(225a) Emerging Basic Science Questions Regarding Water Transport in Polymers for Water Purification, Resource Recovery, and Energy Applications

This presentation will focus on a foundational question regarding the transport mechanism of small molecules through polymers for liquid separations, specifically water transport through membranes being used or considered for use in, for example, desalination, resource recovery, and fuel cell membranes. Historically, the transport of small molecules, such as gases, water, ions, organic solutes, etc. through dense membranes that do not have fixed, permanent pores spanning the membrane, is described by the solution-diffusion model. In this model, small molecules partition from a contiguous fluid phase into the membrane, and the membrane/fluid interface is at thermodynamic equilibrium, so equilibrium partitioning of solutes into membranes can be modeled using the machinery of thermodynamics. The rate limiting step for transport in virtually all membranes is the diffusion of the small molecules through the membrane, typically drive by a concentration gradient, electric field, or both. Water will permeate through a membrane under the influence of a hydrostatic pressure difference. That is, a membrane exposed to a high hydrostatic pressure on one side and a low hydrostatic pressure on the other will permeate water from the high pressure to low pressure side of the membrane. This phenomenon is commonly observed in, for example, desalination membranes, such as reverse osmosis membranes. The solution-diffusion model uses thermodynamic principles to link the pressure difference across the membrane to a concentration gradient in the membrane, with water permeation occurring because of Fickian diffusion of water down its concentration gradient in the membrane.

In the past few years, studies have been published purporting to demonstrate that, in fact, water transport through such membranes is governed by a pore flow model, where water transport is presumed to occur via a network of interconnected, water-filled subnanometer channels or pores, with water flowing through the pores due to the imposed hydrostatic pressure difference across the membrane. Such pores are presumed to be too small to observe directly by any known technique, so evidence for this hypothesis comes indirectly from (primarily) water transport data and computer simulations.

A dispositive distinguishing feature between the solution-diffusion and pore flow models is the existence of a water concentration gradient inside a membrane subjected to a hydrostatic pressure gradient. In the solution-diffusion mechanism, the hydrostatic pressure difference across the membrane induces a water concentration gradient inside the membrane, and in the pore flow model, no such concentration gradient would be observed. Therefore, we have conducted experimental studies to directly measure the water concentration as a function of distance through a series of polymer membranes, including cellulose acetate, Nafion, and crosslinked hydrogels based on poly(ethylene oxide) under hydrostatic pressure differences as high as 200 bar or more. Our studies show distinct concentration gradients in all of the membrane materials considered, with the flux and concentration gradients well-described by Fick’s law of diffusion and conventional solution thermodynamics.