Aqueous phase isotherms were calculated from vapor phase desorption isotherms measured at 15, 30, and 60 °C for trichloroethylene on a silica gel, an aquifer sediment, a soil, a sand fraction, and a clay and silt fraction, all at 100% relative humidity. Isosteric heats of adsorption (Qst(q)) were calculated as a function of the sorbed concentration, q, and examined with respect to the following mechanisms: adsorption on water wet mineral surfaces, sorption in amorphous organic matter (AOM), and adsorption in hydrophobic micropores. Silica gel, sand fraction, and clay and silt fraction 60 °C isotherms are characterized by a Freundlich region and a region at very low concentrations where isotherm points deviate from log-log linear behavior. The latter is designated the non-Freundlich region. For the silica gel, values of Qst(q) (9.5−45 kJ/mol) in both regions are consistent with adsorption in hydrophobic micropores. For the natural solids, values of Qst(q) in the Freundlich regions are less than or equal to zero and are consistent with sorption on water wet mineral surfaces and in AOM. In the non-Freundlich regions, diverging different temperature isotherms with decreasing q and a Qst(q) value of 34 kJ/mol for the clay and silt fraction suggest that adsorption is occurring in hydrophobic micropores. The General Adsorption Isotherm is used to capture this adsorption heterogeneity.
Isothermal desorption rates were measured at 15, 30, and 60 °C for trichloroethylene (TCE) on a silica gel, an aquifer sediment, a soil, a sand fraction, and a clay and silt fraction, all at 100% relative humidity. Temperature-stepped desorption (TSD) rates were measured for these solids in columns prepared and equilibrated at 30 °C, but heated instantaneously to 60 °C after ∼1000 min of slow desorption. Fast and slow elution rates are observed for all solids. Modeling results for the fast eluting fraction of TCE show that fast desorption is controlled by diffusion through aqueous filled mesopores. Rates predicted from diffusion and surface-barrier models are compared to slow isothermal and TSD rates. Diffusion model fits are superior to surface-barrier model fits in all cases. Slow diffusion coefficients and a high activation energy calculated from silica gel data (∼34 kJ/mol) indicate that slow desorption is controlled by activated diffusion in micropores. Initial amounts of slow desorbing TCE do not affect these rates and are found to obey Polanyi's equation. The mass adsorbed in non-Freundlich isotherm regions, where micropores are hypothesized to control adsorption, is 10 times greater than the mass adsorbed at the onset of slow desorption, suggesting that these pores are undulating in nature. TSD column results are consistent with a mechanism where slow diffusion rates are controlled by sorptive forces at hydrophobic micropore constrictions.
In the first paper of this two‐paper series, we present a new model that attributes nonequilibrium sorption of moderately hydrophobia, volatile organic compounds to intragranular diffusion. The model differs from those of previous researchers in that for the first time, it combines the following elements: (1) We account for two distinct intragranular rate‐limiting diffusion processes, occurring in series and at widely different timescales; (2) we describe the slower of the two processes with a gamma distribution of diffusion rates; and (3) we use the disparity of timescales of the two processes to approximate a boundary condition for the distributed diffusion equation, allowing it to be solved analytically. The slower diffusion process is attributed to activated diffusion through very small pores, called micropores. In paper 2 [Werth et al., this issue] we evaluate the capabilities of the model and use it to interpret experimental results.
Trichloroethylene (TCE) elution profiles for purged and unswept columns are presented and simulated with the Distributed Dual Diffusion Model (DDDM) presented in the first of this two‐paper series. Elution profiles were measured at 15, 22, 30, and/or 60°C for a silica gel, a Livermore sand fraction, a Livermore clay and silt fraction, a Santa Clara sediment, and/or a Norwood soil, all at 100% relative humidity. Advection and dispersion control TCE transport through the vapor phase of purged columns. Diffusion controls TCE transport through the vapor phase of unswept columns. For both purged and unswept columns a fast and a slow desorbing fraction of TCE were observed. The DDDM effectively simulated both of these fractions. For the fast fraction the DDDM predicted desorption with no fitting parameters. For the slow fraction the DDDM was not predictive but it simulated desorption using either a single (for silica gel) or a gamma distribution (for soil and sediments) of micropore diffusion rate constant(s) and a micropore capacity factor. Micropore capacity factors obtained by fitting the DDDM to purged column results were used to predict the onset of slow desorption in unswept columns of the same solid.