Date of Award

August 2021

Document Type


Degree Name

Doctor of Philosophy (PhD)


Chemical Engineering

Committee Member

Scott M. Husson

Committee Member

Christopher L. Kitchens

Committee Member

David A. Ladner

Committee Member

Sapna Sarupria


The main challenge of pressure-driven membrane operations is fouling, which refers to materials build-up on a membrane surface that increases the mass-transfer resistance of the membrane and, ultimately, the operating cost associated with higher-pressure operation and chemical cleaning. One way to combat fouling is the physical modification of a membrane surface by adding patterns of structural features that disrupt foulant deposition. My dissertation research goal was to develop the basic science needed to design new fouling-resistant membranes by understanding what factors most influence how foulants interact with patterned membrane surfaces.After introducing the basic methods used for membrane modification, Chapter 1 aims to provide an overview of important characterization methods that I have used to quantify membrane fouling. These methods were applied throughout my dissertation research to assess the effectiveness of the surface modification strategies that I explored. Chapter 2 reports findings from a systematic study to understand the roles of colloidal chemistry and membrane surface properties on membrane fouling using constant flux filtration. Commercial polyamide nanofiltration membranes were modified with a line-and-groove pattern using nanoimprint lithography. Threshold flux measurements were made for as-received and patterned membranes by the flux-stepping measurement method using solutions of silica nanoparticles with different surface chemistry as model foulants. A combined intermediate pore blocking and cake filtration model was applied to the experimental data to determine threshold flux values. Model fits were in excellent agreement with experimental data, indicating that it is an effective tool for determining threshold flux with a sparse data set. Patterned membranes generally exhibited 20–25% higher threshold flux than as-received membranes. Differences in Coulombic interactions and hydrophilicity between the foulants and membrane surface influenced fouling rates. Nevertheless, patterning influenced the threshold flux more significantly than differences in the surface chemistry of foulant particles. Chapter 3 identifies the effect of different pattern geometries on fouling behavior. Nanoscale line-and-groove patterns with different feature sizes were applied by thermal embossing on commercial nanofiltration membranes. Threshold flux values of as-received, pressed, and patterned membranes were determined using constant flux, cross-flow filtration experiments. The earlier derived combined intermediate pore blocking and cake filtration model was applied to the experimental data to determine threshold flux values. The threshold fluxes of all patterned membranes were higher than the as-received and pressed membranes. The pattern fraction ratio (PFR), defined as the quotient of line width and groove width, was used to analyze the relationship between threshold flux and pattern geometry quantitatively. Experimental work combined with computational fluid dynamics simulations performed by my collaborators showed that increasing the PFR leads to higher threshold flux. As the PFR increases, the percentage of vortex-forming area within the pattern grooves increases, and vortex-induced shielding increases. This study suggests that the PFR should be higher than 1 to produce patterned membranes with maximal threshold flux values. Knowledge generated in this part of my study can be applied to other feature types to design patterned membranes for improved control over colloidal fouling. Chapter 4 evaluates the effectiveness of a microscale herringbone pattern for reducing protein fouling on polyvinylidene fluoride (PVDF) ultrafiltration membranes using a combination of flux decline measurements and visualization experiments. Thermal embossing with woven mesh stamps was used for the first time to pattern membranes. Embossing process parameters were studied to identify conditions replicating the mesh patterns with high fidelity and to determine their effect on membrane permeability. Permeability increased or remained constant by patterning at low pressure (≤4.4 MPa) as a result of increased effective surface area; whereas, permeability decreased at higher pressures due to surface pore-sealing of the membrane active layer upon compression. Flux decline measurements with dilute protein solutions showed monotonic decreases over time, with lower rates for patterned membranes than as-received membranes. These data were analyzed by the Hermia model to follow the transient nature of fouling. Confocal laser scanning microscopy (CLSM) provided complementary, quantitative, spatiotemporal information about protein deposition on as-received and patterned membrane surfaces. CLSM provided a greater level of detail for the early (pre-monolayer) stage of fouling than could be deduced from flux decline measurements. Images show that the protein immediately started to accumulate rapidly on the membranes, likely due to favorable hydrophobic interactions between the PVDF and protein, followed by decreasing fouling rates with time as protein accumulated on the membrane surface. Knowledge generated in this part of my study can be used to design membranes that inhibit fouling or otherwise direct foulants to deposit selectively in regions that minimize loss of flux. Chapter 5 evaluates the effectiveness of different microscale herringbone patterns for reducing protein fouling on ultrafiltration membranes. Patterns with different geometries were introduced to membrane surfaces by embossing with different woven mesh fabrics. Having found earlier that CLSM can provide a greater level of detail for the early (pre-monolayer) stage of fouling, I used CLSM in situ to investigate the protein fouling profiles on as-received and patterned membranes. By staining the proteins and membranes with different fluorescent probes, this technique allowed the spatiotemporal imaging of protein accumulation at early-stages of fouling. CLSM images were compared with filtration data to uncover insights on the fouling mechanisms early to late stages of fouling and revealed the effect of pattern geometry on protein fouling. Extending the approach to other patterns and multicomponent solutions is expected to inform surface modification strategies used for the control of protein fouling. Chapter 6 provides concluding remarks and recommendations. Overall, my research generated new knowledge on the roles played by patterning and foulant chemistry on membrane performance. One important practical discovery that I made is that woven mesh fabrics can be used as inexpensive and widely available stamps, which may benefit commercial application of roll-to-roll patterning.



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.