985; FU and LUTHY, 1986). MICELLES Organic chemicals can be quite variable in structure and properties. Many organic molecules have both polar and nonpolar moieties, and the solubility of the material in water will be the result of a balance between the hydrophobic and hydrophilic tendencies. If a molecule containing a hydrophilic region also has a significant hydrophobic region, such as a long carbon chain, the water solubility will be diminished. This diminished solubility can be manifest in several ways. The chemical can sorb onto a surface and thereby diminish the interfacial tension with the water or it can form a separate, immiscible bulk phase. A third possibility exists, whereby the nonpolar moiety can undergo association with the nonpolar regions of other molecules to form smaller subunits within the water matrix. Such an organizational arrangement minimizes the contact between the hydrophobic moieties and the water while allowing the hydrophilic (polar/ionic) moieties to contact the water. Such an aggregate arrangement is frequently referred to as a micelle. Typically, organic chemicals having both polar and nonpolar moieties can form micelles. Such chemicals are often referred to as "amphiphiles" or described as being "amphipathic", which refers to the dual affinity of such species for both polar and nonpola r media. Surfactants and soaps are amphiphiles. They are often characterized by having a polar or ionic end (or "head") and a nonpolar "hydrocarbonaceous" end (or "tail"). These molecules in solution will to be subjected to the forces of interfacial tension or polar affinity as have been so far delineated. The polar or ionic end will be readily solvated by water, which will repel the nonpolar end. Micelles arise when these molecules undergo intermolecular association of the hydrophobic moieties and form a droplet of material which has a hydrophobic interior and a hydrophilic exterior. The interfacial tension between the water and the hydrophobic end is thus minimized and the droplet may be solvated by its outer "shell" of polar or charged ends in association with the polar water phase. This arrangement has been called a "pseudophase" denoting the existence of a hydrophobic interior of the droplets suspended by the interaction of the hydrophilic moieties with the polar water. It has been observed that the association of homogeneous surfactant monomers to form micelles is characterized by some critical concentration of dissolved monomers before true micelle formation (micellization) occurs. A commonly described parameter associated with micelle formation is the critical micelle concentration- or CMC. The onset of micellization, which occurs at CMC, is typically accompanied by some well-defined or observable change at that point. For example, a visible turbidity may accompany the CMC. It is commonly reported that the addition of surfactant monomer to water can cause the surface tension of the solution to decline steadily until CMC is attained, after which continued addition of monomer produces no more drop in the measured surface tension. The transition is typically a sharp one. Experimentally, it is often found that micelles are undetectable in dilute solutions of the monomers, and become detectable over a narrow range of concentrations as the total concentration of solute is increased, above which nearly all additional solute material forms micelles. The concentration at which the micelles become first detectable depends on the sensitivity of the experimental apparatus used to observe the change in surface tension. The concentration range over which the fraction of additional solute which forms micelles changes from nearly zero to nearly unity depends on such factors as the number of monomers in the micelle, the chain length of the monomer, the properties of counterions and other details affecting the monomer-micelle equilibrium. An approximate rule is that the higher the CMC value, the broader is the concentration range over which this transition takes place, in absolute value as well as in relative value in comparison with the CMC (MUKERJEE, 1971). Since different experimental methods may reflect this transition to different extents, some systematic variations in operationally defined CMC's are expected. The impact of salt concentration on the formation of micelles has been reported and is in apparent accord with the interfacial tension model discussed above, where the CMC is lowered by the addition of simple electrolytes (STEIGMAN et al.,1965; GORDON and THORNE, 1967). The existence of a micellar phase in solution is important not only insofar as it describes the behavior of amphipathic organic chemicals in solution, but the existence of a nonpolar pseudophase can enhance the solubility of other hydrophobic chemicals in solution as they partition into the hydrophobic interior of the micelle. A general expression for the solubility enhancement of a solute by surfactants has been given by KILE and CHIOU (1989), in terms of the concentrations of monomers and micelles and the corresponding solute partition coefficients, giving Sw*/ Sw = 1 + Xmn Kmn + Xmc Kmc (8) where Sw* = apparent solute solubility X = total stoichiometric surfactant concentration Sw = the intrinsic solubility in "pure water" Xmn = concentration of the surfactant as monomers Xmc = concentration of the surfactant in micellar form Kmn = partition constant between monomers and water Kmc = partition constant between micelles and water The separation of the concentration terms (Xmn and Xmc) accounts for differences in the partition efficiency of the solute with monomers and micelles. By equation (8), one would expect a plot of the apparent solute solubility (Sw*) versus the total concentration of surfactant to be bilinear; giving a straight line with slope of Kmn from X = 0 to X = CMC, followed by another straight line with slope of Kmc at X greater than or equal to CMC. Because of the markedly greater organic environment of micelles relative to monomers, the increase in slope of the plot on exceeding the CMC should be very sharp. The two distinct slopes define the values of Kmn and Kmc for a given solute-surfactant system. (ascii format does not support this graphic- see original report for graph) Figure 4. Hypothetical plot of surface tension as a function of surfactant concentration for three molecularly homogeneous surfactants (after Kile and Chiou, 1989). While CMC is assumed to be an observable and definite value in the case of surfactant monomers, there are frequent reports in the literature of the formation of "aggregates" or micelle-like associations in solutions of organic solutes so dilute as to apparently preclude the formation of micelles. Work with different types of commercial surfactants has indicated that molecularly nonhomogeneous surfactants do not display the sharp inflection in surface tension associated with CMC in molecularly homogeneous monomers, rather the onset of aggregation is broad and indistinct (KILE and CHIOU, 1989). The lack of well-defined CMC's for nonhomogeneous surfactants is speculated to result from the successive micellization of the heterogeneous monomers at different stoichiometric concentrations of the surfactant, which results in a breadth of the monomeric-micelle transition zone. (ascii format does not support this graphic- see original report for graph) Figure 5. Hypothetical plot of surface tension as a function of surfactant concentration for three molecularly heterogeneous surfactants (after Kile and Chiou, 1989). It is observed that molecularly nonhomogeneous surfactants are able to enhance the solubility of very hydrophobic chemicals such as DDT at surfactant concentrations well below the CMC. This is attributed to the successive micellization of the heterogeneous monomer species. Examination of the solubility enhancement with different types of commercial surfactants reveals that molecularly homogeneous surfactants show relatively insignificant (but linear) solubility enhancement below CMC. Molecularly nonhomogeneous surfactants, on the other hand, show a much greater solubility enhancement at concentrations below the CMC. The plot of the apparent solubility of DDT as a function of the ratio of surfactant concentration (X) to critical micellar concentration (CMC) shows a smooth, gradual upward curvature below the nominal CMC that becomes increasingly steeper near and beyond the CMC as the pseudophase grows. (ascii format does not support this graphic- see original report for graph) Figure 6. Hypothetical plot of apparent solubility enhancement of DDT by surfactants at concentrations approaching nominal CMC (after Kile and Chiou, 1989). These characteristics are not predicted by the conventional theory for homogeneous surfactants; the nonlinear relation near the nominal CMC is strongly indicative of a continuous aggregate formation. This effect may be attributed to a sequential micelliza tion of the heterogeneous monomers because of their unequal solubilities in water. As a result, the monomer-micelle transition of a heterogeneous surfactant may be expected to be considerably less sharp than that of an homogeneous surfactant. This reasoning is in accord with the relatively smooth solubility enhancement curves over the region of the nominal CMC for the surfactants with nonhomogeneous molecular composition. The conventional methods for CMC determinations are not sensitive to incipient formation of aggregates. In comparison, measurement of the solubility enhancement appears to be much more sensitive to surfactant aggregates. This can be partially rationalized by the fact that the change in surface tension (as related mainly to monomers) of a surfactant solution on exceeding the CMC is generally less than a factor of 3, whereas the associated change in the apparent solubility of DDT (which is strongly a function of the concentration of micelles) can be more than 2 orders of magnitude. For this reason, determination of the solubility enhancement data of DDT or other extremely water-insoluble compounds may prove the be the most sensitive method to date for detecting the association transition of the surfactant in water (KILE and CHIOU, 1989). The presence of water-soluble macromolecules in solution at submicellar concentrations has been reported to enhance the water solubility of hydrophobic organic chemicals in several instances (e.g., MADAN and CADWALLADER, 1970; ROHMER et al.,1972). Sorptive interactions or molecular aggregate formation in solution can alter the reactivity of solutes. DUYNSTEE and GRUNWALD (1965) studied the rate of alkaline hydrolysis of methyl-1-naphthoate in a solution of different salts and observed the effect of the additions on the rate of reaction to form methanol and the naphthoate ion. The results suggested that methyl-1-naphthoate formed complexes with the added organic species. If the complexes had a negative charge (organic anions), attack by the hydroxide was impeded; in those with a positive charge, attack was facilitated. The specific manner in which the neutral-salt effects varied with the structure of the organic species suggested that the dominant interactions leading to complex formation involve London or van der Waals force. This suggests that the organic-ion salt effects involve molecular interactions similar to those that cause micelle formation in aqueous solutions of detergent salts. This is stated in spite of the observation that there did not appear to be the actual formation of micelles per se . It appears that the actual formation of micelles created a phase in which the micellar components were unreactive to hydrolysis. The presence of macromolecules in solution can enhance the apparent solubility of solutes by sorptive interactions in the solution phase. The processes by which macromolecules enhance the solubility of contaminants are probably variable as a function of the particular physical and chemical properties of the system. A macromolecule possessing a substantial nonpolar region can sorb a hydrophobic molecule thereby minimizing the interfacial tension between the solute and the water. PRECIPITATION In a process somewhat akin to the formation of micelles, dissolved inorganic solution constituents may precipitate and form a solid phase. This solid phase may form as a bulk phase or exist as dispersed particles in the solution phase with a continual gradation in between the two extremes. At one end of this spectrum is the formation of neutral ion pairs which exist in true solution. The settling of a solid precipitate or the growth of crystals as a discrete solid phase may be considered as the other end of the spectrum. The process of precipitation occurs as a function of the intrinsic properties of the materials involved. The event of precipitation or dissolution may be considered to be governed by the free energy relation described above. The tendency of a material, AB, to dissolve or precipitate may be described in an equilibrium mass-action expression such as Adissolved + Bdissolved = ABsolid (9) The relative tendencies of the reaction to proceed forward or backward as written may be described by measuring the concentrations of all species at equilibrium. The ratio of the products to the reactants may be figured to yield an equilibrium constant according to the following form [{A} {B}] / {AB} = Kequilibrium (10) Where the brackets indicate activities of the reactants. Since the activity of the solid phase is taken to be unity (by defining it as the reference state), the denominator can be eliminated to yield the solubility product constant, Ksp (STUMM and MORGAN, 1981; p.231). {A} {B} = Ksp (11) This seemingly straightforward principle is somewhat complicated in natural systems by the existence of changing conditions, kinetic limitations and coincident or competing reactions. Changing conditions may include changes in oxidation state of the reactants (which may alter the reaction entirely) or changes in temperature. As previously stated, the equilibrium considerations undertaken here do not consider reaction rates or physical limitations which may affect precipitation events. Simply attaining reactant activities exceeding to the solubility product does not guarantee a precipitation event. When it does occur, precipitation can sometimes result in different allotropic modifications, ranging from amorphous to crystalline and can have variations within each form (e.g., WHITTEMORE and LANGMUIR, 1975; STUMM and MORGAN, 1981). Coprecipitation can occur when materials in solution get trapped or caught-up in a precipitation event. This can cause scavenging of solution constituents when a precipitate forms (e.g., JACKSON, 1975). Scavenging can occur on different scales. On the molecular level, dissolved species can become entrapped or bonded in the crystalline lattice if it forms. This may result in phenomena such as isomorphic substitution in clay minerals or simply the existence of "impurities" in the resulting solid. On the macroscopic level, scavenging can occur when dissolved or suspended solution components are taken out of solution by becoming entrapped in a precipitate. This sort of coprecipitation or scavenging event is intentionally created in the use of coagulants in water treatment operations where a slightly soluble salt is rapidly added to water in sufficient amounts to create a saturated solution. When solubility is exceeded, the precipitation event scavenges materials in solution which are responsible for undesirable turbidity. These are trapped in the amorphous matrix and settled out by gravity, thus removing them from solution. pH pH is a fundamental quantity. The pH of the solution can have an impact on the solubility of organic and inorganic solutes. The pH can have an effect on a reaction equilibrium if the reaction, or a related reaction, consumes or produces H+ or OH-. pH relates the activity of protons in solution, i.e., pH = -log {H+} (12) Hydroxide and carbonate typically form insoluble precipitates with polyvalent cations in natural waters. The activity of both of these species increases with pH. The presence of surface functional groups which are capable of exchanging a proton creates pH dependent charge, whereby the ionic character of the surface increases with pH. The molecular configuration of polyelectrolytes may be influenced by pH as the molecules coil and uncoil as the pH decreases or increases. In such a situation, charged sites such as acidic hydroxyl groups or amines can lose or acquire charge as a result of changes in solution pH. In a large, flexible, polyfunctional molecule, intramolecular self-association is thought to occur in the absence of electrostatic repulsion. The tendency to form such intramolecular bonds will vary as charged sites are created or satisfied by pH changes. In such a situation, decreases in pH will satisfy the charge on the surface of the molecule, thereby lowering the hydrophilicity of the surface and also decreasing the coulombic repulsion of the molecular chain for itself and pe rmitting intramolecular bonding (e.g., MATSUDA and SCHNITZER, 1971). REDOX Redox is another fundamental quantity. The redox status of a solution can have an impact on the solution behavior by affecting the oxidation states of the species in the system. Elements form different compounds as a function of differing oxidation states . These compounds frequently differ in their solubilities. Redox relates the activity of electrons in the system. This can be measured using a galvanic cell consisting of two electrodes connected by a conducting solution: oxidation occurs at the negative electrode (anode) and electrons are produced, whereas electrons are consumed and reduction takes place at the positive electrode (cathode). The redox potential is quantified by comparison with a standard redox couple. By convention, the standard redox couple is that present at a hydrogen electrode consisting of a platinum electrode, with hydrogen ions in solution. In the presence of platinum as the catalyst, the reaction is: H2 to 2H+ + 2e-, and the tendency to donate reducing equivalents, as electrons in this case, is measured as the voltage (potential) of the electrical current generated, when the electrode is coupled in series with another redox couple electrode. Under standard conditions, 25C, 1 atm of H2 and pH0, the redox couple H2/2H+ + 2e- is -420 mV. The symbol Eh is used for the redox potential under standard conditions and is measured in volts. Somewhat analogous to pH, pE gives the hypothetical electron activity at equilibrium. The two quantities are related to each other by pE = -log {e-} = Eh /2.3 RTF-1 (13) where R = the gas constant T = temperature (absolute) F = the Faraday constant pE is a measure of the relative tendency of the solution to accept or transfer electrons. In a highly reducing solution the tendency to donate electrons, that is, the hypothetical "electron pressure", or electron activity is large. Just as the activity of hydrogen ions is very low at high pH, the activity of electrons is very low at high pE. Thus a low pE (or a low Eh) indicates reducing conditions. Oxidation-reduction reactions occur by the transfer of electrons between the reactants. Different materials differ in their tendency to accept electrons in reactions, hence preferential reduction of easily reduced materials occurs. Under equilibrium conditions with an abundance of reducing agent, sequential reduction of electron acceptors takes place with those most easily reduced accepting electrons until the concentration of that species is depleted. Under such idealized conditions, the redox potential of the solution will be constant during this process. When the supply of this most-easily reduced reagent is exhausted, the pE will drop to the threshold of the next most easily reduced material and the process will continue. In real systems, kinetic considerations may limit the ability of the redox system to express its equilibrium status. The redox condition of the natural water system will have profound impact on the nature of the overall system and the solubilities of aqueous constituents. In well-aerated systems, oxygen usually serves as the terminal electron acceptor in most oxidation reactions. In the absence of sufficient oxygen, the pE drops until another electron acceptor becomes available. This is commonly NO3-, Mn+4 or Fe+3, although other materials may predominate depending on the geology or other conditions unique to the specif ic situation. Under very reducing conditions in natural waters, sulfide minerals are often formed from the reduction of sulfur containing compounds. Redox changes can also induce changes in the ionic strength of a solution by making relatively insoluble materials into more soluble ones. An example of this is the reduction of iron or manganese. Both of these compounds are more soluble in the reduced form. The association between solubility and pE is linked with the pH of the system. pE and pH affect one another as redox-driven processes produce or consume hydrogen or hydroxide ions. The presence of water itself sets limits to the extremes of pE which can o ccur in water by virtue of water's ability to be oxidized or reduced. Oxidation-reduction (or redox) reactions in biological reactions are normally defined in terms of loss and gain of hydrogens or electrons. Each oxidation is accompanied by a reduction. When both couples combine in a complete redox reaction, the net flow of the reaction can be determined by the relative tendency of each couple to donate or accept reducing equivalents, which is the redox potential. A couple of lower redox potential will always donate reducing equivalents to a couple of higher potential and, during the oxidation of a substrate, reducing equivalents are transferred in the direction of increasing potential. This transfer is accompanied by the release of free energy, the magnitude of which is given by the standard free energy change, Æ G = -nF delta Eh, where n is the number of electrons transferred in the reaction, F is the charge on one mole of electrons, which is Faraday's constant (96.649 kJ/V mol) and delta Eh is the standard electrode potential. Biological systems have evolved to conserve this energy and to convert it into biologically useful forms. In practical terms, the redox potential can be used to indicate which redox reactions will occur within a system. The redox potential gives a measure of the general condition of the liquid. Anaerobic processes will have low values of Eh (<-200 mV), where as aerobic processes will have higher values (>+50mV). More precisely, values of Eh -150mV to -420mV are found in anaerobic environments, whereas aerobic environments vary between -200mV and +420 mV. Facultative environments change from aerobic to anaerobic systems at about +100mV. (GRAY. 1989) In complex natural systems, there are numerous redox couples which are not necessarily in equilibrium with each other. As a consequence, it is not possible to define a unique pE to characterize the whole redox system. In addition, it is generally accepted that there are no free electrons in aqueous solution which could react with a redox electrode like H3O+ ions do with a pH electrode. Despite these facts, redox electrodes are frequently used to characterize the redox state by a single measurement presuming that the solution is in equilibrium. COLLOIDS The presence of colloids in natural aqueous systems will act to influence the distribution and behavior of contaminants. Colloids are formed by some of the physical and chemical processes described in the theoretical section above. These same physical and chemical forces govern the processes by which natural (and perturbed) systems distribute pollutants to and from the colloidal phase. The composition and behavior of colloids are complex and difficult to rigorously define. In the absence of an ability to effectively model colloidal systems in the natural world, we must rely on descriptions of the processes in a more conceptual sense. Colloids are particles in a solution that will not settle out. They are common in natural waters and can enhance the apparent solubility of a wide range of water pollutants, both organic and inorganic. Colloids may be considered as an extension of the solid and aqueous phases and are formed by conditions that can be quite variable in time and space; hence colloids can be dynamic and relatively ephemeral. The composition of colloids can vary with the composition of the solid and aqueous phases. Colloids can be made up of organic, inorganic, or a mixture of material. A colloidal solution has been defined as a solution intermediate in character between a suspension and a true solution. Particles with diameters less than 10 micrometer are usually called colloids (STUMM and MORGAN, 1981), although the distinction based on size is arbitrary. The size of particles is a continuum, and the point at which large macromolecules end and small colloids begin is subject to judgment as is the upper end of the size continuum, where colloids and suspended particles merge. The tendenc y of suspended particles to settle out of solution is not really a function of size alone, rather the relative density of the particles and the motion of the water will determine what is suspended and what settles. The very use of the term "colloid", which defines a behavior and an approximate size, begs the more rigorous definition of the chemical composition of the particles, yet this is a realistic descriptor in common usage. Colloids are present in surface and groundwaters. Surface systems receive terrestrial input as runoff carries soil-derived materials into the streams, rivers, lakes or estuaries. Groundwater receives leachate and percolation water and is frequently well-connected with surface water bodies. Colloids may also be formed in situ by native processes of precipitation and dissolution, suspension or biological activity. Colloids in solution represent a highly disperse solid phase. Because of the sorptive behavior of interfaces, the higher surface area of dispersed colloids tends to make colloids a more effective adsorbent on a mass basis than an equivalent mass of precipitated or solid material. Colloids act to enhance the solubility of slightly soluble contaminants, whether they be organic or inorganic. Hydrophobic organics, and slightly-soluble inorganics, including radionuclides, have been associated with colloids in apparent solution. Colloids as a Third Phase The importance of the colloidal phase in the distribution of water pollutants is a relatively recent issue in the environmental literature. The phenomenon of colloidal solubility enhancement was detected by workers in several fields and was largely unexplained. The concept was apparently developed and forwarded by working with partitioning behavior of water pollutants in water/sediment systems. It was observed that the amount of sediment used in batch isotherms influenced the percentage of pollutant which was sorbed onto the solid phase. Work cited by O'CONNOR and CONNOLLY (1980) found that equilibrium sorption partition coefficients of several radioactive materials into Texas river sediments declined as sediment concentration increased in isothermal studies. This has been interpreted as an indication that colloids in solution were competing with the sediment for sorbate and that the concentration of colloids increased as the concentration of sediment increased. The importance of colloids was recognized by VOICE et al. (1983) when they discussed what they called the "particle concentration effect"; a term coined to describe the observation that the partition coefficient for strongly sorbed or slightly soluble solutes varied with the concentration of the soil/sediment used in the experimental work. They proposed that the observed change in partitioning behavior due to solids concentration could be attributed to a transfer of sorbing, or solute binding, material from the solid phase to the liquid phase during the course of the partitioning experiment. This material, whether dissolved, macromolecular, or microparticulate in nature, was not removed from the liquid phase during the separation procedure and was capable of stabilizing the compound of interest in solution. The amount of material contributed to the liquid phase was thought to be most likely proportional to the amount of solid phase present, and thus the capacity of the liquid phase to accommodate solute would depend upon the concentration of solids in the system. The overall effect can be viewed either as a two-phase system; where the properties of one phase (liquid) vary with the mass of the other (solids), or as a three-phase system consisting of water, solids, and a third phase that is not separated from the water but possesses a higher capacity for the solute than the water itself. This is the colloidal phase. Measurements of "dissolved" sorbing phase (weight of dissolved solids, turbidity, and dissolved organic carbon) demonstrate the increased loading of nonsettling microparticles or macromolecules in the supernatants of batch equilibrium experiments as the solids-to-water ratio increases. It is clear that nonsettling microparticles or macromolecules vary regularly with suspended solid concentration. The observation that dissolved colloidal material was increasing the apparent solubility of analytes in laboratory studies led to the attempt to wash the sediment to try to remove these materials. Successive washings reduced the amount of material in solution, but failed to remove it. After five successive washes, the nonsettling microparticles or macromolecule content dropped about an order of magnitude, yet remained at an amazingly high level of 100 mg/l even after five washes (GSCHWEND and WU, 1985). WALTERS et al.,(1989) confirmed the report that aqueous colloids couldn't be removed satisfactorily by washing or centrifugation. That this occurs should not be particularly surprising. Particle size distributions of natural sediments and soils are undoubtedly continuous and do not drop to zero abundance in the region of typical centrifugation or filtration capabilities. Additionally, there is some evidence to indicate that dissolved and particulate organic carbon in natural waters are in dynamic equilibrium, causing new particles or new dissolved molecules to be formed when others are removed. Experiments with soil columns have shown that natural soils can release large quantities of dissolved organic carbon into percolating fluids (GSCHWEND and WU, 1985 and references therein). In work by BEASLEY and JENNINGS (1984), transport in the solution phase was indicated in the removal of radionuclides from the Columbia River. This river received apparently substantial inputs of radioactive materials as it was used to cool single-pass reactors engaged in weapons production. An inventory of the river was made to determine the radioactivity of the buried sediment. It was observed that the sediment contained levels of radionuclides which were much lower that what was expected from accepted partition models based on a two-phase system. It was suggested that the removal of these contaminants was accomplished in the absence of any major erosional events; hence the transport was assumed to have been accomplished in the solution phase. It has been observed (ALBERTS and WAHLGREN, 1977) that storm-caused turbidity in lake Michigan was associated with elevated levels of plutonium and americium in finished drinking water. The implication is that radionuclides that are strongly sorbed to the sediment are re-suspended into the water as the sediment is stirred by wave action. The apparent solubility of many radionuclides is enhanced through complexation with naturally occurring humic and fulvic acids (NELSON et al.,1985), or through electrostatic binding to colloid clay, metal oxides or other inorganic colloids. Colloids have been repeatedly shown to be important in enhancing the apparent solubility of hydrophobic organic chemicals (e.g., HASSETT and ANDERSON, 1979, 1982; MEANS and WIJAYARATNE, 1982; VINTEN et al.,1983; HASSETT and MILICIC, 1985; WHITEHOUSE, 1985 and others). The solid phase is the source of dissolved or suspended colloidal material that is acting as the third phase. It is observed that the solution phase is in dynamic equilibrium with the solid phase (soil/sediment). The discussion of colloid formation requires a description of the material from which they are formed. To this end, a brief discussion of the nature of the inorganic and organic components of the natural world will be undertaken to provide a background and reference to the discussion of the formation and behavior of colloids. CLAY FRACTION Much of the early work in characterizing the environmental behavior of chemicals was accomplished in the area of agricultural chemistry. Work surrounding the behavior of plant nutrients in the soil has provided a large base of information about the processes of environmental chemistry. Workers investigating the effectiveness of soil-applied herbicides determined that the herbicidal activity of organic chemicals varied with soil properties. It was determined the clay fraction and the organic matter content of the soil were related to the ability of a soil to diminish the effectiveness of an organic herbicide applied to the soil. The clay fraction, which has long been considered as a very important and chemically active component of the soil, has both textural and mineral definitions. In its textural definition, clay is generally assumed to be the mineral fraction of the soil that is smaller than about 0.002 mm in diameter. The small size of clay imparts a large surface area for a given mass of material. This large surface area of the clay textural fraction of the soil makes it very important in processes involving interfacial phenomena such as sorption or surface catalysis. In its mineral definition, clay is composed of secondary minerals such as carbonate and sulfur minerals, layer silicates and various oxides. Layer silicates are perhaps the most important component of the clay mineral fraction. Because of isomorphic substitution of ions in the crystalline lattice of layer silicates, many clay surfaces have a net negative charge which results in the ability of such minerals to exchange cations from the soil solution. The cation exchange capacity varies from about 3 to 200 meq/100g (e.g., THOMPSON and TROEH, 1973). In addition to the existence of a static charge on the clay surface resulting from intracrystalline charge imbalances (isomorphic substitution), soil minerals may acquire charge from the pH-influenced dissociation of surface hydroxyl groups. The magnitude of this type of cation exchange capacity will tend to increase with pH. The sphere of influence or extent of impact of the charged clay surface on the structure of ions of the solution will be, to some extent, determined by the ionic strength of the solution according to the double-layer effects discussed above. In addition to the presence of phyllosilicate minerals which exist in crystalline layers and frequently possess a net surface charge from isomorphic substitution, other products of mineral weathering and dissolution may be present in the clay fraction which do not exist in layers and do not possess an intrinsic charge. These are sometimes called accessory minerals and some of these minerals, such as allophane, may have pH dependent charge. These minerals may exist as uncharged oxides, hydroxides and hydroxyoxides of aluminum, iron and titanium. Finely divided grains of these accessory minerals coat the surface of other mineral grains in the soil. The results of mineral weathering form particles with a size continuum from ions to grains. Mineral dissolution and precipitation occur more or less continuously as a function of ambient conditions. Particles of the clay textural fraction may be suspended in solution as colloids as well as existing as part of the stationary solids. For example, iron in groundwater is often present both in solution and as suspended ferric oxyhydroxides. A significant percentage of the iron in many groundwaters (30-70% is not uncommon) is present as suspended ferric oxyhydroxides. Solubilities of the oxyhydroxides vary greatly depending on such factors as dissolved Fe (II), initial precipitation pH, the base used for hydrolysis, surface area of the precipitate, and time. For a given Eh and pH, waters in equilibrium with freshly precipitated amorphous material can contain approximately seven orders of magnitude more dissolved iron than those in equilibrium with coarsely crystalline goethite or hematite (WHITTEMORE and LANGMUIR, 1975). It is reasonable to assume that clay colloids would exhibit similar surface chemistry to that clay sorbed, bonded or precipitated in the stationary solid phase. Mineral colloids may be formed when precipitation or dissolution forms particles which are resistant to settling. These particles may be formed by any number of conditions whereby the solubility of a particular solute is exceeded (e.g., WHITTEMORE and LANGMUIR, 1975; GSCHWEND and REYNOLDS, 1987) or a stable solid is disrupted mechanically or chem ically (NIGHTINGALE and BIANCHI, 1977; BUDDEMEIER and HUNT, 1988) The composition of the mineral fraction of the soil, being extensively composed of oxygen and silicon bonded with various metals, will lend a relatively polar nature to the surface of most of the inorganic soil components. This polar/ionic nature will cre ate a natural affinity between soil and ionic or polar solutes. In addition to sorption of ionic and polar solutes onto clay in the solution and solid phases, the clay fraction has been shown to be important in the sorptive behavior of neutral hydrophobic organic compounds from the water (KHAN et al.,1979; HASSETT et al.,1981). The large surface area of the clay fraction offers a large sorptive interface upon which hydrophobic bonding may occur (MINGELGRIN and GERSTL, 1983; BANERJEE P. et al.,1985). For such nonpolar compounds, polar/ionic attraction is generally secondary to hydrophobic effects in sorption on most sediments and soils. ORGANIC MATTER PROPERTIES It appears that for very hydrophobic molecules, the organic carbon content of the sorbent is of greater importance than the mineral surface itself (KHAN et al.,1979). However, in the low carbon environments characteristic of the subsurface, mineral sorption may play an important role in affecting hydrophobic pollutant mobility (BANERJEE P. et al.,1985; McGINLEY et al.,1989). It has been reported that if the soil contains more than about 0.2% organic carbon, all of the sorption of hydrophobic organics appears to be due to the organic carbon. If the solid phase contains less than about 0.2% organic carbon, the sorption of hydrophobic chemical from the aqueous phase may be attributed to the clay fraction (BANERJEE P. et al.,1985). In work with sorption of aromatic hydrocarbons by sediments, it was reported by KARICKHOFF et al.,(1979) that when the individual partition coefficients for the sorption of the compounds (Kp) were divided by the organic carbon contents of the sediments (% OC), a unique constant, (Koc) was generated that was independent of sediment properties and dependent only upon the nature of the organic analytes. Koc = Kp/(%OC) (14) These authors reported a significant correlation between the Koc values obtained from the sorption of organic compounds on three local sediments and the partition constants (Kow) for the partitioning of the compounds between octanol and water. log Koc = 1.00 log Kow - 0.21 (15) There have been a number of empirical expressions of this sort developed relating the partitioning behavior of a chemical between water and organic carbon to the octanol-water partition coefficient for the chemical (e.g., MEANS et al.,1980; KARICKHOFF, 1 981,1989; SCHWARZENBACH and WESTALL, 1981). It has been noted that the tendency of a chemical to partition into the organic phase of the soil or sediment is inversely related to the water solubility of the chemical (CHIOU et al.,1979,1982). Hence, the ten dency of an organic chemical to be sorbed by soil or sediment organic matter will be a function of its hydrophobicity. The octanol-water partitioning behavior of a material has also been related to its intrinsic hydrophobicity. Organic solids with high melting-points are reported to behave anomalously in such considerations (BANERJEE et al.,1980; MACKAY et al.,1980). The organic fraction has been determined to be of considerable importance in the environmental behavior of pollutants. The importance of organic matter in the processes of pollutant partitioning warrants some brief description of this material as a preface and background for discussing some of the properties it displays which influence the apparent solubility of water contaminants. HUMUS Humus is considered to be the remains of living things that are no longer visually recognizable as to their origins. The physical nature of humus is that of an amorphous, brownish material with a density somewhat lower than that of mineral soil. In its natural state, humic material is somewhat variable in composition and form. The processes of formation of organic matter are more or less unique to a particular geographic environment on the large scale. Since the material is derived from plant remains which have been more or less degraded by detritovores, it is reasonable to assume that there would be some differences between humic material from different biomes. The vegetation, soil minerals, climate, and microbial population are some of the variables which might act to create differences in the organic matter of a different area. In spite of reported differences, the variations are not so great as to preclude comparisons between humus from one biome to another. The process of plant growth is essentially one of photosynthesis; i.e., the sunlight-driven reduction of oxidized carbon. The reduced carbon takes many forms and combines with many elements in a very complex array of chemicals. When this living material dies, the chemical energy it contains is exploited by heterotrophic organisms for their own life processes. Humus is considered to be produced from that portion of the reduced carbon which was resistant to degradation either as a function of intrinsic nature (e.g., polyphenols such as lignins and tannins) or of ambient conditions which restrict the oxidative processes of degradation (e.g., cold, anaerobic environments). In addition to the reported differences between and within bioregions, it is reported that humic material of terrestrial origin is different from the humic material in freshwater streams. The humic material in aqueous sediments is reported to be less polar (as judged by a lower oxygen percentage) than the soil organic matter (LEE et al.,1981). Humus undergoes changes as it ages. The humus which exists in the soil is the result of extensive alteration of the original component materials and is subject to degradation. Under different conditions, humus undergoes diagenesis and transformation in response to the ambient conditions. Humus buried deep in the subsurface is subject to different forces and will be accordingly different after the passage of time. Coal, peat and oil are examples of organic matter that has undergone extensive diagenesis. Sometimes organic material undergoes diagenesis with earth minerals to form mineral-organic composites such as oil-shale. Diagenesis is reported to increase with depth and time of burial. Maturation is the result of mild heat and pressure; it is possible that interactions with mineral surfaces and complexed metals are also involved (JACKSON, 1975). Thermodynamic stabilization occurs in diagenesis. The least stable and most reactive components or their substituents are gradually eliminated. This process leads, with increasing age and depth of burial, to a gradual stabilization, not necessarily of each individual compound but the sedimentary organic matter as a whole. In terms of structures the transformation of open chains to saturated rings and finally to aromatic networks is favored; hydrogen becomes available for inter or intramolecular reduction processes. Eventually, highly ordered, stable structures of graphite may be formed. It is pointed out the most characteristic feature of organic diagenesis is the appearance of extreme structural complexity and disorder at an intermediate stage, interposed between the high degree of biochemical order of the starting material, and the even greater crystallographic order of graphite, the end product of diagenesis (STUMM and MORGAN, 1981). The long-recognized complexity of organic matter has generally confounded accurate and detailed description of the material and has instead spawned qualitative divisions of the natural material which have been adopted by workers in the field to allow for some agreement on methodology at least. The nature of humus has been studied exhaustively and, in spite of some conflicting reports, a number of points have been agreed on; some of these will be related as they apply to the material at hand. The humic substances are organic polyelectrolytes which are most commonly identified with the organic material present in soils. However, humic substances are also present in practically all of the suspended and bottom sediments of rivers, lakes and estua ries. Humic materials are also apparently soluble in water and are present in surface and groundwaters. This group of compounds enters into a wide variety of physical and chemical interactions, including sorption, ion exchange, free radical reactions and solubilization. The water holding capacity and buffering capacity of soils and the availability of nutrients to plants are controlled to a large extent by the amount of humus in the soil. Humus also interacts with soil minerals to aid in the weathering and decomposition of silicate and aluminosilicate minerals. It is also adsorbed by some soil minerals. The chemical nature of humus is the subject of variable and sometimes conflicting reports in the literature. ALEXANDER (1977) sums up the issue by stating that "humus should be considered as a portion of the soil that is composed of a heterogeneous group of substances, most having an unknown parentage and an unknown chemical structure". For the purposes of this technical appendix, it is important to point out that humus is chemically reactive and has variable chemistry, manifesting both polar and nonpolar tendencies. In general, humus contains a number of chemical functional groups assoc iated with a polycyclic aromatic matrix of varying size. In term of types of compounds, humus contains a number of polymerized substances; aromatic molecules, polysaccharides of several kinds, bound amino acids, polymers of uronic acids and phosphorous containing compounds. Chemical degradation has shown that the basic building blocks of humic acids are benzene carboxylic acid groups, substituted phenolic groups and quinone groups. KHAN and SCHNITZER (1972) proposed that humic material exists in a molecular sieve or clathrate (lattice-like) structure joined by polar bonding mechanisms. They proposed this because they observed that the material would sorb hydrophobic organic compounds and not release them to extraction with organic solvents. Methylation of the mixture enabled extraction of the sorbed analyte. They supposed that the methylation step disrupted hydrogen bonding within the humic material. It is worthwhile to discuss several of the more likely possibilities of mechanism of aggregation of humic materials. The mechanism or mechanisms that bring about the aggregation or disaggregation of humic substances will be determined by charge distributi on and functional-group distribution on the exposed surfaces. A number of workers have shown that humic materials contain abundant polar functional groups. The highly polar nature of some of the functional groups makes dipole bonds and hydrogen bonds probable active mechanisms of structural change (e.g., WERSHAW and PINKNEY, 1973). Reports that humic materials contain both electron rich and electron deficient sites gives evidence that polar bonding will be likely to occur (MELCER et al.,1989). In addition to the hydrogen bonding, coulombic attraction of charged particles will also create bonds in humus. The charged sites on a polyelectrolyte molecule may arise in several different ways. Ionic compounds will dissociate in solution, producing molecules with charged sites. These charged sites may also result from chargeÑtransfer reactions such as the transfer of an electron from a carbanion or a radical anion to another molecule (WERSHAW and PINKNEY, 1973). Humus is also capable of forming covalent bonds with aqueous solutes (PARRIS, 1980). Humus is the site of considerable microbial activity. Living and dead organisms and extracellular enzymes are typically associated with humus as part of the material (e.g ., ALEXANDER, 1977). The presence of enzymes can catalyze reactions. It is also reported that humus contains stable free radicals which make it very reactive and able to form covalent bonds or create ions. It is reasonable to assume that charge-transfer reactions between free radicals are important in the aggregation of humic materials in light of the high concentrations of free radicals that have been detected in both soils and aqueous humic preparations. The free radicals detected in soils and humic acids may arise from the reduction of a diamagnetic molecule by a solvated electron, enzymatic reactions or photolysis (WERSHAW and PINKNEY, 1973). The diverse nature of chemical bonding arrangements exhibited by humus enables the formation of associations both with nonhumic materials and with other humic materials to create a dynamic structure in what WERSHAW and PINKNEY (1973) term a "living polymer". Such a system is capable of undergoing inter and intra molecular bonding to add or lose constituents or change configuration in response to ambient conditions. The chemically diverse and highly reactive nature of the humic matrix imparts the ability of humus to both lose and acquire molecular moieties in a dynamic manner. A fractionation procedure has been derived and widely applied to studies of humic material. The procedure begins with natural organic matter (humus) and uses a basic solution to solubilize a fraction of the material. The basic extract is then acidified which causes a precipitate to form. The precipitate is named humic acid. The fraction which remains in solution is called fulvic acid. Humin is the name given to the insoluble fraction that remains after extraction of humic and fulvic acids. At near-neutral pH, which is characteristic of most natural water, the fulvic acid is the most water soluble of these three fractions. Humic acid is somewhat less soluble, with its solubility increasing as the pH increases. Humin is normally considered in soluble at all pH values. Fulvic acids are soluble in water and so are the majority of the salts of these acids (e.g., OGNER and SCHNITZER, 1970b). The aquatic fulvic acid fraction contains substances with molecular weights ranging from 500 to 2000 and is monodisperse. Aquatic fulvic acids are dissolved rather than colloidal (THURMAN et al.,1982). Fulvic acid has been found to contain branched, cyclic and linear alkanes, as well as fatty acids (OGNER and SCHNITZER, 1970b). It can combine with insoluble organic compounds such as alkanes, fatty acids and dialkyl phthalates to form stable "complexes" that are soluble in water (OGNER and SCHNITZER, 1970a; MATSUDA and SCHNITZER, 1971, BOEHM and QUINN, 1973). Humic acid is pictured as being made up of a hierarchy of structural elements. At the lowest level in this hierarchy are simple phenolic, quinoid and benzene carboxylic acid groups. These groups are bonded covalently into small particles. The molecules of humic acid are reported to be nonspherical, or more probably, nonspherical and hydrated (PIRET et al.,1960). Other workers have reported that humic materials are rigid spherocolloids in solution (KONONOVA, 1961; and references contained therein). Work by GHOSH and SCHNITZER (1980) makes the overall conclusion that the configurations of humic and fulvic acid molecules are not unique; they vary with changes in the environment. These authors report that both humic and fulvic acid molecules are flexible linear colloids at low concentrations, provided hydrogen ion and neutral salt concentrations are not too high. As these factors increase, the macromolecules assume coiled configurations similar to those of uncharged polymers or rigid spherocolloids. It has been reported by KONONOVA (1961) (and references contained therein) that humic acids may have an amorphous structure and furthermore that the size and weight of the humic acid molecules may vary as a function of ambient solution conditions. It has been postulated that the molecular weight of the humic species may vary from 1000 to 50,000 (PIRET et al.,1960). They may consist of particles capable of aggregation or dissociation. Humic acids are larger than fulvic acids and form polydisperse systems (THURMAN et al.,1982). Precipitation is used to isolate humic acid from soil. The humic acid must aggregate to precipitate, therefore, it gives polydisperse systems showing that it exists as aggregates of various sizes. WERSHAW et al.,(1977) contend that humic acids are a mixture of a limited number of more or less chemically distinct fractions of relatively low molecular weight that form molecular aggregates in solution. Particles of similar chemical structure are thought to be linked together by weak bonds to form "homogeneous" aggregates. Two or more different types of aggregates may be linked together to form mixed aggregates. While there is no general agreement on the matter, some workers in the field contend that humic acid is really simply an aggregate of common humic materials and so is chemically similar to fulvic acid and to humin (WERSHAW et al.,1977). The distinctions of humic and fulvic acids and humin are purely operational definitions. WERSHAW (1990) termed the distinction between humic and fulvic acids and humin "a totally artificial division that tends to obscure the close interaction between the organic constituents of natural water systems" Humus can exist in solution as well as in the solid phase. The behavior of water-soluble humic materials is of great relevance to the discussion of solubility enhancement of aqueous pollutants, both organic and inorganic. Freshwater aquatic humic substances originate from soil humic material and terrestrial and aquatic plants. In surface waters these compounds generally account for 30 to 50% of the dissolved organic matter (THURMAN et al.,1982). Systems that contain naturally high levels of dissolved organic matter include bogs, swamps, and interstitial waters of sediments. Interstitial water (porewater) is formed by the entrapment of water during sedimentation, which isolates it from the overlying water. Porewater is considered to be in equilibrium with the sedimentary solid phase and separate from the overlying water column, or bulk water. In high carbon sediments, dissolved organic carbon in porewater can exceed 100 mg/L, whereas overlying waters typically contain less than 5 mg/L of dissolved organic carbon (THURMAN and MALCOLM, 1981; CARON and SUFFET, 1989). The molecular weight of most of humic substances in water is less than 10,000. Although some studies have found humic substances with molecular weights greater than 100,000 (LEE et al.,1981). The ability of humus in solution to form extensive aggregates w as discussed above. INTERACTIONS BETWEEN ORGANIC MATTER AND ORGANIC POLLUTANTS It is suggested that binding to dissolved humic materials could significantly affect the environmental behavior of hydrophobic organic compounds. The rate of chemical degradation, photolysis, volatilization, transfer to sediments, and biological uptake may be different for the fraction of pollutant that is bound to dissolved humic materials (CARTER and SUFFET, 1982). If this is the case, the distribution and total mass of a pollutant in an ecosystem would depend, in part, on the extent of humic material-hydrophobic binding. Agricultural chemists found that organic herbicide activity was inversely related to soil organic matter content (e.g., UPCHURCH and MASON, 1962; LAMBERT et al.,1975). The volatility of organic pesticides was found to be diminished in the presence organic colloids isolated from soil organic matter (PORTER and BEARD, 1968). The relationship between the organic matter content and the behavior of organic chemicals in soils and sediments has been documented extensively (e.g., KHAN and SCHNITZER, 1972; KARICK HOFF et al.,1979; MEANS et al.,1980; BROWN and FLAGG, 1981). It is commonly reported that solutions of soil or sediment-derived organic matter increase the solubility of hydrophobic organic chemicals (e.g., WERSHAW et al.,1969; BALLARD, 1971; HASSETT and ANDERSON, 1979; OGNER and SCHNITZER, 1970a; MATSUDA and SCHNITZER, 1971). The presence of dissolved organic matter in sorption studies will be manifest as a decreased soil/sediment partition coefficient at equilibrium, because the humic material in solution competes with the stationary solid phase for sorption of the analyte. Other workers have reported similar observations of enhanced solubility by organic carbon in the bulk aqueous phase (e.g., BOEHM and QUINN, 1973; CARTER and SUFFET, 1982; LANDRUM et al.,1984; HAAS and KAPLAN, 1985; GSCHWEND and WU, 1985; and others). Pollutants may be bound to humic materials through abiotic or biological processes whereby the formation of bound residues usually results in detoxification of the pollutant. Therefore, enhancing the binding of xenobiotic chemicals to humic materials can serve as a means to reduce toxicity as well as migration of the toxic compounds. Complex formation can occur by an oxidative coupling reaction leading to oligomeric and polymeric products. BOLLAG and BOLLAG, (1990) report the effect of phenoloxidazes (peroxidases, tyrosinases, and lacases) on the binding of substituted phenols and aromatic amines to humus monomers as well as to humic substances. Copolymerization largely depends on the chemical reactivity of the substrates involved. Certain phenolic humus constituents, such as guaiacol or ferulic acid are highly reactive in the presence of phenoloxidases. When one of these compounds was incubated together with a phenoloxidase with less or even non-reactive phenols, anilines or other chemicals, a synergistic reaction took place, resulting in increased formation of bound residues of these compounds. Phenyloxidases are able to catalyze the polymerization and/or binding of numerous organics to humic constituents. The inclusion of phenolic humus constituents, such as syringic acid, vanillic acid or vanillin, resulted in the enzyme induced formation of various cross coupling products. A wide variety of xenobiotics can become cross coupled to naturally occurring humic monomers by the action of phenyloxidases. These xenobiotics include phenols such as various mono-, di- and tri- substituted chlorophenols and 2,6-xylenol, and anilines such as 4-chloroaniline, 3,4-dichloroaniline and 2,6-diethylaniline. This enhanced removal of a xenobiotic is by no means unique to fulvic acid. It has been shown that the addition of a highly reactive humic monomer, such as syringic acid, to a phenoloxidase containing system can initiate the effective polymerization and/or binding of a molecule which by itself is only poorly transformed, if at all. There seems to be no shortage of examples of enzyme induced polymerization and binding of xenobiotics. It is thought that the enzyme induced oxidation of naturally occurring phenols yields free radical quinonoid structures. This is a common pathway in the phenoloxidase catalyzed polymerization and binding of both naturally occurring and man made compounds. Another pathway is the decarboxylation of a highly reactive compound such as sytingic acid and the formation of a covalent bond at that site to generate phenolic oligomers. Binding of a pollutant to humic acids, clays, or other materials would be expected to decrease its toxic effects. Binding can reduce the amount of a compound available to the biota and as the quantity of an available xenobiotic is reduced, toxicity also declines. The distribution of hydrophobic organic pollutants between sediments and water has typically been viewed as a surface adsorption phenomenon and, as such, has been studied with batch sorption isotherm techniques. Adsorption isotherms of nonpolar organic co mpounds on a number of soils and sediments are linear over a wide range of equilibrium solute concentrations (e.g., CARON and SUFFET, 1989). If the mechanism of solubility enhancement of hydrophobic organics is one of surface sorption, it might be expected that partition coefficients of aquatic humic substances may be less than those of the organic matter on particles, since macromolecules in solution must be relatively hydrophilic (GSCHWEND and WU, 1985). This view is supported by the reports describing heteroatom compositional differences between fulvic and humic acids recovered from natural water. The smaller, more water soluble fulvic acids have higher oxygen-to-carbon ratios compared to the larger humic compounds (THURMAN and MALCOLM, 1981). Thus, smaller, more water soluble macromolecules may be expected to be more polar sorbents (i.e., exhibit relatively lower Koc's) than related larger macromolecules and particulate matter. It has been proposed that the association between nonpolar compounds and the organic carbon fraction of sediments, soils, and natural waters is better described as a liquid-liquid partitioning phenomenon than as a surface adsorption process (CHIOU et al., 1979, 1983; KILE and CHIOU, 1989a). An organic matter partitioning process is supported by a number of observations, including; - linear sorption isotherms to near aqueous saturation concentrations of nonpolar organic substances, with no evidence of isotherm curvature at the higher concentration range; -isotherm curvature at higher concentrations is predicted by adsorption theories ; - small temperature effects on solute sorption; - absence of competition in experiments using binary solute systems; - data covering seven orders of magnitude in which sediment-water partition coefficients were inversely proportional to aqueous solubility and well correlated to octanol-water partition coefficients. CHIOU et al.,(1987) considered the mechanism for water solubility enhancement of nonionic organic solutes by dissolved organic matter of soil and aquatic origins. Such enhancement effects were effectively explained in terms of a partitionlike interaction of solutes with dissolved high molecular weight humic materials on the basis of the properties of the solutes and humic materials. The observed solubility enhancement of the solute by dissolved organic matter (DOM) can be expressed by Sw* = Sw(1 + X KDOM) (16) where Sw* = apparent water solubility in the solution Sw = apparent water solubility in pure water X = concentration of dissolved organic matter KDOM = partition coefficient between DOM and water The difference in values of KDOM for a solute with different types of fractionated humic materials has been explained in terms of the polarity, molecular size, and molecular configuration of the humic materials (based on elemental data analysis) gives a reasonable estimate of the relative enhancing effects among humic extracts. Conceivably, the compositions and structures of humic materials in different aquatic systems can be significantly different because of such environmental factors as water pH, biological processes, and the presence of other chemical species that affect the concentration (solubility) of humic materials. In more acidic streams or rivers, there appears to be a tendency for the humic material to contain a larger percentage of oxygen compared to samples from a neutral or basic water. A decrease in oxygen content of the humic materials from acidic to neutral water can also be accompanied by an increase in carbon content. The solubility enhancement effects of individual humic samples appear to be closely correlated with the polarity of the materials (using elemental data as approximate indices), suggesting that differences in molecular sizes of humic materials are not as much a critical factor as the polarity in affecting the partition interaction with organic solutes (CHIOU et al.,1987). It is supposed that the solubility enhancement cannot be explained by the cosolvency theory forwarded by Yalkowsky and others (e.g., YALKOWSKY et al.,1972,1975; AMIDON et al.,1974) because the magnitude of solubility enhancement is greater than that which would be predicted from cosolvent effects alone. This was investigated by CHIOU et al.,(1986) who used phenylacetic acid, synthetic organic polymers (poly(acrylic acid)) and dissolved humic and fulvic acids to assess the solubility enhancement effects on different chemicals. They found significant solubility enhancements of relatively water-insoluble solutes by dissolved organic matter of soil and aquatic origins. The concentrations of the humic materials varied from 0 to 94 ppm. They observed that the apparent solute solubilities increased linearly with dissolved organic matter concentration and showed no competitive effect between solutes. With a given dissolved organic matter sample, the solute partition coefficient increased with a decrease of the solute's water solubility or with an increase of the solute's octanol-water partition coefficient. The partition coefficient values of solutes with soil-derived humic acid were approximately 4 times greater than with soil fulvic acid and 5-7 times greater than with aquatic humic and fulvic acids. The effectiveness of dissolved organic matter in enhancing solute solubility appeared to be largely controlled by the molecular size and polarity of the material. The organic acid and polymer (molecular weight varied in separate experiments from 2000 to 90,000) created no observable solubility enhancement. The investigation of phenylacetic acid as a cosolute, with the concentration exceeding 600 mg/l, shows slight enhancement for DDT, which was the most hydrophobic analyte in the experiment. The magnitude of DDT solubility enhancement per unit mass of phenylacetic acid was much smaller than with the humic or fulvic acids. They found that the solubility enhancement exhibited by the dissolved humic material may be described in terms of a partition-like interaction of the solutes with a "microscopic nonpolar organic environment" associated with the high-molecular weight humic species. The relative inability of high-molecular-weight poly(acrylic acids) to enhance solute solubility was attributed to their high polarities and extended chain structures that do not permit the formation of a sizable intramolecular nonpolar environment. This observed "partition-like" interaction between hydrophobic organic solutes and dissolved humic material has led to the proposition that humic micelles may exist in solution. WERSHAW (1986) has proposed an elaborate model for this micelle structure. In this model, humic materials are pictured as existing as membrane-like aggregates which are made up of partially decomposed plant-derived compounds, which are held together in the aggregates by weak bonding mechanisms, such as pi bonding, hydrogen bonding and hydrophobic interactions. The humic membrane-like structure consists of polar hydrophilic exterior surfaces with hydrophobic interiors. Polar compounds will interact with the exterior polar groups of the humic structures, while hydrophobic compounds will partition into the hydrophobic interiors of the structures. This model is consistent with much of the reported information in the literature, especially with regard to the "membrane-like" behavior. Such a structure might explain the report of HASSETT and ANDERSON (1979), wherein it was reported that the solubility of cholesterol was enhanced by high-molecular-weight dissolved organic matter in river water. These authors found that solvent extraction of the radiolabeled cholesterol was ineffective as a means of recovery unless the organic matter was destroyed by U V radiation. The existence of a polar interface like the outside of a "membrane-like" structure, as proposed by Wershaw (1986), would explain the inability of a nonpolar solvent to recover the cholesterol. It might be envisioned that such a polar or ionic region might be a zone of high interfacial tension which would be relatively inhospitable to transversal by nonpolar molecules. Similar results were reported by FISH et al.(1989), who observed that solvent recovery of sorbed hydrophobic organics was enhanced by a digestion technique which degraded the dissolved organic matter. Simple adsorption onto a the nonpolar region of humic molecules by van der Waals forces and hydrophobic interfacial tension would probably not impede solvent recovery of adsorbed hydrophobic organics. Acceptance of the micelle model for dissolved humic material has been slow to arrive (e.g., MINGELGRIN and GERSTL, 1983; MaCINTYRE and SMITH, 1984; MACKAY and POWERS, 1987; and others). PIRET, et al.,(1960) reported that peat-derived humic acid had a critical micelle concentration of about 18 g/L. Since most of the observed solubility enhancement has been associated with humic concentrations orders of magnitude lower, it has been assumed that micelles were not present. More recent work with micelle formation has indicated that molecular aggregate formation may occur below the critical micelle concentration with molecularly nonhomogeneous surfactants. Natural humic materials may be considered as molecularly heterogeneous amphiphiles. It has also been shown (KILE et al.,1990) that a commercial surfactant which consists of a diverse admixture of monomers (made by reacting petroleum with concentrated sulfuric acid) does not exhibit behavior typically associated with micelle formation i.e ., a sharp inflection of solvent properties as the concentration of surfactant reaches CMC. These surfactants exhibit gradual change in solvent behavior with added surfactant. It is proposed that this gradual solubility enhancement indicates that micelle formation is a gradual process instead of a single event i.e., CMC does not exist as a unique point, rather it is a continuous function of molecular properties. This type of surfactant may construed as more realistic representation of humic material in water and may indicate that dissolved humic substances form molecular aggregates or colloids in solution; which comprise a third phase in aqueous environment. This phase will have the effect of increasing the apparent solubility of very hydrophobic chemicals. Polar solutes will not be noticeably affected by these colloids because they are already pretty soluble. These conclusions are consistent with the observations of BOEHM and QUINN (1973) who observed that the solubility of hydrophobic hydrocarbons was increased by humic material dissolved in sea water. There was evidence to suggest that the mode of solubilization of the hydrocarbons was by incorporation into micelles formed by intermolecular association of the surface active humic-type monomers. The solubilized hydrocarbons were determined to exist in a semicolloidal or micellar state formed by interaction of humic-like monomers in solution. The presence of salts in solution appeared to be prerequisite for the formation of the aggregates responsible for the solubility enhancement. The authors also observed that addition of the hydrocarbon being solubilized appeared to lower the concentration at which humic micelles formed. HUMIC/MINERAL ASSOCIATIONS In addition to the ability of humic substances to form associations with hydrophobic organic species, humic material also reacts readily to form associations with inorganic minerals and polar and ionic organic materials as well. These sorts of association s are involved in colloid formation with a wide variety of materials. It is reported that thorium(IV) was strongly bound to colloidal humic materials via metal-polyelectrolyte binding. This was determined to be electrostatic in nature, with complex stability increased with increasing ionization of the colloidal organic matter (NASH and CHOPPIN, 1980). The dissociation of thorium bound to humate in aqueous solution has further been studied, (CHOPPIN and NASH, 1981) and it is suggested that Th(IV) is bound by at least four types of sites with different basicities and different local polymer structure. CACHERIS and CHOPPIN (1981) reported that thorium(IV) was bound to humic material in solution and reported that the complex appeared to be characterized by two mechanisms with different dissociation tendencies. Fulvic acids can interact with clay minerals (OGNER and SCHNITZER, 1970a) and are known to form stable complexes with metal ions and hydrous oxides (e.g., JACKSON, 1975). The operational technique of isolation of humic acid involves a pH-induced precipitation and it is likely that accessory minerals may be associated with the precipitation process. Complexes of humic acid and clay minerals are also formed, the increased ash content of humic acid suggests that amorphous silica and clay may aggregate with the humic acid fraction (WERSHAW et al.,1977). Humin is that part of humus which is not solubilized by alkaline solution. It is thought to comprise an amorphous aromatic matrix interlinked by strong bonds. Humin is the least polar of the commonly studied fractions. Because of its extensive nonpolar ar omatic network, humin is probably the best sink for hydrophobic organic chemicals (MANAHAN, 1989). Extraction of the humin fraction of soils and sediments with methylisobutyl ketone (MIBK) demonstrates that the humin fraction also appears to be composed of several different components which can be separated by relatively gentle techniques. The humin can be fractionated into a hydrophobic fraction and a hydrophilic fraction. The hydrophobic fraction is white in color and it appears to be a lipid-like material, being at least partially soluble in solvents such as hexane, chloroform and benzene-methanol mixtures. This lipid-like fraction may be a mixture of plant lipids. The hydrophilic fraction is composed of two subfractions: an inorganic subfraction which settles out with time and a brown organic subfraction which remains in solution (WERSHAW, 1986). Alternate treatment of the humin fraction with strong mineral acids and strong bases generally renders most of the humin soluble in basic solutions. It was concluded by KONONOVA (1961), that humins are humic acids that are bound to the mineral constituent of soils. The amphiphilic nature of dissolved humic substances lends them the ability to associate with both hydrophobic organics and polar or ionic species (e.g., WERSHAW et al.,1977). Inorganic ions or mineral colloids in solution will interact with the electrically active surface of humic material in solution or in the solid phase according to the same bonding forces which lead to the association between soil organic matter and the soil mineral matrix. Humic matter in water is associated with various metal ions, clays and amorphous oxides of iron and aluminum (e.g., DAVIS, 1982). In aqueous environments, oxide mineral surfaces are generally covered with hydroxyl groups. Organic macromolecules can sorb onto these surfaces both by ligand exchange and by van der Waals forces to create a very strong association. In his model of "membrane-like" humic micelles, WERSHAW (1986) points out that metal ions cause aggregation of humic materials and suggests that inorganic ions may be associated with such structures. There are frequent reports of polyvalent cations causing humic material to aggregate (e.g., LEE et al.,1981; ARES and ZIECHMAN, 1988).The associations between humus and mineral matter is manifest as residue upon ashing of humic materials. Humic materials may be bound to the clay surfaces by the amino acids of proteins. This binding is most likely due to electrostatic charge effects. Metal ions could also act as bridges between the humic materials and the clay mineral surfaces. Oxides with relatively acidic surface hydroxyls, e.g. silica, do not react strongly with the organic matter (DAVIS, 1982) via coulombic attraction, however van der Waals forces will certainly exist between the large humic species and the silica surface. Humic materials are polyfunctional macromolecules with a number of surface groups capable of dissociating as ions. These groups will be more or less dissociated to impart some measure of ionic character to the material. This will tend to attract polar species such as water as well as counterions (e.g., JACKSON, 1975; LYTLE and PERDUE, 1981; BARKER et al.,1986). Humus is frequently considered to be able to form stable complexes such as chelates with polyvalent cations. Soil organic matter is capable of strong polydentate binding to transition metals in a chelate (e.g., DALANG et al.,1984; BOHN et al.,1985). The speciation of trace metals in natural waters is controlled by the interaction of the metals with a complex and varying mixture of inorganic anions, organic ligands, reducible or oxidizable dissolved chemical species, reactive surfaces, and organisms. Filterable concentrations of metals may include fine colloidal particles as well as organic and inorganic metal complexes (HERING and MOREL, 1990). In many natural waters high concentrations of colloidal organic material are frequently associated with high concentrations of iron (KNOX and JONES, 1979). Humus has been reported to act as a reducing agent. VISSER (1964) measured the formal redox potentials of neutral humic acid derived from tropical sphagnum peat and reported that the mean value of normal potentials was between +0.32 and +0.38 volts and decreased with increasing depth. More recent measurements in different soil reports that humic material is an active redox system with an E value of + 0.70 V (MANAHAN, 1989). Sometimes humus will reduce a metal and then bind it in a polydentate complex. For example, Cr2O7-2 is reduced by humic acids to chelatable cationic Cr3+ (MANAHAN, 1989). DOUGLAS and QUINN (1989) showed a very stable complex with chromium(III) and humic materials in reduced sediment. The accumulation of vanadium and molybdenum in peats has been attributed to reduction and chelation of soluble oxides of these elements. The reduction of iron(III) to iron(II) and subsequent retention in the reduced form has been demonstrated on the oxidized surface of coal (MANAHAN, 1989). In summary, the chemical and structural nature of humus makes it very active in the environmental behavior of many types of pollutants. The presence of bound enzymes and free radicals in the material allows it to form covalent bonds with a variety of molecules. The existence of nonpolar regions of the humus introduces the possibility of intramolecular sorptive partitioning of nonpolar organic materials into the humic matrix. The extent and polarizability of the humic surface enable it to bind to materials by the van der Waals force. The existence of electrostatic charges on the surface of the substance make it reactive with respect to water, ions and mineral surfaces. The nature of the surface chemistry grants humus a surface charge which is pH dependent, hence the tendency to coil or uncoil, to flocculate or disperse, will be more or less a function of pH and ionic character of the solution. COLLOID STABILITY It is commonly reported that dissolved humic material tends to coat mineral particles and thereby affect the surface chemistry of those materials. Dissolved organic matter coats the surfaces of soil particles even when it is present at very low concentrat ions (DAVIS, 1982). It furthermore imparts a negative charge to the surfaces which it coats. The organic coating is expected to have a great significance on subsequent adsorption of inorganic cations and anions (DALANG et. al.,1984). The importance of adsorbed organic material on trace metal uptake will be considerable because of the cation exchange capacity of the organic matter. Anion adsorption will also be greatly influenced by surfaces coated with organic material. This may be due to competition for the adsorption sites or possibly by electrostatic repulsion (DAVIS, 1982). In addition to the ability of organic matter to coat mineral particles and thus enhance the cation exchange capacity of the soil minerals, a thin organic coating may tend to increase the disperse nature of small mineral particles by imparting a net negative charge and creating a repulsion between the particles (GIBBS, 1983). The pH dependent nature of the charge on such coated particles may create a pH-dependent dispersion tendency; as the pH drops and the surface functional groups of the organic matter become electrically neutral, the particles coated with this organic matter would become less mutually repulsive and intraparticle collisions might result in the formation of van der Waals bonds. Such an event might result in flocculation of the particles. The intraparticle repulsion of such coated minerals will also diminish as the ionic strength of the solution increases. Experimental evidence has verified this (GIBBS, 1983). This is in accord with the model of double layer compression at higher ionic strength, which allows closer approach between particles. Organic coated particles coagulate much slower than the particles with the coatings removed. They will also resist sorption onto the stationary phase (as in saturated groundwater flow) if the stationary phase is also coated with negatively-charged organic matter (GSCHWEND and REYNOLDS, 1987). Such behavior is important in environmental management. Solid surfaces in natural aqueous systems are the sites of important geochemical phenomena. Coagulation, sedimentation, adsorption, and other processes are usually controlled by physical chemistry of the solid/liquid interface. Most models for these processes are based on studies of colloidal systems in the absence of organic matter. However, almost all part icles are negatively charged due to adsorbed organic material. This raises the question of whether adsorption models based on clean oxide surfaces are useful for a description of natural systems (DAVIS, 1982). The dispersal and sedimentation of clay minerals and other mineral colloids may be influenced appreciably by sorbed humic matter. While humic matter may keep clay particles in a dispersed state under conditions otherwise conducive to flocculation, humic matter could conceivably "cement" clay particles together, as a polyelectrolyte bridge, to form stable aggregates, as in soil, thereby promoting deposition of clay in a hydraulic regime in which individual colloids would be kept in suspension (JACKSON, 197 5). The sorptive nature of the colloidal surface creates the possibility for aggregation between colloids. Aggregation (or coagulation or flocculation) can cause settling of the colloids as the particle densities increase. The tendency of colloids to coagulate is a function of conditions such as pH, ionic strength, solution composition and, as discussed above, repulsion between colloids. In natural (and polluted) waters, these conditions causing flocculation can change and the aggregated particles can disperse back into the solution. The stability of colloids in natural waters cannot be explained by electrostatic theory alone, but must be considered as a combination of electrical, kinetic and purely chemical forces (STUMM and MORGAN, 1981, p. 660). Dissolved organic matter in solution will influence the sorption chemistry and aggregation behavior of mineral particles in aqueous systems. The presence and nature of suspended and dissolved minerals, in turn, will influence the behavior of the dissolved organic matter. The aqueous phase will thus contain suspended and dissolved mineral/organic colloids at greater or lesser concentration as a function of ambient chemistry and physical conditions. Organic material can form colloids when aggregates or micelles form. Mineral/organic colloids can exist when mixed aggregates coprecipitate or agglomerate in solution, or when conditions bring mixed material into apparent solution. BIOCOLLOIDS The term "biocolloids" is frequently applied to microbes in solution. Bacteria, algae, protozoans and many other biological agents present in the aqueous phase may be considered to exhibit colloidal behavior. Insofar as these species are able to sorb contaminants like other colloids, the distinction between living and nonliving colloids is relatively unimportant. It is also known that biological exudates or subcellular fragments may exist in solution (AWWA COMMITTEE, 1981; XUE et al.,1988). The sorptive nature of bacterial or algal exterior membranes is well documented. Biological particles can influence the distribution of heavy metals in natural waters because the functional groups on the cell surfaces are able to bind metal ions (XUE et a l.,1988). The mechanism of sorption of metals onto biological surfaces seems to be of different sorts. CRIST et al.,(1990) report that adsorption of Sr on Vaucheria released an equivalent amount of Ca and Mg, indicating that metal adsorption by alkali and alkaline-earth metals is an ion-exchange phenomenon based on electrostatic interactions. Release of protons when Cu was adsorbed demonstrated additional covalent bonding for this transition metal. Protonated ethylenediamine is adsorbed both as a cation similar to metals and as a neutral species, indicating the presence of additional bonding sites. Anions such as carboxylate groups of pectin, the polymer of galacturonic acid, are the most likely sites for electrostatic bonding. These substances are found in microbes (CRIST et al.,1990). Microbes are ubiquitous in the subsurface and as such may play an important role in groundwater solute behavior. Microbes in the subsurface can influence contaminants by solubility enhancement, precipitation or transformation (biodegradation) of the contaminant species. Microbes in the groundwater can act as colloids or participate in the processes of colloid formation. Bacterial attachment to saturated granular media can be reversible or irreversible and it has been suggested that extracellular enzymes are present in the system. Extracellular exudates (slimes) can be sloughed-off and act to transport sorbed materials (AWWA COMMITTEE, 1981). The stimulation of bacterial growth in the subsurface may be considered as in situ formation of colloids. In the same way as described for surface water, inputs of dissolved organic matter from the surface tend to stimulate microbial activity because they constitute reduced carbon which can be utilized as a substrate. Subsurface microbial activity associated with inputs of organic substrate will consume oxygen and create reducing conditions if oxygen demand exceeds supply. The availability of oxygen in the groundwater can influence the redox status of system and hence the water chemistry and colloidal status. Biological activity surrounding soil deposits of oxidizable organic matter has been found to create sharp and highly localized drops in the redox potential (PARKIN 1987). Such an event could cause a reduction of a metal in complex and result in a change in the status of the complex. Ion complexes such as chelates may be altered by changes in the redox potential (DOUGLA S and QUINN, 1989). If such an event was truly localized, a contaminant metal ion might undergo reduction and then re-oxidation as it was transported out of the localized reducing zone and into a zone of higher pE. In addition to the colloidal behavior of microbial species acting to transport or influence the availability of pollutants, biotransformation or degradation may occur as well. Biodegradation can occur even when concentrations are very low. It is reported that compounds such as chlorinated diphenylamines (products of microbial degradation of pesticides), naphthalene, styrene, chlorobenzenes, 2,4-D, and Sevin are biodegradable at concentrations below 100 µg/L (AWWA COMMITTEE, 1981). COLLOIDS IN GROUNDWATER Groundwater was long considered to be relatively invulnerable to contamination from surface activities; a consideration which has been widely reevaluated as water quality concerns and analytical capabilities have advanced (WILSON et al., 1981). Contamination of groundwater has been widely documented and is of great concern. Groundwater is frequently valuable and groundwater contamination is often very difficult to remedy. The role of colloids in groundwater contamination is a topic of current interest and development. Groundwater which underlies the soil surface receives input from the soil as the soil solution migrates downward in the percolating water column under the influence of gravity. Because of variations in physical properties such as porosity and permeability involved in geologic depositional patterns, the hydraulic connections between the groundwater and the surface can be rather complex. The directions and velocities of subsurface flow are also subject to variation for the same reasons. Groundwater is connected to surface through unsaturated flow from human surface activities such as agriculture or surface landfills (WILSON et al.,1981; BARKER, et al.,1986). In addition to unsaturated flow through the vadose zone making inputs to groundwater, surface water is frequently well-connected with the groundwater and polluted surface water can move into the subsurface through saturated flow (SCHWARZENBACH and WESTALL , 1981; SCHWARZENBACH et al.,1983). Groundwater contamination from surface activities such as sewage infiltration, agricultural practices or waste disposal is becoming well known. Colloids may be transported directly from the surface through the matrix of soil pores. The soil will tend to filter out some materials, but since many colloids are smaller than soil pores through which water moves, the filtration effect will probably be relatively small. It is likely that the processes of surface sorption, whereby colloids are sorbed onto the solid phase are more important in removal of colloids during infiltration through the surface than are processes of physical sieving or straining by the soil media. Organic macromolecules can move with the regional groundwater flow as demonstrated by ROBERTSON et al.,(1984) who found macromolecular tannins and lignins transported 1000 m from a waste pulp liquor lagoon. A change in the salt balance of percolation water can cause deflocculation of the surface soil and result in the transport of mineral colloids into the groundwater. This was documented by NIGHTINGALE and BIANCHI (1977) who described groundwater turbidity which became evident in wells near a groundwater recharge facility in Fresno, California. These workers reported that the colloidal turbidity was transported from the surface and through the aquifer to the wells; a distance of several miles in some cases . The turbidity was determined to be caused by surface application of recharge water with low ionic strength which caused clay in the surface soil to disperse and form colloids which moved with the bulk flow. Metals can be transported to the subsurface in colloidal state. Colloidal transport of metals, especially polyvalent metals able to form complexes with organic polyelectrolytes, has been suggested as a possible explanation of groundwater contamination by cobalt and uranium at the Oak Ridge National Laboratory in Tennessee (MEANS et al.,1978). It is reported that dispersed colloids have been observed to transport DDT and paraquat through vertical soil columns. The DDT was sorbed onto sewage sludge and the paraquat onto montmorillonite. Transport was enhanced by water of low ionic strength (VINT EN et al.,1983). Colloidal transport in the saturated subsurface occurs with the bulk flow of the groundwater. The potential for colloid transport is suggested because the colloid particles are far smaller than the pores in permeable and fractured media, and their high surface area per unit mass means that they will be effective sorption substrates. Colloid removal from solution by capture onto fixed media surfaces is controlled by the Brownian motion of the colloids and the attachment efficiency following collision. In general, for natural waters with low ionic strengths, colloid attachment to surfaces is hindered by electrostatic repulsion, but predictions based on double-layer theory underpredict observed attachment by orders of magnitude (BUDDEMEIER and HUNT, 1988). In unconfined, shallow aquifers, unpolluted groundwater has been reported to be well-supplied with oxygen, i.e., oxygen levels were measured and found to be near saturation, while pollutant plumes containing dissolved organic matter are strongly reducing in the interior of the plume where oxygen diffusion rates are inadequate to meet the demands of microbes oxidizing the organic matter (THURMAN et al.,1986). Such plumes are considered to have an oxic interface where the contaminant plume meets the uncontaminated groundwater (ROBERTSON et al.,1984). Precipitation events might be expected at such an interface where oxidation of reduced iron or other elements occurs. These reduced species such as Fe(II) which are formed in the reducing environment of the pollutant plume will be oxidized at the oxic interface to form less soluble species such as Fe(III). The precipitation event might serve to scavenge dissolved solution components and remove them from solution if the precipitate was settled and became associated with the solid phase. It has been suggested that such a precipitation event might serve to modify the hydraulic conductivity of the aquifer as small particles deposited in the interstitial pores prevented water from flowing through the pores. If, on the other hand, the precipitated mineral was of small enough size or low enough density to be suspended in the groundwater, the resulting colloid might serve as a vehicle for transport of the pollutant. There is good evidence that colloids are formed in situ in the subsurface as changes in water chemistry perturb the local equilibrium and cause precipitation of dissolved minerals. This was pointed out by GSCHWEND and REYNOLDS (1977), who analyzed colloid s from groundwater near a secondary sewage infiltration site. It was determined that these microparticles consisted primarily of iron and phosphorous. The authors concluded that these microparticles were formed by sewage-derived phosphate combining with ferrous iron released from the aquifer solids, and that these colloids may be moving in the groundwater flow. The chemical nature of the colloids recovered from the groundwater was different from that of the colloids in the wastewater influent, indicating that the colloids were formed in the subsurface and did not simply move into the aquifer from the infiltration ponds. These authors described a scenario of a drop in the redox potential of the groundwater driven by dissolved organic matter in the sewage. This solubilized iron from the aquifer matrix which combined with the phosphate in the sewage to precipitate as an insoluble colloidal suspension in the groundwater. They suggested that such a subsurface transport process could have major implications regarding the movement of particle-reactive pollutants traditionally viewed as non-mobile in groundwater. This same process was envisioned by PENROSE et al.,(1990) who were investigating groundwater contamination by radionuclides in a shallow unconfined aquifer. Their water analyses revealed a drop in dissolved oxygen in the groundwater relative to the surface water. Under anoxic conditions, they expected that ferrous iron would be leached from the aquifer matrix. It would reprecipitate upon encountering oxygen, producing colloidal materials and particulates. Confined aquifers (those which are overlain by an aquiclude and are thus relatively isolated from the surface) are observed to sustain comparatively reducing conditions as the water moves downslope from the recharge area. Presumably, organic matter contained in the recharge water contributes to the oxygen demand in excess of supply. This results in a depleted oxygen condition which the confining conditions sustain by preventing oxygen diffusion from the terrestrial atmosphere. A steady source of organic m atter into a confined aquifer can produce very reducing conditions. (CHAMP et al.,1979) Under extremely reducing conditions, sulfide minerals may form as insoluble precipitates with colloidal dimensions (ROBERTSON et al.,1984). Co-precipitation might involve otherwise soluble contaminant species in these colloids by the mechanisms described previously. COLLOIDAL METAL BEHAVIOR In addition, dynamic interactions at the solution-solid interfaces determine the transfer of metals between aqueous and solid phases. Thus, trace metals may be in a suspended, colloidal, or soluble form. The suspended and colloidal particles may consist of (1) compounds or heterogeneous mixtures of metals in forms such as hydroxides, oxides, silicates, or sulfides or, (2) clay, silica, or organic matter to which metals are bound by sorption, ion exchange or complexation. The soluble forms are usually ions , simple or complex, or unionized organometallic chelates or complexes. Most highly charged metal ions (e.g. Th+4, Fe+3, Cr+3) are strongly hydrolyzed in aqueous solution. Fe(H2O)6+3 + H2O= Fe(H2O)5OH2+ + H3O+ Hydrolysis may also proceed further by the loss of one or more protons from the coordinated water. Fe(H2O)5OH2+ + H2O= Fe(H2P)4(OH)2+ + H30+ Many divalent metals (e.g. Cu+2, Pb+2, Ni+2, Co+2, and Zn+2) hydrolyze within the pH range of natural waters. The hydrolysis of aqueous metal ions can also produce polynuclear complexes containing more than one metal ion, for example, 2FeOH2+ = Fe2(OH)24+ Polymeric hydroxo forms of metal ions (e.g., Cr+3) may condense slowly with time to yield insoluble metal oxides or hydroxides. Polymeric species are important in moderate to high concentrations of metal salt solutions. Metal ions also react with inorganic and organic complexing agents present in water from both natural and contaminant sources. Dominant inorganic complexing ligands include Cl-, SO4-2, HCO3-, F-, sulfide, and phosphate species. These reactions are somewhat similar to the hydrolysis reactions of metal ions in that sequences of soluble complex ions and insoluble phases may result depending on the metal and ligand concentrations and pH. Inorganic ligands are usually present in natural waters at much higher concentrations than the trace metals they tend to complex. Each metal ions has a speciation pattern in simple aqueous solutions that it dependent upon (1) the stability of the hydrolysis products and (2) the tendency of the metal ion to form complexes with other inorganic ligands. For example, Pb(II), Zn(II), Cd(II) and Hg(II) each form a complex series when in the presence of Cl- and/or SO4-2 at concentrations similar to those of seawater. The pH at which a significant proportion of hydrolysis products are formed is dependent upon the concentration of the ligand, for example, Cl- competing with OH- for the metal ion. Metals can also bond to natural and synthetic organic substances by way of (1) carbon atoms yielding organometallic compounds, (2) carboxylic groups producing salts of organic acids, (3) electron donating atoms O, N, S, P, and so on forming coordination complexes, or (4) pi-electron donating groups (e.g. olefinic bonds, aromatic ring, etc.). It is reported that under aerobic conditions free metal ions occur mainly at low pH, and with increasing pH the carbonate and then the oxide, hydroxide, or even silicate solids precipitate. Metal speciation is also controlled by oxidation-reduction conditions. COLLOIDAL BEHAVIOR OF RADIONUCLIDES Transport of colloidally bound material has been indicated in the movement of radionuclides in groundwater in several instances. This has been well-documented (e.g., SHORT et al.,1988; McCARTHY and ZACHARA, 1989 and others. The contemporary relevance of environmental contamination by radionuclides warrants special consideration of this topic. In their 1977 report by a Group of Experts, the OECD NUCLEAR ENERGY AGENCY reports that plutonium hydrolyzes in the pH range of natural waters. The hydrolysis products are stated to exhibit colloidal behavior and be easily adsorbed on particle surfaces. The solubility of uranium and plutonium is enhanced by the presence of bicarbonate ions, with which they form stable carbonate complexes. This report mentions difficulties of forecasting the behavior of plutonium in the aqueous environment due to the various possible oxidation states of the element and the great chemical variability of natural waters. The work cited above reports at the burial ground at Maxey Flats, Kentucky, plutonium has been detected in surface soil, in soil cores 90 cm deep, in monitoring wells, and in streams which drain the site. It is not clear if ground water has been the main dispersing medium, since spreading of contamination above ground has been suggested. However, in case of underground transfer, the data would indicate migration of tens and possibly hundreds of meters in less than 10 years. The most likely migration paths would be along the fractures and joints that are fairly well developed in the burial formation. In a case like this the high adsorption capacity of the shales containing the waste would not constitute an effective barrier since only a small fraction of the water moves through the intergranular interstices. MEANS et al.,(1978) found that radionuclides (cobalt and uranium) in nuclear waste buried at the Oak Ridge National Laboratory facility were migrating in chelated form with EDTA, which is used to cleanup accidental spills. The chelated material was buried and is migrating into groundwater as a stable complex. The same report also states that radionuclide complexes with humic material were found, although this was not thought to be responsible for the initial movement of the contaminant from the site of burial. It is suggested that the humic/radionuclide complex was formed in the environment after the EDTA/radionuclide complex moved away from its place of burial. This suggests the importance of the stability of the complex; the dissociation tendency of the ligand from the bound material, and the degradability of the chelating agent must be considered. Many humic or biologically produced macromolecules may be able to complex contaminants and may also be susceptible to degradation in the environment. Such an event would release the bound pollutant, perhaps resulting in the formation of another complex. The reported presence of humate-bound radionuclides would seem to indicate that the EDTA/radionuclide complex was dissociating to some degree to enable the humate/radionuclide complex to form. SHEPPARD et al.,(1980) used gel filtration to determine the particle size distribution of radionuclides in association with the colloids. They report that particles of colloidal dimensions are shown to be important potential vehicles for the transport of radionuclide elements in soils and groundwater. For the soils studied, the distribution of radionuclides between the soil and aqueous phases is determined by a characteristic particle size spectrum of radionuclide-bearing particles. These spectra, which are related to the physical and chemical composition of the soils, include uncomplexed ions, complexes of fulvic and humic acid polymers, and larger radionuclide-bearing particles, such as clay. For the soils and experimental techniques used, the authors report that three broad classes of particles determined the radionuclide distribution ratios; (1) ionic particles containing radionuclides which have radii less than 1 nm, (2) complexes of humic matter, possibly humic acid polymers with molecular weights between 8000 and 50,000 (2-3 nm radii), and (3) larger soil particles bearing radionuclides and with radii in the 10-60 nm range. In this particular study, the authors found a maximum in radionuclide capacity occurring at sizes corresponding to molecular weights of 8000-50,000 (2-3 nm.). This is in the range of mean molecular weights of soil humic acids reported by KONONOVA (1961). Strontium is reported to be strongly bound by humic matter. The reported distribution ratio of Sr at 3 nm is three orders of magnitude larger than its ionic component (SHEPPARD 1980). Uranium is often found in natural deposits of organic matter (JACKSON, 1975). This indicates an affinity of uranium or its geologic precursors for organic matter. Workers in Canada (CHAMP et al.,1984) investigated radionuclide contamination of groundwater. The authors suggested that organic ligands enhanced the mobility of Co, Ce, Cs, Eu, Sb and Zr. An in situ glass-block-leaching experiment is one of the four study sites at the Chalk River Nuclear Facility in Canada. Field measurements of 137Cs migration indicated transport four times farther than predicted from Kd values. Soil columns prepared from undisturbed, uncontaminated cores showed Cs transport by 0.2 - 1.0 micrometer particles. There was also an indication that microorganisms were involved in particulate Cs release and transport. Work by BUDDEMEIER and HUNT (1988) conclusively demonstrates radionuclide transport by colloids in groundwater flowing through fractured media at the Nevada Test Site. Groundwater samples collected from within a nuclear detonation cavity and from approxim ately 300 meters outside the cavity were analyzed for chemical composition and radionuclide activity. In samples from both locations, approximately 100% of the transition element (Mn, Co) and lanthanide (Ce, Eu) radionuclides were associated with colloids . Their presence outside the cavity indicates transport in the colloidal form. Equilibrium distribution coefficients calculated for Ru, Sb, and Cs nuclides from both field sample locations indicate partitioning on the 0.05-0.003 micrometer colloids. Calculation of transport efficiencies relative to colloid mass concentrations and dissolved neutral or anionic nuclides indicates that both the cations and the radiolabeled colloids appear to experience capture by, or exchange with, immobile aquifer surfaces. The following conclusions are offered from their work: (1) There is a strong association between colloidal solids and radionuclides, particularly in the case of elements that are strongly sorbing and/or insoluble under groundwater conditions. (2) Both the dissolved and colloidal radionuclide species undergo hydrological transport through the fracture-flow system. (3) A substantial proportion of the radionuclides associated with suspended colloids pass through the conventional filters traditionally used to distinguish between "particulate" and "dissolved" species. (4) Measurements indicate that most of the colloidal material consists of natural minerals. There is some evidence that the high colloid concentrations are not unique to the specific study site. (It was observed that no noticeable turbidity accompanied the colloids). (5) Chemical analyses, isotopic comparisons and distribution coefficient calculations all suggest that the radionuclides have chemically equilibrated with the groundwater system, and that their behavior is representative of their stable analogs and of trace element geochemistry in general. (6) Colloidal nuclides and soluble cations are transported less efficiently than soluble neutral or anionic species, indicating that both colloids and cations experience some additional processes of loss to or exchange with the aquifer during hydrogeological transport. The work of PENROSE et al.,(1990) describes the release of treated aqueous wastes containing traces of plutonium and americium into a desert canyon, within the site of the Los Alamos National Laboratory, N.M. The wastes infiltrate a small aquifer within the canyon. Although laboratory studies have predicted that the movement of actinides in subsurface environments will be limited to less than a few meters, both plutonium and americium are detectable in monitoring wells as far as 3390 m downgradient from the discharge. Investigation of the properties of the mobile actinides indicates that the plutonium and part of the americium are tightly or irreversibly associated with colloidal material between 25 and 450 nm in size. A fraction of the americium exists in a low molecular weight form (diameter less than or equal to 2 nm) and appears to be a stable, anionic complex of unknown composition. COLLOID SUMMARY Application of contaminant behavior models which neglect the colloidal phase may result in inaccurate and unfortunate estimations of apparent solubility and transport. The impact of colloidal solubility enhancement will be the most pronounced for the least water soluble solutes. The affinity of a solute for a colloid is a function of the same tendencies which drive a material to sorb onto the stationary solid phase; namely bonding interactions and hydrophobicity. Hence, colloids will manifest the greatest solubility enhancement for those materials which are the least soluble in water or the most attracted to the solid phase. Materials which are soluble in water are less likely to be sorbed onto the solid or colloidal phase in the absence of specific bonding interactions. A 1985 report by CARON et al., reported that dissolved organic matter increased the apparent solubility of DDT significantly, while the solution behavior of lindane was virtually unchanged. Lindane is reported to be 3 orders of magnitude more soluble in pure water than DDT. The colloidal carbon effects on the apparent solubility of neutral organics is the most significant for low solubility, highly sorbed, solutes. It is for these hydrophobic organics that the relative increases in solubility due to sorption on colloidal material is the greatest (BOUCHARD et al.,1989). Because the affinity of organic solutes for binding to humic substances is related to their hydrophobicity, the potential impact of organic colloids may be assessed from the octanol-water partition coefficient of the solute. Several different estimates ha ve been made as to what constitutes hydrophobicity. Where one draws the line is a matter of judgment and experience. In general, it may be assumed that the larger the Kow, the greater the tendency of the solute to partition out of the aqueous phase. Octanol-water partition coefficient values are widely cataloged. In a similar manner to the assessment of the hydrophobicity of organic material, the water solubility of inorganic solutes may be considered as an important parameter in estimating the apparent solubility of metals due to the presence of colloids in solut ion. Because of the potential diversity of the solution and colloidal composition, assessing the behavior of metal-colloids in solution may be very challenging. The presence of insoluble counterions and complexing agents may act to influence the status of a metal ion. At any rate, the composition and concentration of the colloidal phase will obviously have an impact on the ability of such a phase to transport slightly-soluble contaminants. Colloids are more or less an extension of the solid phase, i.e., biological, mineral and organic matter, and they will vary in composition in similar ways. Research has shown that colloidal organic matter does differ in its ability to sorb neutral organics, and that organic colloid sorptivity increases with size and hydrophobicity. The concentration of colloids in the water will have an effect on the degree of solubility enhancement. It is often reported that dissolved organic carbon contents of natural waters are typically in the region of 5 mg carbon/L (THURMAN and MALCOLM, 1981). Given the proven importance of molecular size and polarity, such a number really doesn't mean much in an absolute sense (CHIAN and DeWALLE, 1977). Not surprisingly, dissolved organic carbon levels are reported to have a very wide range. A reported higher carbon-to-oxygen ratio of groundwater than surface water humic material suggests a lower polarity and therefore greater sorptivity for neutral organics by groundwater humic material (BOUCHARD et al.,1989). While it is possible that colloids may be involved with water-soluble contaminants as well, it is doubtful if the role of colloids in transporting such materials will be significant relative to transport in the water. It is possible that sorption of any material onto colloids will affect the degradation or transformation of that material. The water soluble nature of a colloid would seem to indicate a hydrophilic surface which could conceivably sorb a polar or ionic species. It has been frequently reported that sorption/association with a surface affects the chemical transformation of the sorbed/associated species. The increased surface area of colloids in solution will introduce increased interfacial area into the system and hence will tend to enhance surface-related reactions. This aspect of colloidal behavior makes colloids an important consideration in the BARR process. ----------------------------------------------------------------------------- This document is from the WELL's gopher server, gopher.well.sf.ca.us in the Science and Environment menus. For info e-mail to gopher@well.sf.ca.us .