Fundamental and applied studies of polymer membranes

Abstract

With the recent increase of feed titer concentrations in the biotechnology industry, the majority of the costs of a given process have been shifted downstream. In order to overcome this problem, we have developed a novel synthesis method in order to increase the breadth of our high throughput screening library. This library was generated using maleimide chemistry to react a common methacrylate linker with a variety of different functions groups (R groups) in order to form new monomers that were grafted from the surface of PES ultrafiltration membranes. From this work, we discovered that the chirality of a membrane can affect performance when separating chiral feed streams. This effect was observed when filtering bovine serum albumin (BSA) and ovalbumin in a high salt phosphate buffered saline (PBS, 150 mM salt). The Phe grafted membranes showed a large difference in performance when filtering BSA with selectivity of 1.13 and 1.00 for (S) and (R) Phe, respectively. However, when filtering ovalbumin, the (S) and (R) modified surfaces showed selectivity of 2.06 and 2.31, respectively. The higher selectivity enantiomer switched for the two different proteins. Permeability when filtering BSA was 3.06 LMH kPa-1 and 4.31 LMH kPa-1 for (S)- and (R)- Phe, respectively, and 2.65 LMH kPa-1 and 2.10 LMH kPa-1 when filtering ovalbumin for (S)- and (R)- Phe, respectively. Additionally, these effects were no longer present when using a low salt phosphate buffer (PB, 10 mM salt). Since, to our knowledge, membrane chirality is not considered in current industrial systems, this discovery could have a large impact on the pharmaceutical and biotechnology industries.

The high throughput platform described herein is both robust and versatile. It allows for the modification, screening, and optimization of membranes and surfaces for a wide array of research areas. This powerful method shows great potential for discovering new surfaces that can be tuned for specific applications.

Stem cells are currently grown on biologically based substrates, which do not allow for optimization or tailoring to specific cell type. Using our high throughput screening platform, we were able to determine that (N-[3-(dimethylamino)propyl] methacrylamide), DMAPMA, supported strong attachment and long-term self-renewal of mouse embryonic stem (ES) cells while preventing differentiation (maintaining pluripotency). After developing this platform, it was used to screen for a surface that could instead induce differentiation of bovine and human retinal pigment epithelium (RPE) cells while promoting cell growth. Several PEG based surfaces were able to induce cobblestone morphology of the RPE cells, which is indicative of differentiation.

Three model surfaces were studied: unmodified PES, hydrophobic alkane (C18) modified PES, and poly(ethylene glycol) (PEG) modified PES. In the presence of the low salt phosphate buffer (10 mM salt), phosphate anions were excluded from the PEG-modified PES film. This led to a charge separation between the phosphate anions and sodium cations, creating a surface potential which strongly ordered water molecules into the bulk. When using high salt PBS (138 mM salt) the sodium chloride ions screened this charge and reduced water ordering. Interestingly, this effect was the greatest for the PEG modified surface, with minor or no effects observed for the C18 modified PES and unmodified PES, respectively. This large effect of phosphate ions, present in all phosphate buffers, on water structure at PEG modified surfaces demonstrates the importance of studying modified surfaces in situ instead of using pure water.

Fundamental understanding of the interactions that occur during membrane filtration will be critical for improving future membranes. In order to achieve this, we have been using SFG to study interfacial water on membrane surfaces. We believe that water interactions with the membrane surface and with the feed species, e.g. proteins, play a critical role during the fouling process. Previous studies have only studied the relationship between surface structure, pure water structure, and resistance to fouling at various antifouling surfaces. They have not, however, studied the role of ions and buffers in these systems. Relevant buffers, such as phosphate buffered saline (PBS) and phosphate buffer, contain ions that are known to restructure water at interfaces. Sum frequency generation spectroscopy (SFG) was used to characterize interfacial water structure at poly(ether sulfone) (PES) thin films in the presence of 0.01 M phosphate buffer (low salt) and 0.01 M phosphate buffered saline (high salt).

Ultrafiltration membranes are widely used in the biotechnology, desalination, and water treatment industries. They also see extensive use at laboratories across the world. The major drawbacks of membranes are fouling and concentration polarization, which ultimately lead to a negative change in separation performance. In this work, the effect of surface chemistry on membrane performance, particularly fouling, is examined. We have developed a high throughput screening platform that uses atmospheric pressure plasma to activate a membrane surface, followed by covalent grafting of many different monomers to the surface to impart new chemical functionalities. Four major areas have been studied in this research: 1) synthesizing novel monomers, e.g. chiral monomers, to produce new types of functionalized membranes for the biotechnology and pharmaceutical industries, 2) hydrophobic brush membranes for desalinating brackish water, sea water, and separating organics, 3) fundamental studies of water interactions at surfaces using sum frequency generation (SFG), and 4) discovering new surface chemistries that will control the growth and differentiation of stem cells. Using our high throughput screening method, we are able to evaluate hundreds of different surfaces in only a couple of days. This allows us to select for the best performing surfaces in a short period of time, for a wide variety of applications.

Chemical intermediates and organics, e.g. isobutanol, are currently in high demand due to their potential energy applications. We have developed hydrophobic brush membranes that were able to selectively separate valuable organics (isobutanol) from water, while rejecting other undesirable species, such as enzymes, using pervaporation (PV). These membranes (grafted from nanofiltration (NF) support membranes) had a selectivity ~1.5x higher than the current industrial standard, polydimethylsiloxane (PDMS), with α = 10.1 ± 0.9 for our brush membranes and α = 6.7 ± 0.1 for PDMS membranes. Since the mechanism of pervaporation is based on the solution diffusion (SD) model, it follows directly that these membranes may be used to desalinate water or fractionate gases since they are also based on the SD mechanism. Currently, all membranes used to desalinate water are hydrophilic. In this work, we have discovered that hydrophobic brush membranes are able to reject monovalent salt ions as well. This type of membrane is analogous to carbon nanotubes (CNTs), which are believed to have extremely high water fluxes through them due to near frictionless flow caused by a lack of hydrogen bonding. Our PV results and this theoretical model are the main motivation for this work. Using these brush membranes we were able to achieve 42% monovalent (NaCl) salt rejection of simulated seawater (32,000 ppm salt). These membranes are easier to scale-up than current composite membranes produced using interfacial polymerization. This is due to the fact that we are modifying the surface chemistry of NF membranes that are already being produced at the industrial scale.