Tuesday, 7 October 2008: 3:45 PM
George R. Brown Convention Center, 362F
The purpose of this paper is to propose a conceptual model for soil water repellency based on Langmuir-Blodgett and self-assembled monolayers of di- and multivalent fatty acid salts. The goal is to explain time-dependent wetting and many hitherto unexplained features of repellent soils. Soils have many negatively charged surfaces that can interact with ionized fatty acid headgroups via divalent cation bridges or directly via lattice cations of positively charged mineral surfaces. The long alkyl chains will be oriented nearly perpendicular to the surface, rendering it hydrophobic. In contact with water, some of the molecules undergo desorption or reorientation, or hydrophilic surface moieties become hydrated; all of these mechanisms can contribute to a time-dependent decrease in hydrophobicity. To test this model, we examined the effect of mono- and divalent cations at different pHs on repellency persistence (as measured by drop penetration time) of water repellent and wettable soils. Repellency persistence increased at elevated pH in the presence of Ca2+ but not Na+, increased at high cation concentrations, and decreased at elevated temperature. These results are consistent with the behavior of monolayers of fatty acid divalent salts adsorbed at solid substrates, but not with the behavior of other hydrophobic film-forming compounds. Abundant long chain fatty acids were identified in the test soils and in many other repellent soils. Conditions required for development of soil water repellency include: (i) organization of fatty acid molecules with hydrophobic tails facing outward from soil particle surfaces; (ii) slow rates of reorientation or desorption of surface molecules when in contact with water; and (iii) low abundances of easily-hydrated hydrophilic moieties at particle surfaces. The extent of repellency persistence will be substantially affected by small differences in these conditions, and the absence of any one will render a soil wettable.