How membrane permeation is affected by donor delivery solvent.

We investigate theoretically and experimentally how the rate and extent of membrane permeation is affected by switching the donor delivery solvent from water to squalane for different permeants and membranes. In a model based on rate-limiting membrane diffusion, we derive explicit equations showing how the permeation extent and rate depend mainly on the membrane-donor and membrane-receiver partition coefficients of the permeant. Permeation results for systems containing all combinations of hydrophilic or hydrophobic donor solvents (aqueous solution or squalane), permeants (caffeine or testosterone) and polymer membranes (cellulose or polydimethylsiloxane) have been measured using a cell with stirred donor and re-circulating receiver compartments and continuous monitoring of the permeant concentration in the receiver phase. Relevant partition coefficients are also determined. Quantitative comparison of model and experimental results for the widely-differing permeation systems successfully enables the systematic elucidation of all possible donor solvent effects in membrane permeation. For the experimental conditions used here, most of the permeation systems are in agreement with the model, demonstrating that the model assumptions are valid. In these cases, the dominant donor solvent effects arise from changes in the relative affinities of the permeant for the donor and receiver solvents and the membrane and are quantitatively predicted using the separately measured partition coefficients. We also show how additional donor solvent effects can arise when switching the donor solvent causes one or more of the model assumptions to be invalid. These effects include a change in rate-limiting step, permeant solution non-ideality and others.