Development of hollow fiber forward osmosis membranes
Due to the COVID-19 crisis measures the PhD defence of Mehrdad Mohammadifakhr will take place online without the presence of an audience.
The PhD defence can be followed by a live stream.
Mehrdad Mohammadifakhr is a PhD student in the research group Membrane Science & Technology (MST). His supervisor is prof.dr.ir. H.D.W. Roesink from the Faculty of Science and Technology.
Forward Osmosis (FO) is an emerging technique used to purify water or to concentrate valuable stream. Unlike the conventional membrane processes, FO utilizes an applied osmotic gradient to transport water through a semi-permeable membrane. While most available FO membranes are based on spiral-wound geometries, many applications of FO actually prefer a different geometry. Especially hollow fiber membrane geometries are of interest, as these types of membranes can cope with far more challenging feed streams. The research presented in this thesis aims at a better understanding of the role of these hollow fiber supports on FO performance.
Due to the challenges associated with IP-coating on hollow fibers, an alternative selective layer made from a polyelectrolyte complex (PEC) is manufactured. The complexation is applied on the inner surface of the hollow fiber simultaneously during fiber spinning (Chapter 2). For this, two oppositely charged polyelectrolytes are added separately to the bore and the dope solutions of the spinning, respectively. The purpose of this single-step approach is to circumvent the difficulties and time-consuming procedure affiliated to the IP post-coating. It is shown that a successful single-step PEC FO membrane can be made exhibiting adequate FO performance once a proper draw solute is selected. The choice of polymer and FO process orientation are found to be vital for the polyelectrolyte-based membranes. Due to the low rejection of the PEC membrane towards the most common draw solute (NaCl), it is concluded that the IP remains the most promising method for fabrication FO membranes when NaCl is used.
In Chapter 3, the possibility of increasing the reproducibility (success rate) of an IP‑coating on hollow fibers is investigated by the addition of a polyelectrolyte (PE) intermediate layer via Layer-by-Layer (LbL) assembly. We demonstrate that a sufficiently thick intermediate layer can provide a uniform and smooth layer which increases the reproducibility of the IP coating when compared to a fiber without an intermediate layer. The nascent membranes show an IP success rate of 72% for a PSS/PDADMAC intermediate layer and 90% for a PSS/PEI intermediate layer by the addition of 3.5 PE bilayers, as compared to 40% for the support without an intermediate layer.
Since the deposition of an LbL-based intermediate layer requires several time-consuming steps and is susceptible to variations, a single-step fabrication of the intermediate layer is investigated in Chapter 4, as an attempt to also increase the IP success rate. The single‑step approach can be advantageous in terms of reducing the intermediate layer formation time and a better adhesion of this layer to the support. Here, the IP success rate even is increased to higher values of 86% and 100% for PSS/PEI and PSS/PDADMAC, respectively, as compared to 29% for the support without an intermediate layer. While the NaCl rejection of the fibers with intermediate layer showed acceptable NaCl rejection values for FO, their water fluxes (0.28 to 0.32 L·m-2×h-1×bar-1) are found to be rather low.
It is known that the thickness of the support together with its porosity and tortuosity influences the ICP inside the support and thus the FO performance. Chapter 5 investigates the influence of the wall thickness on the FO performance. Here, a single‑step surface modification with only PEI is chosen instead of the complexation of two oppositely charged polyelectrolytes. This is done to acquire higher initial water permeances as compared to the polyelectrolyte complex based membranes, while still having a smooth and defect-free surface area to enable a successful IP coating. It is proven that the wall thickness of a hollow fiber significantly influences the FO water flux; the thinner the wall becomes the higher osmotic water flux is achieved, due to the reduction of the internal concentration polarization. Additionally, it is shown that there is a correlation between the fiber dimensions, polymer tensile strength and the burst pressure of the nascent fiber. This correlation is helpful to choose the most suitable polymer to spin thin fibers.
Chapter 6 combines all our findings from previous chapters into one. The implementation of a single-step modification of the surface with PEI has shown to be a good method to improve the IP-coating. In this chapter, this approach is coupled with opening-up the support to further decrease the ICP. This is done by substantially reducing the polymer content by 3 wt% compared to the dope composition used in our previous chapter. The permeability of the newly spun hollow fiber support is increased to 365 ± 64 (L·m‑2×h‑1×bar‑1) from the very low value of 48 ± 2 (L·m-2×h-1×bar-1). These highly permeable, inner surface PEI modified fibers possess a high NaCl rejection after IP coating, indicating a successful coating. The 3 wt% reduction results in a 4-fold improvement in FO water flux. It is shown that a high permeability and an open structure of the hollow fiber support, in addition to a low wall thickness, results in significant improvement in the FO performance of the membrane.
This research highlights that IP-coating is very challenging for hollow fibers whose success rate is pretty much dependent on the physicochemical properties of the hollow fiber (inner) surface.