The common approach to investigating the extent of contamination at a hazardous waste site is to drill monitoring wells to obtain geotechnical and chemical information. A new technology is now being utilized at various hazardous waste sites in order to facilitate the hydrogeologic data collection process. The use of cone penetration testing has proven to be a relatively quick and inexpensive method for initial site investigations compared to current techniques. Cone penetrometer testing can quickly determine the stratigraphy of an area, the soil properties, and (by taking samples) the extent of contamination. The data are obtained immediately, which allows in-field evaluation and "on the spot" decisions, resulting in enhanced placement of monitoring wells and in some instances a substantial reduction in the number of wells needed. Fewer monitoring wells means a large cost savings. Applications of cone penetrometer testing are presented and recommendations for further research are discussed.

The rate of movement of contaminants in groundwater is a topic of environmental importance and public concern. Early work in contaminant transport showed that the ratio of the velocity of the water to the velocity of the contaminant (v$\sb{\rm w}$/v$\sb{\rm c}$ or retardation, R) in a flowing saturated soil column can be determined using known values of soil density, soil porosity, percent organic carbon in the soil, and the octanol-water partition coefficient of the contaminant. More recent studies indicate that the retardation of the contaminant in a soil may vary due to soil/contaminant isotherm non-linearity and non-equilibrium conditions in the column. In this research, the effect of velocity on the observed retardation of methylated benzenes is evaluated using three different soils (organic carbon range--0.008% to 0.27%) and a wide range of linear velocities (0.2 ft/day to 40 ft/day). Contaminant solutions of 10 mg/l benzene, toluene, p-xylene and mesitylene, 1 mg/l naphthalene, and 50 mg/l ethylbenzene were pumped through a 0.9 cm ID saturated soil column into a UV/Vis flow-through spectrophotometer for detection of contaminant breakthrough. Breakthrough and flush-out curves, in general, were found to be symmetric, regardless of velocity or contaminant used. Below a velocity of 3 ft/day, compounds with K$\sb{\rm ow}$ values of $>$10$\sp3$ showed an increase in retardation with decreasing velocity. Data were fitted to both an equilibrium and non-equilibrium model with one variable parameter. Results indicated that the data could not be modeled by the non-equilibrium model but did fit the equilibrium model with variable retardation.

The sorption and desorption behaviors of chlorinated hydrocarbons and naphthalene to and from Lake Charles sediments were studied. The sorption of 1,2- dichlorobenzene, 1,2,4-trichlorobenzene, and naphthalene was determined to be linear via hydrophobic effect, and the partition coefficients of these compounds were in the range of literature reported values. The desorption of chlorinated hydrocarbons and naphthalene were studied with two systems--a desorption with continuous removal of aqueous phase contaminants and a desorption by the repetitive water extraction. The desorption of both historically-existed and newly-added compounds showed great deviations from the equilibrium conditions. A biphasic desorption, with a labile phase and a resistant phase, has been observed to exist for both freshly- and aged-contaminated sediments. The desorption from the irreversible compartment reached a near-constant concentration, which was orders of magnitude lower than that predicted by equilibrium equations. The size of the resistant fractions seemed highly related to the initial solid phase concentration. Several factors, including pH, temperature, and competitive sorption, were found to significantly affect the release of historically-existed contaminants from the sediments. The laboratory observations in this study agreed well with other laboratory and field observations. The existence of the irreversible fraction may have a significant impact on regulatory, modeling and remediation activities.

The biodegradation patterns of polycyclic aromatic hydrocarbon (PAH) mixtures were studied in systems with and without sediments. Naphthalene showed no change in its behaviour in the presence of any other PAH. In sediment-free systems, acenaphthene showed a diauxic pattern in the presence of naphthalene, but was unaffected with fluoranthene. Diauxy was not observed in the same mixture in sediment-containing systems. Sediment-free experiments with fluoranthene indicated it was the only PAH not degraded when present alone, and it was only degraded with naphthalene. In sediment-containing systems, fluoranthene degradation only occurred above a critical naphthalene concentration. Experiments using all three PAHs in both sediment-free and -containing systems showed results similar to those obtained using pairs. This work indicated that mixtures of PAHs do have different biodegradation patterns to single compounds. These patterns may be additive. Sediments act as a sink for PAHs, highlighting the concentration dependence of the patterns observed.

This thesis addresses the modeling of the dissolution of residually-trapped multi-component non-aqueous phase liquids (NAPLs) from a porous medium. As part of this work, a model was constructed in which the mass transfer behavior of the NAPL is characterized by a probability density function (pdf) which describes the initial distribution of residual NAPL "blob" volumes. Mass transfer from these trapped NAPL "spheres" is simulated using an empirical relationship for the Sherwood Number, a linear concentration driving force, and multi-component solubility effects. This "Distributed-Site" model is then used to match two sets of dissolution data from a laboratory column test.

Resistant desorption has been widely observed for hydrophobic organic contaminants. It is an important process that impacts sediment quality and contaminated sediment management. Conventional approaches to measure the desorption resistant fraction of sorbed contaminants are by conducting repetitive water desorption, in which the desorption resistant fraction is determined by measuring the solid phase concentration after extended desorption times. This method is time-consuming and impractical for widespread applications. In addition, the final measurement of the solid phase concentration is subject to considerable error, because there is always a fraction of water associated with the solid due to the incomplete separation of the liquid phase from the solid. In this research, a new experimental protocol was developed to rapidly and accurately measure the desorption resistant fraction based on the study of desorption of phenanthrene from Utica sediment.

Association of hydrophobic organic compounds with dissolved organic matter in natural water systems may impact a contaminant's ability to transport across synthetic membranes. Importantly, these interactions can create interferences when monitoring ambient levels of contaminants with a potential for biouptake. The influence of water-quality matrix conditions on the transport and lipid uptake of five hydrophobic organic compounds of environmental concern was investigated by partnering semipermeable membranes with a model lipid phase in a batch dialysis system. Contaminants fell into two characteristic groups based on the response of transport and lipid uptake to exposure conditions: one for which behavior was largely independent of water-quality matrix conditions and one for which alterations to the bulk aqueous phase were impactful. For short exposure periods, the abiotic technique demonstrated the potential to qualitatively replicate the root-to-shoot translocation behavior of non-ionized hydrophobic organic compounds in plant systems.

Samples of atmospheric particles were analyzed for organic and elemental analysis at three sites in Houston, TX. Samples for the quantification of individual organic compounds were collected during August 2000--September 2000 and analyzed for molecular speciation. A chemical mass balance (CMB) model was applied to the organic speciation data to estimate the contributions of the eight possible sources to the fine particulate matter mass in Houston. Major contributors to PM2.5 included gasoline vehicles, diesel vehicles, meat cooking and wood combustion with smaller contributions from vegetative detritus. It was found that PM2.5 mass was also dominated by other organics and secondary sulfate. Samples for airborne metal analysis were collected and analyzed by two different chemical analysis methods; Inductively Coupled-Plasma Mass Spectrometry (ICP-MS) and X-Ray Fluorescence (XRF). Positive matrix factorization (PMF) was applied to the elemental concentration data for source identification and apportionment. PMF resolved five physically interpretable factors at each site of which four were found to be common at all sites: crustal material, road dust, wood burning, and sea salt. The composition of the remaining factor was similar, but not identical at the three sites and had an elemental composition similar to industrial combustion. Crustal material is the most important contributor at each site.

On-lattice Monte Carlo simulations were performed to investigate the effect of collision efficiencies alpha of particles on the morphology of heterodeposits. The model predicts that the structure and surface chemistry of such deposits are determined by particles with large and small values of alpha, correspondingly. The case of homodeposition onto a rough substrate was also studied. The best "memory" of the deposit measured as a rate of decay of the Fourier amplitude corresponding to the wavelength of the initial roughness, was found to correspond to intermediate values of alpha. Cross-flow membrane filtration experiments using monodisperse latex suspensions were performed for validation of the Sethi's extended model. A slightly more complex structure at the cake-suspension interface was hypothesized to explain differences between model calculations and experiments. A representation of the cake with a gradually changing solids concentration allowed for better matching of the extended model and experimental results for particles diffusively depositing onto membrane.

Alumoxane-derived membranes are characterized and compared with commercial anodized alumina and polycarbonate track-etched membranes. The alumoxane-derived membranes were produced using two different types of ligands, acetic acid and (methoxyethoxy)acetic acid, or mixes of the two to alter the membrane characteristics. Membranes were studied using scanning electron microscopy, atomic force microscopy, nitrogen adsorption-desorption, cleanwater flux experiments, goniometer measurements, and Zeta-meter measurements to determine membrane morphology, pore size distribution and shape, permeability, hydrophobicity, and surface charge. Alumoxane-derived membranes were found to have a nodular morphology with over 90% of pores between 5 and 25 nm and permeability ranging from 0.3 to 1.5 nm$\sp2$. The two ligands used did not produce large differences in the overall characteristics of the membranes. Alumoxane-derived membranes exhibited similar permeabilities to the commercial membranes tested. Carboxylate-alumoxanes show considerable promise as precursors to membranes and other alumina products.