Hyperfiltration-induced isotopic fractionation: Mechanisms and role in the subsurface

This study consists of four series of hyperfiltration experiments. The experiments described in the first chapter concern membrane-induced fractionation of hydrogen and oxygen isotopes. The conclusion is that hydrogen and oxygen isotopes are fractionated when distilled water is passed through a clay membrane because isotopically heavier \u27flickering\u27 clusters of water molecules are preferentially retarded within the membrane structure. The second set of experiments (Chapter II) showed that an uncharged, synthetic membrane preferentially retained $\sp7$Li during hyperfiltration whereas a charged smectite membrane preferentially retained $\sp6$Li. Therefore, hyperfiltration-induced isotopic fractionation of solutes in uncharged membranes is attributed to simple diffusion. Fractionation of solutes in charged membranes is attributed to back diffusion, wherein the charge on the membrane more successfully repels the lighter isotope against the advective flow and therefore reduces its ability to contact the membrane. It was also found that the magnitude of isotopic fractionation, no matter what the mechanism, decreases with the buildup of solute at the high-pressure membrane face. The third set of experiments (Chapter III) used a cross-flow hyperfiltration configuration in which only a small portion of the feed is allowed to permeate through the membrane. This is the configuration used in commercial reverse osmosis desalination. The membrane used was an FT-30 (FilmTec Corp.). Two conclusions were reached from these experiments. First, the membrane has nearly identical rejection characteristics for LiCl and NaCl. Second, the FT-30 membrane preferentially retains $\sp7$Li while the permeate is enriched in $\sp6$Li. The maximum fractionation obtained was 10.7 per mil. This suggests that there may be commercial applications of hyperfiltration-induced isotopic fractionation. The forth set of hyperfiltration experiments (Chapter IV) demonstrated that increasing the solute concentration in a smectite membrane results in increased permeability of the membrane to the solute. The increase in solute permeability occurs because the double layer thickness within the clay membrane decreases as solute previously rejected by the membrane and concentrated at the high-pressure membrane face gradually enters the membrane in ever increasing amounts. Eventually, due to the solute permebility decrease, the effluent concentration becomes greater than that of the feed solution for a time as solute previously accumulated behind the membrane dissipates. In other words, the death of the clay membrane is a natural consequence of its ability to reject solute--its death is self induced. This suggests that the membrane effect in the subsurface has the ability to hide its tracks