Graduation Year


Document Type




Degree Granting Department

Civil Engineering

Major Professor

Robert P. Carnahan, Ph.D.


Arsenite, Cake Layer, Chabazite, Copper, Iron, Membrane


Carcinogenic health concerns over arsenic in drinking water caused the USEPA to reduce the maximum contaminant level (MCL) from 50 to 10 ppb, effective on January 23, 2006. This has forced many smaller utilities into expensive treatment or discontinuation of water distribution. Researchers throughout the world are working to develop an inexpensive method for arsenic removal to meet this MCL. Aluminum silicates, or zeolites, are naturally occurring ionic sorbents. Modification of a zeolite may enhance adsorption capacities and ion selectivity. This research investigates the arsenic adsorption capacities of a modified Chabazite. This adsorption, coupled with a hollow fiber, microfiltration membrane substrate, allows for the use of finer zeolite particles. Powdered zeolite creates a cake layer on the filtration surface through which the arsenic solution must filter.

The research goal was to develop an overall mathematical model for the adsorption of arsenic through the adsorption equilibrium isotherms, the cake layer, and the microfiltration operational settings. Baseline adsorption isotherms where performed in distilled water. Solutions containing counter ions were then used to determine any counter-effects. The final isotherms were found using dechlorinated tap water, which is similar to many groundwaters found in the United States. Various runs were used to determine the most efficient modification and loading rate.Initial characterization of the membrane system defined membrane permeability and inherent arsenic rejection. Variable mass loading in both deadend and crossflow filtration determined that the cake layer was not compressible due to linear pressure increases. This process also determined the maximum cake layer permissible hydraulically on the membrane surface.

Membrane system operational characteristics and arsenic dosing were chosen to adhere to these parameters as well as the adsorption isotherms. Adsorption runs were conducted which varied the flux through the membrane, the arsenic feed concentration, and the cake layer thickness. Through the data collected, a mathematical model based on irreversible adsorption was developed. This novel approach to arsenic removal and the predictive mathematical model can be used as an effective method for removal of aqueous arsenic, and may provide small water utilities with a cost effective way to meet the recommended new MCL.