UTFacultiesTNWMSTNewsPhD defense Erik Vriezekolk

PhD defense Erik Vriezekolk

On Friday February 5th 2016, Erik J. Vriezekolk will defend his PhD thesis entitled ‘All the Same: Isoporous Membranes for Water Purification’. In his thesis, Erik investigated three approaches to produce membranes with pores that are isoporous, or in other word, that are all the same. Isoporosity makes membranes much more efficient compared to current commercial membranes that suffer froma a high polydispersity. Additionally, Erik investigated methods to carefully tune the pore size of these isoporous membranes.

The defense will start at 12:45 in the Prof. Dr. G. Berkhoff Room (Collegezaal 4) in the Waaier building at the University of Twente. The introduction to the thesis will start at 12.30 in the same room.

Summary PhD Thesis:

In this thesis, the focus is on three approaches that allow fabrication of films and membranes that contain ordered and uniform pores with pore sizes in the ultrafiltration range. Special attention is given to the tuning of pore sizes by varying simple parameters during the fabrication process.

For the first approach, solvent exposure is studied as a simple method to decrease the dimensions of polymeric microsieves (Chapter 2). Microsieves, having very uniform and straight-through pores, can be fabricated via phase separation micromolding (PSμM), which is a technique that combines the principle of polymer phase inversion to fabricate membranes with the use of microstructured molds to fabricate replicates. Pore sizes of microsieves are usually in the range of micrometers due to limitations in the size of the molds, but need to be reduced to below 1 μm to make the microsieve attractive for other aqueous filtration applications such as ultrafiltration (UF). Polyethersulfone (PES)/polyvinylpyrrolidone (PVP) microsieves were fabricated with (perforated) pores of 2-8 μm in diameter and a very open internal structure. The pore sizes of the microsieves were then further reduced by solvent-shrinkage, where the microsieves were immersed in mixtures of acetone and N-methylpyrrolidone (NMP). Microsieves shrink because of swelling and weakening of the polymers, and subsequent collapse of the open internal structure. Size reduction in terms of pore size and porosity of the perforated pores was monitored over time. The pore size of microsieves was reduced from an initial pore diameter of 2.6 μm to only 0.2 μm. A higher NMP concentration leads to a higher shrinking rate. Shrinking typically occurs in two stages: first, a stage where both the perforated pores and periodicity shrink, and second, a stage where the perforated pores continue to shrink while the periodicity remains constant. For the microsieve to retain a high porosity, it is desired that both the perforated pores and periodicity shrink at similar rates, a process called isotropic shrinkage. The shrinking rate of the periodicity depends on the geometry of the microsieve and the structure of the polymer matrix. The maximum obtained isotropic shrinkage is ~35%, which is determined by the amount of voids in the polymer matrix.

The second approach focuses on the fabrication of composite membranes with a thin top layer based on self-assembling diblock copolymers (dBCPs) and homopolymers (Chapter 3 and 4). The block copolymers self-assemble into a morphology of hexagonally packed cylinders perpendicular to the surface, while the homopolymers reside in the core of the cylinders because of favorable interactions. Subsequent selective removal of the homopolymers leads to the formation of ordered, nanoporous layers. First, fundamentals of this approach are investigated, where layers are fabricated on model surfaces (Chapter 3). A system of polystyrene-block-poly(4-vinyl pyridine) (PS-b-P4VP) block copolymer, poly(4-vinyl pyridine) homopolymer and chloroform was used. The pore size and porosity can be tuned by varying the homopolymer content and molecular weight of the block copolymer. In this way, pore sizes were obtained between 10 and 50 nm. Uniformity of the pore size, however, is lost when the average pore size exceeded 30 nm because of macrophase separation. In a continued investigation, composite ultrafiltration membranes were fabricated by coating a thin nanoporous BCP layer on top of a support membrane (Chapter 4). A system of polystyrene-block-poly(ethylene oxide) (PS-b-PEO) block copolymer, poly(acrylic acid) (PAA) homopolymer dissolved in tetrahydrofuran (THF) was used. The pore size and morphology of the polymer layer could be controlled by varying the content of homopolymers, as in agreement with results in Chapter 3. The polymer concentration of the coating solution influenced both the morphology and the thickness of the layer. The different pore sizes and morphologies lead to membranes with different molecular weight cut-offs (MWCO) and permeabilities. The work thus demonstrates clearly that membranes with different performances can be fabricated using just a single type of BCP by simply varying the homopolymer content.

In the third approach, freestanding, asymmetric BCP membranes having a selective top layer were fabricated via dry-wet phase separation (Chapter 5). A mixture of PS-b-P4VP, volatile THF and non-volatile NMP was used as polymer solution. After casting a film, THF was allowed to evaporate for a very short duration (1 second). The still liquid film was then immersed in a coagulation bath where the polymer solidifies. The evaporation step increases the polymer concentration locally at the top of the film, which results in a very thin top layer. It also gives the BCP time to form cylindrical micelles that can lead to an ordered honeycomb structure formed by threadlike cylinders. Changing the ratio of THF/NMP and polymer concentration leads to membranes with different structures, hence, different performances. The best ordered honeycomb structures were obtained using a 18 wt% polymer solution with a THF/NMP ratio of 70/30 and a 21 wt% polymer solution with a THF/NMP ratio of 60/40. The THF/NMP ratio also influences the morphology of the support layer, which determines the permeability of the membrane (349-1320 L·m-2·h-1·bar-1). Filtration experiments with 30 and 10 nm silver nanoparticles and BSA showed that a sharper size cutoff is obtained when the pores are more ordered. This indicates that the ordered honeycomb-like pores indeed have a more narrow pore size distribution that allows for more selective filtration. In the general discussion, all three approaches were revisited, with a focus on the current limits and limitations of these techniques and with recommendations to address these.


For more information about this work, please contact Dr. Ir. Wiebe de Vos (w.m.devos@utwente.nl; phone: +31 (0)53 489 4495).