Membrane research worldwide is still concerned with the development of new techniques and the comprehension of the phenomena in membrane formation, due to the difficulties to obtain membranes with the desired properties, i.e., ultra-thin and defect-free dense skin. Different methods of polymer membrane preparation have been covered in several reviews. Dense homogeneous polymer membranes are usually prepared from solution by solvent evaporation only or by extrusion of the melted polymer. There may be a number of reasons for polymer to be chosen, e.g., thermal, chemical and solvent stability, price, etc. but the most important one is that membranes prepared from these polymers can separate solute from solvent. Ultrafiltration (UF) is a process of separating extremely small particles and dissolved macromolecules from fluids using asymmetric membranes of surface pore size in the range of 50 to 1 nm and often operated in a tangential flow mode where the feed stream sweeps tangentially across the upstream surface of membranes as filtration occurs, thereby maximizing flux rates and membrane life. It imposes specific requirements on the membrane material and membrane structure and the efficiency of UF is determined by the porosity and the pore size of the membrane.
Polyethersulfone (PES) polymer is well used for membrane fabrication as it
produces high performance asymmetric membrane. Aromatic polysulfone family of
polymers is extensively used due to their wide temperature, pH and chlorine
tolerance. PES consists of phenylene ring structures connected together with
sulfone groups (SO2) or ether linkages (-O-) in the backbone chain
to form a polymer. The sulfone groups tend to make the polymer stiff with a
high glass transition temperature and together with the ring structures, it
makes the polymer chemically resistant and relatively hydrophobic (Kesting,
1985). For better membrane properties, a third component as additive can
be dissolve in the casting solution (Barth et al.,
2002). These additives affect membrane in developing spongy structure and
prevent the formation of macrovoid as it enhances pore formation and hydrophilicity
of membrane (Katarzyna, 1989; Wienk
et al., 1996). There are many kinds of additives including polymers,
nonsolvents and inorganic salts. Inorganic salts were found effective for the
fabrication of membrane having an appropriate structure and high performance
(Kim et al., 1999). Although, the membrane preparation
techniques are well known, the precise membrane casting procedure outlining
choice of solvent, additive, concentration and other relevant details are not
available for several polymer candidates.
Less fouling behavior was found in modified polyethersulfone membranes and
it has a wide range of retentate pH values, giving more protection (Wienk
et al., 1996). The water permeability, salt permeability and water
regain studies of sulfonated PES membranes have been studied (Brousse
et al., 1976) and constant research are carried out by applying polymer,
nonsolvent and inorganic salt as additives in casting dope to study the performance
and structure improvement. Inorganic salt additive was found to be more effective
for membrane performance and structure improvement (Kesting,
1985; Kraus et al., 1976) as it affects the
thermodynamic/kinetic properties of the membrane-forming system, thus resulting
in changes in the membrane structure and performance.
Several commonly used additives include low-molecular-weight inorganic salts such as lithium chloride (LiCl), zinc chloride (ZnCl2), magnesium chloride (MgCl2), calcium chloride (CaCl2), magnesium perchlorate, (MgClCO4) and calcium perchlorate (CaClCO4). However no work has been reported regarding the use of LiF additive in PES membranes. Thus the objective of this study is to investigate the influence of LiF on separation performance of PES membranes.
MATERIALS AND METHODS
Commercial grade polyethersulfone (PES) in resin form was supplied by BASF. The solvent N, N-dimethylformamide (DMF: HCON (CH3)2/Mw; 80.14 g mol-1, 99.8%) was purchased from Labscan Asia Co. Ltd and used without further purification. Inorganic salt additive lithium fluoride anhydrous (LiFH2O) 99.7% (mol. wt. 12) was obtained from Fischer Scientific Chemicals. Tap water was used as the coagulation bath. For UF experiments, PEG with various molecular weights (PEG 600, PEG 1000, PEG 3000, PEG 6000 and PEG 10,000) were obtained from Fluka.
Dope solution preparation: PES and LiFH2O were dried in microwave for 10 min at medium to high pulse before dissolving them in DMF. The 500 mL polymer solution consists of 20% PES in various concentrations of DMF and LiFH2O as shown in Table 1. The dope solution was prepared in the modified microwave at low to high pulse.
|| Dope solution compositions
Temperature was kept at 90-95°C. Temperature was monitored by a thermocouple.
Membrane casting: The membranes are prepared by phase inversion method. The dope solution thus obtained was spread over a smooth glass plate with the help of a casting knife. The thickness of the membranes was 200 μm. The casted polymer film was then immersed in a tap water at room temperature, where exchange between the solvent and water is induced. It was then transferred to another container containing distilled water. All membranes were inspected for defects and good areas were chosen for evaluation.
Viscosity measurements: The average apparent viscosity of the dope solutions were measured with Brookfield Digital Rheometer (model DV-III ultra, USA) equipped with a suitable-sample adaptor (SC4-31). The viscosity for dope solutions was measured at room temperature. The spindle was SC4-31 type.
Determination of permeation flux and solutes rejection: Pure water permeation fluxes (PWP) and solutes water permeation fluxes (PR) of membranes were obtained as follows:
where, J is the permeation flux of membrane for PEG solution (L/m/h) or pure
water and Q is the volumetric flow rate of permeate solution. Δt is the
permeation time (h) and A is the membrane surface area (m2). Solutes
rejection of membranes was evaluated with various molecular weight of PEG solutions
ranging from 600 to 40,000 Da at 4.5 bar. The concentration of PEG solution
used is 500 ppm. The concentration of the feed and permeate solution were determined
by the method described elsewhere (Sabde et al.,
The membrane rejection (SR) is defined as:
where, Cf and Cp are the polyethylene glycol concentration in the feed solution and permeate solution, respectively. The concentration of PEG was determined based on absorbency in a UV-spectrophotometer at a wavelength of 535 nm.
Pore and pore size distribution: The pore size of PES membrane produced
was determined using transport data (Sabde et al.,
1997). Solute diameter (ds) is given by:
where a is the Stokes radius of polyethylene glycol, with a function of molecular weight, M. This is given by:
Stokes radius of a macromolecule can be obtained from its diffusivity in a
solution using Stokes-Einstein equation as reported by Singh
et al. (1998). The Stokes radius was explained as radius of hypothetical
sphere that would diffuse with the same (Ani et al.,
2007) speed as the particle under study (Park et
al., 2000). To determine the mean pore size (μp) and
standard deviation (σp) of the membranes, the data of solute
separation versus solute diameter that formed solute separation curve was plotted
on log normal graph. The mean pore size was calculated with ds, corresponding
to R = 50% on the linear regression line. The standard deviation was calculated
from the ratio ds at R = 84.13% and at 50%. Moreover, MWCO can be
measured from the regression line at R = 90%.
RESULTS AND DISCUSSION
Performance of the membranes: Table 2 observed that membranes produced from dope solution containing LiF exhibits high pure water permeation (PWP) compared to those prepared without additive. Membranes containing 4% LiF exhibits highest PWP. In general, the PWP for membranes having 3-5% LiF has approximately 50% higher than the 1% LiF.
Table 3 indicates that membranes with LiF showed both higher
permeate rate and rejection compared to that without LiF. However, membrane
with 5% additive has low rejection rates with MWCO greater than 40 kDa although
it has the highest permeate rate. Improved permeation flux of the membranes
has confirmed that the hydrophilic property of the membrane has improved by
the presence of LiF. The presence of LiF has helped improved both the PR and
rejection rates of the membranes but its concentration is best kept to 3%. Increasing
LiF concentration to above 3% will not only increase the permeation rates but
decrease its rejection rates.
|| Pure water permeation of LiF membranes
|| Permeation rate and rejection rate of LiF membranes
|| Solute separation curve
Mean pore size and pore size distribution: From Fig. 1,
the mean pore size (μp), standard deviation (σp)
and MWCO were calculated. Mean pore size is define as pore diameter when solute
separation is 50% (Ani et al., 2007). Results
tabulated in Table 4 showed that the membrane without additive
has 35 kDa MWCO and 3.75 nm mean pore size. As for 1, 2 and 3% LiF additive
membranes, results found membranes has 8.1, 7.1 and 8.45 kDa MWCO and the mean
pore size were 0.884, 0.649 and 1.043 nm, respectively.
||Geometric mean pore size (μp) and genomic
standard deviation (σp) for various membranes
These values displayed a linear relationship between the mean pore size and
MWCO. Smaller pore size of 1 and 2% LiF membrane contributed to low flux although
the rejection rate is high. While 4% LiF membrane has high flux but less rejection.
The pore sizes obtained in Table 4 further explains the performance
of the membranes. Increasing LiF concentration to more than 3% will not reduce
the pore size of membranes. Thus the best concentration of LiF that should be
used is 3% as the membranes produced has small pore sizes which displays high
rejection rate and permeate rates.
Membranes produced from dope solutions containing LiF are superior in terms of permeation flux rates, rejection rates and quality of membranes compared to those prepared without additive. The addition of LiF has a significant effect on solution properties and resulted in high permeation rate which implies the membrane porosity has increased. It can be concluded that LiF additive with formulation 1-3% exhibited the best rejection rates and permeation rate with MWCO of 8.9 kDa.
Financial support from the Ministry of Science, Technology and Environment through the IRPA funding vote No. is gratefully acknowledged.