The objective of this study was to examine membrane filtration of a single stranded DNA (ssDNA) with 60 thymine nucleotides, and to elucidate the variables controlling its transmission across track‐etched porous membranes. Dead end filtration measurements were performed using different pore size membranes (10, 15, and 30 nm) at different transmembrane pressures in solutions with ionic strength ranging from 0 to 1000 mM NaCl. The diffusivity of the ssDNA was determined using fluorescence recovery after photobleaching, yielding hydrodynamic radii ranging from 1.6 to 2.8 nm, with values decreasing with increasing solution ionic strength. Despite the small ssDNA/membrane pore size, nearly 100% rejection was observed for measurements performed with the 10 and 15 nm pore size membranes under low‐ionic strength conditions. These high rejections can be attributed to strong repulsive electrostatic ssDNA‐membrane interactions. With increasing ionic strength, electrostatic interactions as well as the effective size of the ssDNA decreases and the flexibility of the ssDNA increases, leading to a reduction in ssDNA rejection. A design of experiments approach was used to plan filtration experiments that adequately covered the variable space with a manageable number of experiments. The results yielded an empirical expression relating ssDNA rejection to pore size, solution ionic strength and transmembrane pressure. There was evidence of flow induced elongation at high‐transmembrane pressures in the 30 nm pore size membranes, but not in the smaller pore size membranes. These results are consistent with critical flux estimates developed using a free draining model for the ssDNA.

No abstract

In this paper, predictions from a theoretical model describing the rejection of a rigid spherical solute from porous membranes are compared to experimental results for a single stranded DNA (ssDNA) with 60 thymine nucleotides. Experiments were conducted with different pore size track-etched membranes at different transmembrane pressures and different NaCl concentrations. The model includes both hydrodynamic and electrostatic solute–pore wall interactions; predictions were made using different size parameters for the ssDNA (radius of gyration, hydrodynamic radius, and root mean square end-to-end distance). At low transmembrane pressures, experimental results are in good agreement with rejection predictions made using the hard sphere model for the ssDNA when the solute size is described using its root mean square end-to-end distance. When the ssDNA size is characterized using the radius of gyration or the hydrodynamic radius, the hard sphere model under-predicts rejection. Not surprisingly, the model overestimates ssDNA rejection at conditions where flow induced elongation of the DNA is expected. The results from this study are encouraging because they mean that a relatively simple hindered transport model can be used to estimate the rejection of a small DNA from porous membranes.

No abstract