Absorption and evaporation of volatile organic solvents from human skin in vitro

Accurate estimation of evaporation and permeation rates of volatile organic solvents through human skin in vivo is important for occupational and environmental risk assessments, which include factors such as the solvent's capacity to be a local sensitizer or a systemic toxin. Current risk assessment models for volatile organic solvents often include the assumption that 100% of the dermal dose of toxicant is absorbed, which results in an overly conservative risk assessment. The overall goal of this research was to further develop and improve an existing skin diffusion model by experimentally confirming the predicted evaporation and absorption rates of volatile, topically applied organic compounds based on their physicochemical properties, the known biological properties of skin and principles of diffusion theory.;Twenty-one compounds with varying physicochemical properties were studied to determine evaporation rates from human skin in vitro mounted on Franz diffusion cells under two exposure conditions, bench top and fume hood. Correlations of the data according to five evaporative mass transfer models from the occupational and environmental safety literature were examined. We found the US EPA evaporation model (Peress J. Chem Eng Prog April, 2003, 32-34) to be the most suitable model for the analysis of laboratory skin diffusion studies with volatile compounds. The effective values for wind velocity in bench top and fume hood experiments were found to be 0.23 and 0.92 m s-1, respectively.;In vitro human skin permeation of two hydrophilic solvents (acetone and ethanol) and two lipophilic solvents (benzene and 1,2-dichloroethane) were studied in the same diffusion cells under the fume hood conditions. Four doses of each compound were tested -- 5, 10, 20, 40 muL cm-2 , corresponding to specific doses ranging in mass from 5.0 to 63 mg cm-2. 14C-radiolabed compounds were employed. A decrease in percent of absorption of each solvent with increasing dose was observed. This decreased could be explained in terms of a stratum corneum deposition region in the diffusion model. Overall permeation of acetone, benzene and 1,2-dichloroethane was under predicted; however the diffusion model satisfactorily described ethanol. As large doses of the first three compounds are known to extract skin lipids, lipid disruption by even small doses of these compounds was postulated to contribute to the under prediction.;In order to more closely describe the permeation data, two key parameters in diffusion model required modification: the stratum corneum/water partition coefficient, KSC and diffusion coefficient of the permeant in the stratum corneum, DSC. In most cases the magnitude of change of the fitted values of KSC and the corresponding skin concentrations were not physically realistic when a fixed stratum corneum thickness was employed. This finding provided strong evidence for both skin swelling and barrier disruption. Recommendations for future work included conducting accurate stratum corneum-water partitioning studies for solvents under varying levels of skin hydration and the development of a more flexible skin-swelling model to count for co-solvent interaction with the stratum corneum.