The partition coefficient of solutes into the polyamide active layer of reverse osmosis (RO) membranes is one of the three membrane properties (together with solute diffusion coefficient and active layer thickness) that determine solute permeation. However, no well-established method exists to measure solute partition coefficients into polyamide active layers. Further, the few studies that measured partition coefficients for inorganic salts report values significantly higher than one (∼3-8), which is contrary to expectations from Donnan theory and the observed high rejection of salts. As such, we developed a benchtop method to determine solute partition coefficients into the polyamide active layers of RO membranes. The method uses a quartz crystal microbalance (QCM) to measure the change in the mass of the active layer caused by the uptake of the partitioned solutes. The method was evaluated using several inorganic salts (alkali metal salts of chloride) and a weak acid of common concern in water desalination (boric acid). All partition coefficients were found to be lower than 1, in general agreement with expectations from Donnan theory. Results reported in this study advance the fundamental understanding of contaminant transport through RO membranes, and can be used in future studies to decouple the contributions of contaminant partitioning and diffusion to contaminant permeation.

A reliable, user-friendly, bench-top method was developed and evaluated for the measurement of negative charge density in the active layers of thin-film composite and thin-film nanocomposite membranes. The method consists of isolating the active layer on a quartz crystal microbalance (QCM) sensor (i.e., AL+sensor sample), exposing the AL+sensor sample to an aqueous cesium solution at any pH of interest, and measuring with a QCM the mass of cesium ion that associates with the negative sites of the active layer. Results showed that QCM measurements of charge density in active layers were: (1) repeatable within 3% for tests performed with the same AL+sensor sample under the same experimental conditions; (2) reproducible within 3.8% for tests performed with the same AL+sensor sample when the ionic strength of cesium solutions was varied by 300%; (3) reproducible within 4% for active layers isolated from nearby locations of 2 the same membrane sheet; and (4) consistent within 2.1% at pH = 10.5 with results obtained using the previously reported Rutherford backscattering spectrometry method on non-isolated active layers. The results therefore demonstrate the robustness, repeatability, reproducibility, and accuracy of the QCM method. We also demonstrated that the ionization behaviors of the polyamide-based thin-film composite and nanocomposite membranes tested were similar: both membranes had bimodal pK a distributions and negative charge densities of |0.5 M at full ionization.

Multi-gram synthesis and densification is presented for a porous polymer network (PPN-4) examined as a possible vehicular methane storage material. Compaction at 17,000 psi doubled the bulk density of the material and unexpectedly increased microporosity within the material. As a result, the densified material exhibits higher excess gravimetric methane uptake and improved total volumetric methane uptake relative to the powder.

No abstract