Desalination Fibrous Membranes
| Accessible fresh water is a global concern that is expected to worsen as the impact of global warming on our planet becomes more severe in the next decades. With a modest amount of energy, direct-contact membrane distillation (DCMD) can produce freshwater by creating a temperature difference between the feed (warm salty water) and permeate (cold purified water) separated from one another by a hydrophobic membrane. The temperature difference across the membrane creates a vapor pressure gradient that serves as the driving mechanism for the transport of water vapor from the feed to the permeate through the pores of the membrane (see the figure below). Mass transfer through a DCMD membrane is directly related to fresh water production rate and should be maximized, while heat transfer should be minimized to maintain a temperature difference across the membrane. |
 Schematic of a direct contact membrane distillation (DCMD) is shown in the figure on the left. Flow on both sides of an DCMD membrane is shown schematically in the figure on the right [Ullah et al., 2018]. |
| While majority of existing membrane are produced as porous films, electrospinning is also becoming a method of membrane production for its unique microstructural attributes. |
  The figure shows a schematic drawing for the electrospinning unit used to produce polystyrene membranes (the SEM image) used for desalination experiment [Esteves et. al., 2020 and Abdelrazeq et. al., 2020]. |
| We have recently developed a novel numerical framework in which a realistic model of the membrane’s microstructure is generated and used to simulate the DCMD process. Virtual membranes resembling electrospun structures were generated using the Discrete Element Method (DEM). An example of our virtual membranes compressed to 65% is shown below. |
|
|
| An in-house Pore Morphology Method (PMM) code was also developed to predict the shape and location of the Air-Water Interface (AWI) on the cold and hot sides of the membrane and to evaluate its Liquid Entry Pressure (LEP). The figure below shows an example results from our PMM simulation of warm (red) and cold (cyan) air-water interfaces under an intrusion pressure of 142.2 kPa across a virtual membrane. The magnified figures show the warm and cold interfaces from the inside the membrane. The figure on the right shows the location where the two interfaces have come into contact with one another leading to membrane failure [Gildeh et. al., 2026]. |
 |
| The Navier-Stokes and species equations were then solved to predict the rate of vapor transport between these AWIs. Using such realistic geometries in the simulations allowed us to examine the counteracting impacts of membrane thickness and porosity on LEP, air dissolution, and vapor transport. Our simulations also quantified the negative impact of temperature polarization on freshwater production rate. Contours of fiber temperature is shown in the figure below for one of our virtual membrane in (a). Vapor pressure field inside the membrane is shown in (b) along with velocity vectors illustrating vapor transport from the hot air-water interface to the cold air-water interface. The relative humidity contours across the computational domain are shown in (c). |
 |
| The computational framework developed in this work enables the design and optimization of fibrous DCMD membranes for highest transmembrane vapor flux while preventing membrane flooding. An example of AWIs over a membrane colored with temperature is shown in the figure below. The temperature polarization effect can be seen in the AWI temperature varying in the in-plane direction. The colder regions on the warm AWI shows locations where the AWI sagged deeper into the membrane. Likewise, warm regions on the cold AWI shows locations where it penetrated deeper into the membrane. |
 |
| This classical approach for modeling the performance of an DCMD membrane relies on predetermined empirical factors/relationships that compensate for the inaccuracies in the existing models as they fail to incorporate the microstructure of the membranes. To overcome these limitations we have considered a new modeling approach in which the membrane microstructure is used in the calculations. Such a computational tool can help one to predict the rate of heat and mass transport through a DCMD membrane from first principles rather than trial-and-error. Such simulations can be used to study the importance of membranes’ microstructure without the need for empirical correction factors. More specifically, the effects of fiber diameter, membrane porosity, and membrane thickness are simulated. |
 Molecular Dynamics simulation of direct contact membrane distillation(in of a simplified fibrous membrane) showing the evaporation of water molecules from the warm feed side (the lower side) toward the cold permeate side (the upper side). Note the condensation on the permeate side [Hemeda et. al., 2018]. |
Representative Publications:
|
Acknowledgement:
|
