Permeation studies, with hollow-fiber

The experimental procedure for permeation studies with hollow fiber modules was Identical to that described by Bhave and Slrkar (12) for flat Celgard films except that the flat film test cell was replaced by the hollow fiber module (see Figure 1 in reference (12)). 

Many researchers have assessed the effect of pulsatile flow on different membrane processes with wide range of feeds. One of the first studies was by Kennedy et al. [48] who showed that flux in the RO of sucrose solution could increase by 70% by pulsatile flow at 1 Hz. Gupta et al. [49] reported a 45% enhancement of flux in MF of raw apple juice with a pressure waveform provided by a fast piston return followed by a fast forward stroke at 1 Hz. Jaffrin [50], using hollow fiber filters, demonstrated a 45% enhancement in flux in plasma filtration. Using the collapsible-tube oscillation generator described above, Bertram et al. [47] demonstrated that pulsation resulted in a 60% increase in permeate flux in the filtration of silica suspensions. 

Experimental. The hollow fiber membranes used for this study were Naflon 811, which Is a copolymer of polysulfonyl fluoride vinyl ether and polytetrafiuoroethylene, and sulfonated and/or quaternated derivatives of polyethylene (kindly supplied by Dr. E. Korngold from Ben Gurlon University In Israel). The aqueous alcohol solutions studied thus far are those of methanol, ethanol and 2-propanol. The separations were accomplished via the pervaporatlon process as described In Reference 9. Counter Ions were replaced In the hollow fiber by soaking the permeator for twenty four hours In 1 molar solutions of the pertinent Ions. For example, experiments were conducted with Na as a counter Ion. When this set of experiments was finished, the sodium was exchanged by Ll etc. Each data point shown In Figure 14 consists of 6 to 10 measurements taken over a time period of 8 hours. Re-runs with the various counter Ions proved that the intrinsic properties of the membrane remain unchanged and the permeability measurements are reproducible. 

An alternative approach to solving stability problems with ILMs is presented by Bhave and Sirkar (114). Aqueous solutions are immobilized in the pore structure of hydophoblc, polypropylene hollow fibers by a solvent exchange procedure. Gas permeation studies are reported at pressures up to 733 kPa with the high pressure feed both on the shell and lumen sides of the laboratory scale hollow fiber permeator. No deformation of the hollow fibers is observed. Immobilizing a 30 weight % KjCO, solution in the hollow fibers greatly improved the separation factor, a(C02/Na). from 35.78 with pure water to 150.9 by a facilitated transport mechanism. Performance comparisons with commercial COj separation membranes are made. 

Immobilized Liquid Membranes. A pilot plant study of the recovery of ethylene and propylene from a polypropylene reactor off-gas stream was presented by Hughes et al. (23). Aqueous solutions of Ag Ion were Immobilized In the pore structure of commercial reverse osmosis hollow fiber modules. The pilot plant operated at feed pressures of 414-827 kPa, feed flow rates of 1.42-4.25 m /h at STP, and sweep flow rates of 3.79 10" - 0.114 m /h hexane. Permeate streams with propylene concentrations In excess of 98 mole % were observed in pilot plant operation with modules containing 22.3 to 37.2 m membrane area. 

The data of Table III (Set A) also demonstrate the high pressure capability of Celgard X-10 hollow fibers. The CO2-N2 mixture separation with aqueous 30% wt/wt 200 ILM-s in Celgard X-10 fibers was studied at applied AP values up to about 550 cm Hg without any significant effect on permeation due to any fiber deformation. 

The membrane model in the implemented ACM module was first validated with the experimental results of Pan (1986). In the experiments, the effect of varying stage cut (that is, the fraction of feed that permeates through the membrane) on mole fraction of H2 in the permeate was studied using asymmetric hollow fiber membranes in co- and counter-current configurations. Parameters of the membrane module can be found in Table lO.A.l stage cut varies from 0.34 to 0.55 in the counter-current configuration and from 0.35 to 0.6 in the... 

The developed membrane model/ACM module was also validated against the results of Davis (2002). In his study, a mathematical model of a hollow-fiber membrane was developed in Aspen HYSYS for air separation. Parameters used in the simulation of this separation are given in Table 10.A.2. As can be seen in Table 10.A.3, the ACM model predictions are very close to the simulation results of Davis (2002), with a maximum difference of 0.41% in permeate O2 concentration. 

Wang D, Tong J, Xu H, Matsamura Y (2004) Preparation of palladium membrane over porous stainless steel tube modified with zirconium oxide. Catal Today 93-95 689-693 Nair BKR, Harold MP (2007) Pd encapsulated and nanopore hollow fiber membranes synthesis and permeation studies. J Memb Sci 290 182-195... 

Asymmetric carbon membranes were made by carbonization of asymmetric PI hollow fiber membranes by Tanihara et al. [43], In their study, it was found that the permeation properties of caibon membranes were hardly affected by feed pressure and exposure to toluene vapor. Furthermore, there was only little change in the permeation properties of the caibon membrane with the passage of time. 

In Section 5.4.4, we studied a variety of chemical reaction facilitated separation where the reaction was taking place in a thin liquid layer acting as the liquid membrane Figure 5.4.4 illustrated a variety of liquid membrane permeation mechnisms. Here we will identify first the structural configuration of the liquid membranes as they are used in separators with countercurrent flow pattern (as well as for the cocurrent flow pattern). There are three general classes of liquid membrane structures emulsion liquid membrane (ELM) supported liquid membrane (SLM) or immobilized liquid membrane (ILM) hollow fiber contained liquid membrane (HFCLM). Each will be described very briefly. 

---