Two-region linear/nonlinear sorption modeling: Batch and column experiments

Batch and column experiments were conducted to assist in elucidation of the mechanism and rates of slow diffusion of 1,2,4-trichlorobenzene into Borde n sand. Batch rate experiments were;conducted at three concentration levels , and a full isotherm was developed over 550 d. The equilibrium isotherm sh owed significant nonlinearity (Freundlich exponent = 0.75). Three column ex periments were conducted-two at low velocity (each at a different input con centration) and one at high velocity. The data were analyzed with four diff erent models; the two most, successful are described in detail here. These two models divided sorption between instantaneously equilibrating and diffu sion-limited domains, with either a linear or nonlinear sorption isotherm a ssumed for the instantaneously equilibrating fraction of sorption sites. Ei ther model could simulate any single dataset by adjustment of two parameter s; however, no single parameter set could simulate all data sets. The model with a linearly sorbing instantaneous fraction described the batch data be tter than the model with a nonlinearly sorbing instantaneous fraction; howe ver, a nonlinearly sorbing instantaneous fraction described the slow column experiments much better than a linearly sorbing instantaneous fraction. Th e high-flow-rate column experiment indicated that the assumed instantaneous fraction is actually influenced by-rate limitations, suggesting that more complex models are needed in order to simulate this shorter time scale beha vior. Overall, our work illustrates how one needs to include a wide variety of experimental conditions in order to fully test the complex set of sorpt ion equilibrium and rate mechanisms that are present in natural solids.