Noel Jacob Kaleekkal,A.Thanigaivelan,M.Tarun,D.Mohan*

Membrane Laboratory,Department of Chemical Engineering,Anna University,Chennai,India

Keywords:Sulfonated polyethersulfone Hydrophilicity Diffusive permeability Biocompatibility Molecular weight cut-off

ABSTRACT In this work,we evaluate the properties of solution casted polysulfone(PSf)/sulfonated polyethersulfone(SPES)blend membranes prepared by non-solvent induced phase inversion technique.The morphologies of these blend membranes,observed using scanning electron microscopy(SEM)and atomic force microscopy(AFM)imaging,indicated a smoother skin layer and an increased number of highly interconnected pores in the sub layer.The efficacy of the prepared membranes was evaluated in terms of porosity,ultra filtration rate(UFR),molecular weight cut-off(MWCO)and mean pore size.The hydrophilicity of these membranes was in consonance with contact angle values.It was observed that the selectivity and the UFR of the blend membranes were higher when compared to pristine membranes.Furthermore,these blend membranes demonstrated an increase in biocompatibility—prolonged blood clotting time,suppressed platelet adhesion,reduced protein adsorption and lower complement activation.These membranes were also investigated for uremic solute removal.Diffusive permeability of middle molecular weight cytochrome-c revealed an increase from 8× 10-4 cm·s-1 to 18×10-4 cm·s-1 and illustrates the possibility that these sulfonated PES/PSf blend membranes can be used to prepare membrane modules for hemodialysis applications.

1.Introduction

Chronic kidney diseases(CKDs)are widespread and on the rise worldwide.Globally,the number of reported cases of end stage renal disease(ESRD)patients is estimated to be three million with nearly a 7%growth rate at the end of 2012[1].The use of membrane technology forhemodialysis has been proven vital.These membranes act as a semipermeable barrier which permits the diffusion of uremic toxins like urea,creatinine etc.from blood to dialysate without the loss of important blood proteins such as albumin, fibrinogen etc.[2].Worldwide statistics of ESRD imply that the demand for hemodialysis membrane is expected to continuously increase,and dialysis treatment will become a multi-million dollar industry[3].Among the polymeric membrane materials used in hemodialysis,polysulfone and PSf based membranes are commercially available and are most-widely used[4]because of its thermal stability,mechanical rigidity and chemical inertness.Moreover it is one of the few biomaterials that can withstand all sterilization techniques viz.steam,ethylene oxide,gamma radiation[5].Another key feature of the PSf membranes are that they hardly swell or shrink,and are found to have a water absorption rate of 0.30 wt.%per 24 h as determined by the ASTM D570 method[6].Further,PSf has an excellent film forming capacity that allows for the easy preparation of flat-sheet and hollow fiber membranes by immersion precipitation techniques[7].

However,PSf membranes are inherently hydrophobic and when in contactwith blood rapidly adsorbs proteins.This apartfrom the adverse effects on the body leads to an irreversible fouling of both membrane surface and the internal pores,thus reducing the permeation flux and selectivity[8].The hydrophobic–hydrophilic interaction between membrane surface and the organics in feed solution can explain this irreversible fouling[9].Increasing the biocompatibility of the hemodialysis membranes has been the subject of intense investigation for the past decade.Several earlier attempts have been made to increase the biocompatibility of polymeric membranes by increasing the hydrophilicity of the membrane surfaces.

Krueger et al.employed polyvinylpyrrolidone as a hydrophilic modifier to polysulfone-based membranes to increase the hydrophilicity and hence the biocompatibility of the dialysis membranes[10].The N-methyl-D-glucamine grafted PSf membranes prepared by Shi et al.were observed to have a hydrophilic top-surface.These membranes demonstrated a more open porous structure,higher permeation flux and better anti-fouling properties than the PSf control membrane[11].Mahlicli et al.concluded that blending small amounts of heparin with alginate significantly decreased protein adsorption while prolonging the blood clotting time[12].Yue et al.suggested that grafting of poly(sulfobetaine methacrylate)onto polysulfone membranes improved the biocompatibility of the PSf membranes,which then were characterized by protein adsorption,hemolysis assay,platelet adhesion,plasma recalcification time,activated partial thromboplastin time(APTT),thrombin time(TT)and cytotoxicity experiments.All of their results indicated that the modified PSf membrane had better blood compatibility and cytocompatibility[13].Wang et al.developed a SPES blend membrane which effectively reduced the BSA adsorption and prolonged the blood clotting time,thereby increasing the blood compatibility[14].Takashi et al.obtained a proteinadsorption-resistant membrane for hemodialysis,by blending 2-methacryloyloxyethyl phosphorylcholine(MPC)polymer with PSf[15].

However,the drawbacksofthese methods,such as low modification efficiency,complex procedures,high cost,deterioration of mechanical strength,and damages of porous structure are pronounced.Therefore,a simpler and a more cost effective route to introduce hydrophilic moieties within the membrane is imperative.

One approach to improve the hydrophilicity without compromising any characteristics of the membrane is functionalization of the polymer backbone.Sulfonation is a versatile modification of this method which increases hydrophilicity as well as other membrane properties such as water flux and permeability[16].

In our study,PES was sulfonated,and the resultant hydrophilic SPES was blended with the base polymer PSf.This blending of polymers is known to be an extremely attractive and commercially viable method of obtaining new structural materials with improved hydrophilicity,enhanced biocompatibility and variant morphology[17].The prepared membranes were evaluated for performance as well as biocompatibility.Initially the casting conditions were optimized for the preparation of the blend PSf/SPES membranes,followed by characterizations which involved porosity measurement and ultra filtration rate to ensure reproducibility of the membranes.The MWCO and average pore size of the prepared membranes were investigated to ensure that they meet the criteria for hemodialysis membranes.The cross sectional and top surface morphology were obtained using SEM and AFM respectively.The contact angle of a water droplet on the membrane surface is useful in determining the wettability assay and as a corollary,the work of adhesion.Parameters like protein adsorption and platelet adhesion onto the membrane samples were evaluated in order to ensure their biocompatibility.Freshly drawn blood samples were used to determine the Blood coagulation Time and Complement Activation when the membranes were in contact with blood.The membrane efficiency and rejection of uremic toxins were studied using a single stage dialysis module.An exploratory attempt has been made to determine the influence of SPES,in the casting solution,on performance and biocompatibility of the membranes.

2.Materials and Methodology

2.1.Materials

Commercial grade polymers polysulfone(Udel P-3500)and polyethersulfone(Gafone?-3000)having molecular weights of 77 kDa and 58 kDa respectively were procured from Amoco Chemical Inc.US and used after being dried in a vacuum oven at 80°C for 48 h.The solvent N-methyl-2-pyrrolidone(NMP)from SRL Chemicals Ltd.,India was sieved through molecular sieves(Type-0.4 nm)to remove moisture and stored in dry conditions prior to use.Dichloromethane and chlorosulfonic acid from SRL chemicals were used as sulfonating reagents.Sodium lauryl sulfate(Glaxo India Ltd.)and polyethylene glycol-400(Merck Co.,)were used as surfactants and additives.A series of PEGs from Merck and CDH were used to evaluate the MWCO of the membranes.Urea(MW-60.06 Da),creatinine(MW 113.12 Da),cytochrome-c(MW 11,800 Da)and bovine serum albumin(MW 64 kDa)purchased from Himedia Laboratories India(Pvt.)Ltd were used in the solute diffusion studies.Urea nitrogen(Diacetyl)and creatinine(Direct)reagents from Erba Mannheim were used for the determination of the corresponding solute concentration.

2.2.Sulfonation of PES

10 g of PES was taken in a three neck,round bottom flask and dissolved in 80 ml of dichloromethane(DCM)by continuous stirring at room temperature till a homogenous solution was obtained.A mixture of 3 ml chlorosulfonic acid and 25 ml DCM was added drop wise to it,in about 90 min with continuous stirring[18].The nitrogen atmosphere was maintained over the reaction period to remove HCl effluents formed during the substitution reaction.The mixture was further stirred for 150 min at ambient conditions at 400 r·min-1,after which it was transferred into a separating funnel and held for approximately 10 min.The lower layer of the solution was precipitated into cold water under constant stirring.The precipitate was then filtered,washed and dried for 2 days at 60°C under vacuum.

The ATR-FTIR spectra of pure PES and SPES were recorded using a spectrometer(Thermo Nicolet,Avatar 370)at room temperature.The samples were dried before FTIR measurement and spectrum was recorded in the region up to 4000 cm-1.

2.3.Membrane Preparation

Dope solutions were prepared by dissolving the polymers(PSf and SPES)in different compositions(Table 1)in the presence of an additive PEG-400 using NMP as the solvent.The solution was homogenized using constant mechanical stirring for 4 h at 40°C.The air bubbles in the blend solution were eliminated by introducing the solution in an ultra-sonication bath.The homogenous solution was,then,casted on a glass plate using a doctor's blade maintaining a thickness of(0.20±0.02)mm.The cast film on the glass plate was immediately immersed in a gelation bath containing 2%(V/V)NMP(solvent)and 0.2%(by mass)SLS(surfactant)in distilled water(non-solvent).Strict conditions were maintained during casting and gelation to ensure membranes with improved physical properties such as homogeneity,thickness and morphology were formed[19].The nascent membranes were left overnight in the gelation bath and then washed thoroughly with warm distilled water to remove residual solvent.These prepared membrane sheets were stored in distilled water containing 1%(by mass)sodium azide to prevent any bacterial contamination.

Table 1 Polymer dope solution composition and casting conditions of the membranes

2.4.Porosity

The porosity of the membrane is calculated by assuming that the skin layer of the membrane is solely effective in separation.Porosity is calculated by a gravimetric method as shown below[20]

where ε is the porosity of the membrane,Wwis the wet membrane mass and Wdis the dry membrane mass(g),Dwand Dmare the densities of the water and membrane(g·cm-3)respectively.

2.5.Ultra filtration rate

A batch type,dead end UF kit(Model 8400,Amicon,USA)was used in the determination of UFR of the prepared PSf/SPES membranes.The schematic representation of this experimental set-up has been given in an earlierpaper[21].The effective membrane area available forultrafiltration experiments is 38.5 cm2.The UFR of pure water through the PSf and PSf/SPES membranes was measured at an operating pressure of 33.331 kPa

where V is the quantity of permeate collected in(ml),A is membrane surface area(m2),P is the operating pressure(kPa),and t is the time for which the permeate is collected(h).

2.6.Molecular weight cut off and mean pore size

Molecular weight cut-off(MWCO)of all the prepared membranes was determined by the ultra filtration of polyethylene glycols(PEGs)of increasing molecular weights.This method provides a very simple technique for indicating the performance of a given membrane.A series of PEGs–4,6,8,10,12,15,18,20 and 35 kDa–were used for solute rejection studies and consequently for the estimation of MWCO.Standard solutions were prepared by dissolving the PEGs in deionized water to obtain solutions having a required concentration of 0.5 mg·ml-1.The UF test cell was filled with these solutions and a constant pressure of 33.331 kPa was applied throughout the experiment.The concentrations of feed and permeate were determined by a total organic carbon analyzer(Shimadzu,TOC-V CPH).The percentage solute rejection(SR,%)was calculated from the concentration of the feed(Cf)and the concentrate of the permeate(Cp)by the following equation[22].

The average Stokes radii of the membrane pores were calculated based on molecular weight cut-offs[23].

Further the mean pore size can be calculated by the following expression.

where r is the average Stokes radius of the membrane(m)and M is the molecular weight of the PEG having rejection greater than 90%.

2.7.Scanning electron microscopy(SEM)analysis

The cross sectional morphology of the prepared PSf and PSf/SPES blend membranes was obtained by the scanning electron microscope(FEIQuanta-400 FEG).The membranes were cut and dried using solvent exchange method.These dried membrane pieces were frozen in liquid nitrogen for 60 s.The samples were then fractured and sputtered with gold for the improvement of electrical conductivity.The images were then captured in very high vacuum conditions operating at 20 kV.

2.8.Atomic force microscopy

The surface morphology of the pure PSf and PSf/SPES blend membranes were investigated using a non-contact type atomic force microscope(Park System,Korea model no.XE-100).A 1 cm2sample of the prepared membranes was affixed onto the substrate followed by imaging of the surface.The parameters thatindicate the surface roughness of membranes such as mean roughness(Ra),the mean value of the surface relative to the center plane,the root mean square of the Z data(Rq)and the mean difference between the highest peaks and lowest valleys(Rz)were also calculated.For increased accuracy the measurements were carried out at three different locations on the membrane surface and their average value was reported.

2.9.Wettability assay and work of adhesion

The wettability assay of the pristine PSf and PSf/SPES blend membranes was measured by a sessile drop method using a goniometer(GBX Instruments,Germany).Milli-Q(5 μl)was deposited by a microsyringe as a sessile drop onto the loaded membrane surface.Instant readings were captured with the help of a high frame rate camera.The value reported is the average of five measurements[24].The work of adhesion(W)explains the interactive forces between liquid and solid surfaces.This was calculated using the Young–Dupree equation

where,γLis the surface tension of the water(7.2 × 10-2N·m-1)and θ is the contact angle.

2.10.Biocompatibility studies

2.10.1.Protein adsorption test

Bovine serum albumin was selected as the test protein for this study.Phosphate buffer saline solution containing 4.5 mg·ml-1of BSA was prepared while maintaining a pH of 7.2.One square centimeter samples of all membranes immersed in 10 ml of prepared protein solution were incubated at 37 °C by continuous shaking for 2 h at 100 r·min-1in a 24 well culture plate.At the end of the incubation period the unbound proteins from membrane surfaces were removed by repeatedly rinsing with PBS buffer,followed by elution of the adsorbed proteins from the membrane with 1%(by mass)sodium dodecyl sulfate(SDS).The concentration of adsorbed protein present in the SDS solution was quantified by bicinchoninic acid(BCA)method[25].

2.10.2.Platelet adhesion

To evaluate the adhesion of platelets,equal sized membrane samples were incubated with 2 ml of human platelet rich plasma(PRP)in a 24 well plate for two hours at 37°C.The platelet-adhered membranes were equilibrated in 1 ml Triton-100 solution for half an hour following which the suspension was aspirated from each well.The on-adherent platelets were,then,rinsed away by filling and aspirating the wells five times with PBS.Platelet adhesion levels were quantified using a lactate dehydrogenase(LDH)assay[26].

2.10.3.Blood coagulation time

10 ml of fresh blood was drawn into a citric acid lined tube to prevent coagulation.This blood sample was initially centrifuged at 100 rpm for half an hour at 4°C to obtain platelet-rich plasma(PRP)as supernatant.A further centrifugation of the PRP at 2000 r·min-1for 20 min at 4°C yielded platelet poor plasma(PPP)as supernatant.A one cm2sample of all the prepared membranes was incubated individually in 0.5 ml of PPP at 37°C for 1 h[27].The activated partial thrombin time(APTT),prothrombin time(PT),of reacted PPP was determined by an automated blood coagulation analyzer(Sysmex CA 1500).

2.10.4.Complement activation

20 ml human blood was drawn from healthy adult volunteers without the addition of any anticoagulant.The collected blood was incubated for 1 h at 37°C.The resulting clotted blood was then centrifuged at 2000 r·min-1for 20 min to separate the serum.Membranes(1×1 cm2)were immersed in 2 ml of serum and further incubated at 37°C for 1 h.The concentrations of complement components C3 and C5 left in the serum after incubation with membranes were determined with immunoturbidimetry and Nephelometry assay[28].

2.11.Permeability studies

The experiments were carried out to investigate the diffusive solute permeability for the prepared membranes.The dialysis experiment was run for 3 h at 37°C in a single stage counter current membrane module.The solute reservoir was filled with a mixture of solutes containing urea(1 mg·ml-1,molecular weight:60 Da),creatinine(0.1 mg·ml-1,molecular weight:113 Da),cytochrome-c(0.05 mg·ml-1,molecular weight 11800 Da)and bovine serum albumin(1 mg·ml-1,molecular weight:64000 Da)in order to mimic the actual conditions of dialysis treatment.The dialysate reservoir was filled with commercial dialysate solution(Acid:Base:Water=1:1.83:34).Constant flow rates of 50 ml·min-1and 100 ml·min-1were maintained for solute and dialysate reservoirs correspondingly.

The experimental set-up comprised of a flat sheet module with an effective membrane surface area of 38 cm2and a cell volume of 10 ml.This single stage dialysis test module is inspired by that described by Kee and Idris[29].

The urea concentration of sample was analyzed using urea nitrogen(diacetyl)reagent,while creatinine(direct)reagent was used to determine the creatinine concentration.BSA concentration was measured by the Biuretreagent.Cytochrome-c was quantified using UV–visible spectrophotometer at 550 nm.Three runs were carried out for each prepared membrane to ensure consistency of performance.

Diffusive permeability P of a solute through the membranes can be given by.

where t1and t2are the sampling times,ΔC(t)is the difference between the solute concentrations of both cells at each sampling time,Vband Vdare the solution volumes in each cell(Vb=100 cm3,Vd=200 cm3)and S is the effective membrane area(38.5 cm2)

3.Results

3.1.Sulfonation of PES

Fig.1 shows the FTIR spectra of pristine PES and sulfonated PES.The peak observed at1025 cm-1is characteristic of the aromatic SO3Hsymmetric stretching vibrations while those observed at1180 cm-1are due to the asymmetrical stretching vibrations of sulfonic acid groups.The FTIR spectroscopy also shows another distinct peak at 3420 cm-1which is associated with the stretching of the hydroxyls of sulfonic acid groups[18,30].These results confirm the insertion of sulfonic acid groups onto the polymer backbone.

Fig.1.FT-IT spectra of pristine and sulfonated polyethersulfone.

3.2.Membrane preparation,porosity and UFR

The compositions of the prepared membranes are as shown in Table 1.The percentage of the SPES in the dope solution varied from 0 to 30%(by mass).Integrity of the membranes was ensured before further characterizations.The porosities and ultra filtration rate of all the fabricated membranes were studied and results are shown in Tables 1 and 2 respectively.Porosity of the blend membranes was found to be higher than the pristine PSf membrane.The pristine membrane has a porosity of 34.65%.The porosity was found to increase with the increase in SPES concentration showing a maximum of 52.89%for the 70/30 PSf/SPES blend.The pristine PSf membrane shows a UFR of(0.6857±0.0376)ml·m-2·h-1·Pa-1,that increases to(1.0925±0.0902)ml·m-2·h-1·Pa-1for the membrane M3 demonstrating the highest UFR.The UFR for all membranes were measured at a constant 33.331 kPa pressure.On increasing the SPES concentration further it was observed that there was a slight decrease in UFR.The UFR of membrane M4 was found to be 1.0436 ml·m-2·h-1·Pa-1.From both these studies it has become significantly clear that the water flux depends notonly on the membrane porosity but also on the hydrophilicity and properties of constituent polymers in the blend[31].

3.3.MWCO and mean pore size

Table 2 shows the corresponding MWCOs and the mean pore size of the membranes.The molecular weight cut-offs were investigated using PEGs of increasing different molecular weights.The pristine PSf membrane was found to have a molecular weight cut-off of 20 kDa.A higher rejection rate was observed for the blend membranes compared to the pristine.The M4 membrane was seen to have a molecular weight cutoff of 10 kDa.The mean pore radius of the membranes calculated from the sieving experiments was found to be 4.16 nm for the pristine PSf membrane which reduced to 3.54 nm for the membrane M3.

3.4.Cross section morphology-SEM

SEM has been applied as a powerful tool for characterizing the morphology of materials.All the membranes as seen in Fig.2 reveal typical asymmetric structure—a thin dense skin layer and a porous substructure.The top layer of membrane is responsible for the permeation or rejection,whereas the sub layer of the membrane provides mechanical support.Instantaneous demixing occurs in the formation of the pristinePSf membrane due to high mutual affinity of NMP for water,and results in the formation of finger like cavities in the sub layer of the prepared membranes.The cross sectional images of subsequent membranes indicated that incorporation of SPES in the casting solution produced membranes with sponge like structure in the sub layer with highly interconnected pores as evidenced from the figure.This is due to the presence of more hydrophilic SPES being incorporated into the dope solution which causes delayed demixing[32].

Table 2 Physico-chemical characteristics of the PSf/SPES blend membranes

3.5.Top surface morphology-AFM

To study the effect of SPES on the blend membrane morphology,AFM analysis PSf/SPES membrane surface were carried out by non contact mode in a scan area of5×5μm.The three dimensional AFM images of the membranes are shown in Fig.3.The surface roughness of membranes is a crucial factor affecting its biocompatibility.A smooth surface tends to decrease cell adhesion and thus enhances the biocompatibility of the membranes[33].It was observed that all the membranes possess a nodule-valley like structure which is typical of an asymmetric membranes prepared by phase inversion technique.The PSf/SPES membranes surface roughness were measured and tabulated as shown in Table 2.The pristine PSf membrane showed a Rzvalue of 33.36 nm which was suppressed to 19.63 nm for membrane M3.From the results,it was clear that the surface roughness of the membranes decreased with an increase in composition of SPES in the casting solution.

3.6.Wettability assay and work of adhesion

Wettability assay is an effective methodology to determine the hydrophilicity of the membranes.It was carried out by the measurement of the contact angle(CA)by the sessile drop method.Hydrophilic as well as the hydrophobic nature of the membranes play vital roles in determining the protein adsorption and celladhesion during hemodialysis[34].The CAs of the membranes are shown in Fig.4.The contactangle of hydrophobic PSf membrane was found to be 64.3°.A lower CA of 37.7°was observed on increasing the SPES composition to 30%(by mass).This demonstrates that on the addition of SPES,the hydrophilicity of the membrane surfaces rises.The work of adhesion(wA)values calculated from the contact angle values is also shown in Fig.4 The SPES incorporated PSf membranes were more hydrophilic since a greater amount of work is required to uniformly lift the formed liquid layer from the blend membrane surface[35].

Fig.2.Cross sectional morphology of membranes by scanning electron microscopy.(a)M1,(b)M 2,(c)M 3 and(d)M4.

Fig.3.Top surface imaging of the membranes by atomic force microscopy.(a)M1(b)M2(c)M3 and(d)M4.

Fig.4.Wettability assay(contact angle)and work of adhesion of the membranes.

3.7.Biocompatibility studies

3.7.1.Protein adsorption

The protein adsorption is an initial event on blood-membrane surface contact and denotes a rapid adsorption of plasma proteins from the blood onto the membrane surface.The surfaces of synthetic biomaterials are generally not bioactive themselves[36].It is found that the surface bioactivity is initiated by the proteins that adsorb onto the biomaterial surface which could lead to coagulation,complement activation and fibrinolysis[37].Protein adsorption on the membrane surface or within the pores is irreversible,occurring in the initial minutes of contact between blood and membrane.The amount of protein adsorbed onto the pristine PSf and SPES incorporated blend membranes are shown in Fig.5.It is observed that the amount of protein adsorption onto membranes decreases with an increase in concentration of SPES in the blend.Adsorption of BSA onto the pristine PSf membrane was quantified at 3.10 μg·cm-2,conversely the values were found to be 0.98 μg·cm-2for the membrane M4.This is in accordance with the CA values,as the incorporation of SPES into the blend introduces hydrophilic domains on the membrane surfaces.

Fig.5.Protein adsorption onto the membranes.

3.7.2.Platelet adhesion

Platelet adhesion is another critical attribute in the evaluation of potentially more biocompatible materials[38].Platelet adhesions onto the prepared membranes were counted by the LDH assay and results are shown in Table 3.The adhered platelet count was found to be 14223(±850)cells per square millimeter on the pristine PSf membrane and 8753(±655)cells per square millimeter on membrane M4.The mixture of hydrophobic–hydrophilic domains on the membrane reduces the membrane fouling due to platelet adhesion.These results suggest that SPES incorporated blend membranes suppress platelet adhesion.The outcome of platelet adhesion follows a similar trend as observed in protein adsorption studies.This suppression of platelet adhesion is generally believed to be due to reduction of protein adsorption[39].

Table 3 Complement activation,platelet adhesion and clotting time of the membranes

3.7.3.Complement activation

Foreign materials on contact with blood activate the complement system to some degree via the alternative pathway depending on their physico-chemical surface and structure.Complement activation represents an inflammatory response of the host towards ‘non-self’structures and plays a central role in the recognition of,and defense against,‘non-self’materials[40].Therefore,determining the complement activation of dialysis membranes is highly significant.The serum complement activation could be assessed by determining the generated anaphylatoxins C3 and C5.The concentrations of C3 and C5 left in the serum after incubation for 2 h with the prepared PSf/SPES blend membranes are shown in Table 3.The concentrations of C3 and C5 generated by the pristine PSf membrane were 135.3 and 29.8 mg·dl-1respectively.These findings proved that pristine PSf membranes activated the complement systems.Reduced concentrations of C3 and C5(121.9 and 22.4 mg·dl-1,respectively)were exhibited by M4 membranes.This points to the fact that blend membranes have only a modest ability to activate the complement cascade[41].

3.7.4.Blood coagulation time

The blood coagulation time in relation to exposure to membrane surface was studied using PTT and aPTT and results were tabulated in Table 3.For the pristine PSf membranes,the aPTT and PTT values were found to be(37.3±2)s and(14.2±1)s.These observations were similar to the control values.The clotting time significantly improved as the composition of SPES in the blend was increased to 30%(by mass).aPTT and PTT values were also found to be extended up to(54.4±1)s and(22.7±2)s respectively for M4.These results suggested that the sulfonic groups on the membrane surface might be effective in prolonging the blood clotting time.In addition,the improvement of anticoagulant activity was in conformity with the increased hydrophilicity,decreased protein adsorption,and suppressed platelet adhesion and modest complement activation.

3.8.Permeability studies of solutes

The diffusive permeability of the prepared membranes were studied using uremic toxin solutions such as urea,creatinine and cytochrome C.The value for urea was found to be 32×10-4cm·min-1for pristine PSf membranes and this value increased to a maximum of 96 × 10-4cm·min-1in membrane ‘M3’as presented in Fig.6.The diffusive permeability of creatinine for membrane ‘M3’was found to be 52×10-4cm·min-1in contrast to 24×10-4cm·min-1for the pristine membrane.A similar trend can also be observed for the middle molecular weight solute cytochrome-c,the diffusive permeability of which increased from 8 × 10-4cm·min-1for membrane ‘M1’to 18 × 10-4cm·min-1while using membrane ‘M3’.

Fig.6.Diffusive permeability of solutes,urea(60 Da),creatinine(113 Da)and cytochrome-c(11800 Da).

4.Discussion

The consequences of an escalation in SPES concentration on the increase in porosity of the blend membranes can be explicated both by the thermodynamics as well as kinetics of the dope solution during membrane formation[42].The presence of a more hydrophilic polymer in the casting solution causes a thermodynamic delay in the phase separation as evidenced by morphological analysis.Secondly,it also causes a kinetic impediment for solvent and non-solvent interaction during phase separation by increasing the viscosity of the casting solution resulting in delayed demixing.Water exists in a cluster of100 molecules held together by hydrogen bonds to the closest four to six neighbors.So the permeation of water through the membrane is dependent on the physical and chemical interaction of water with the surface of the membrane and the walls of the pores[43].Transport of water through the membrane barrier is facilitated by the presence of the polar sulfonic acid group in competition with the inherent self-associating tendency of water molecules to form clusters.However,in the case of pure PSf membranes the hydrophobicity of the membrane surface hinders the passage of water molecules resulting in a lower ultra filtration rate.

Though the dope solution is homogenous,the SEM image of the membrane cross section shows an asymmetric morphology.This is because mass transfer induces a phase separation and the subsequent vitrification eliminates further reorganization of a polymer rich phase[44].The greater affinity of the PSf/SPES dope solution to water results in longer time being required for solvent-non-solvent exchange.This leads to the formation of a spongier sublayer[45].Consequently,it can be deduced that the longer the exchange time between solvent and nonsolvent in the coagulation bath,the more developed the process of polymer-lean phase growth and coalescence.Generally,more pores on membrane surfaces and the better interconnectivity inside membranes would contribute to enhancing pure water flux and improving selectivity.Pure water permeation is strongly dependent on the top layer and sub-layer of the membranes.

Biocompatibility has always been the crucial requirement for a good hemodialysis membrane in order to prevent any immune system activation and thrombosis during dialysis treatment.Masayo Hayama et al.[46]proved that better hemocompatibility,i.e.lower platelet adhesion ratio,was exhibited by membranes with smoother surface.Fukuda et al.[47]revealed similar findings using commercial dialysis membranes,where cellulosic membrane with the smoothest surface demonstrated excellent anti-thrombogenity.It has been observed that a doubling in membrane peak to valley(PV)roughness leads to a 3.5 fold increment in the number of platelets adhered.

On introduction of SPES into the blend it introduces hydrophilic domains on the membrane surfaces.Considering the thermodynamic relationship of ΔGads= ΔHads–TΔSads,it is clear that to provide protein resistance,the interaction of a surface with a protein should thus ideally result in ΔHads≥ 0 and ΔSads≤ 0;or in other words,the unbound hydrated state of the surface functional group should be lower in enthalpy and higher in entropy than when bound to a protein[48]The enthalpy will be provided if the surface functional group is able to bind with water more favorably than with functional groups of the protein.As for entropy,the-SO3H functional group must be able to maintain a higher state of system entropy when bonded to water than when bound to the protein.

When SPES is added to the blend mixture,the hydrogen bonding SO3H groups are oriented in such a manner that water can access the groups much more readily than the functional groups of a protein.A lower state of enthalpy will subsequently be provided for the hydrated,non-adsorbed state.Analysis showed an increase in BSA adsorption with a decrease of the free surface energies.It can be expected that BSA will undergo unfolding on low energy surfaces to minimize the interfacial energy and maximize protein–surface bonds[49].

Suppression of platelet adhesion is generally believed to be due to the reduction of protein adsorption.Our results are in concordance with the hypothesis of Tanaka et al.and Tzoneva et al.that less wettable surfaces cause unfolding by tight binding of the adsorbed protein and thus promotes platelet adhesion due to the exposure of the binding sites[50,51].

The large portion of platelet or plasma protein surface is recognized as charged negatively,resulting in an attractive electrostatic interaction with the positively charged membrane surface.Lee et al.reported the effects of different functional groups introduced on the PE sheet.Carboxylic acid groups for AA grafted surfaces and sulfonate groups for NaSS grafted surfaces are negatively charged.These groups demonstrated poor platelet adhesion[52].

The immune system of a healthy individual is provoked by antigens such as bacteria,viruses,and polymeric materials that are distinctly different from body tissues[53].The body elicits an immune response to eliminate these antigens.The cascade of enzymes forming the complement system(immune response)consists of approximately 30 fluidphase and cell-membrane bound proteins,all of which play a vital role in the body's defense systems against pathogenic xenobiotics[54].Tang et al.[55]reported that negatively charged SO3H and COOH showed lower protein adsorption,suppressed platelet adhesion and reduced thrombin–antithrombin(TAT)generation.Moreover the percentage platelets positive for CD62p expression was found to be fewer,and they exhibited only moderate complement activation.The SPES modified membranes inhibit the C3 and C5 activation mechanism in blood plasma.

The coagulation cascade has three pathways:intrinsic,extrinsic,and common.The intrinsic and extrinsic pathways lead to the formation ofa fibrin clot.APTT and PT tests were performed to exhibit these pathways during blood coagulation.This confirmed the presence of the sulfonic acid group responsible for improved hydrophilicity and lower platelet adhesion.

In diffusive permeability studies,the rate of the solute movement through the blend membranes depends mainly on the size of the solute.It was found that none of the prepared membranes allowed the permeation of BSA as expected because the Molecular weight cutoff of these membranes was less than 64 kDa.In hemodialysis,the solutes with low molecular weights are removed from the solution by allowing them to diffuse into a region of low concentration.Hence for all the prepared membranes,it can be observed that the diffusive permeability of urea＞creatinine＞cytochrome-c.There is only a minor pressure difference across the membrane and the flux of each solute is proportional to the size as well as concentration difference.Urea is a standard solute marker and has high clearance during hemodialysis.Its movement in and out of the RBCs is rapid.The dialysis clearance of urea is greater than the plasma flow.Therefore,the clearance of urea is higher than other solutes.Creatinine shows much lower clearance since its removal is based on size alone[56].These results are also in accordance with the UFR of the membranes.Maximum diffusive permeability was observed for membrane ‘M3’because of higher UFR and more number of evenly distributed pores.From AFM imaging,it can be observed that the blend membranes are smoother.This implies that given the possibility of the same number of particles being deposited they would most likely be more evenly spaced out resulting in a reduced overall flux decline.

5.Conclusions

An effective and commercially viable way to modify polymeric membranes with excellent biocompatibility has been realized by blending sulfonated polyethersulfone with PSf.We find that on increasing concentrations of SPES,the MWCO of the membranes decrease.80/20 PSf/SPES blend membranes show an optimum cut-off of 12 kDa,which is suitable for the removal of middle molecular weight toxins.The blend membranes display an increase in UFR making it a viable alternative for hemodialysis membranes as it reduced the overall dialysis time.An escalation in hydrophilicity with an increase in the concentration of SPES is observed by a decrease in contact angle values and a corresponding rise in work ofadhesion values.The increase of clotting time when blood is exposed to the membranes ensures a better biocompatibility of the membranes by lowering the chances of complement activation.The presence of SPES on the surface of the blend membranes ensures lesser protein adsorption onto the membrane surface,thereby lowering the cascade up to thrombus formation.All these results indicate that the PSf/SPES blend membranes show a tremendous potential in hemodialysis applications.

Acknowledgments

We would like to thank Dr.S.Senthil Kumar for his valuable assistance with the biocompatibility characterizations.This study was supported by the Department of Science and Technology(DST),Government of India(IDP/MED/2010/17/2(General)).