The effect of hydrostatic gradient on solvent drag reflection coefficient in heteroporous membranes: differences between model-dependent and model-independent computations.

The accepted reason for apparent increase in solvent drag reflection coefficient (sigma f) with mean hydrostatic pressure gradient (delta p) across a heteroporous membrane is thought to be twofold. First, fractional volume flux traversing small pores (F) is thought to rise as delta p rises. Second, since it is believed that sigma f is directly related to F, then sigma f should rise with delta p to an asymptote. The first feature of a rise in F with delta p to an asymptote was shown to be true for two published two pore models. However, sigma f was closely correlated with F only if sigma f was computed from summation of known components of flux across each pore. If sigma f was computed by a model independent, purely thermodynamic method, then sigma f was not closely tied to F, and approached its asymptote at considerably lower values of delta p than the summation form of sigma f. It is concluded that although F rises with delta p above critical values of delta p, this alone cannot be the cause of a rising sigma f computed by model independent methods. Another explanation, such as anatomic change is required. In addition, model independent methods for computation of sigma f require fewer assumptions, less information and lower pressures than model dependent methods.