Preparation of polysulfone-based PANI–TiO₂ nanocomposite hollow fiber membranes for industrial dye rejection applications

Abstract

Polysulfone-based polyaniline–TiO₂ containing hollow fiber membranes were prepared via a dry wet spinning method. Polyaniline (PANI) coated TiO₂ nanotubes were prepared via chemical oxidative polymerisation and were incorporated into the hollow fiber membranes at different compositions. The hollow fibers were fabricated by varying the air gap distance during the spinning process. The effects of the addition of PANI coated TiO₂ and the variation in the air gap distance on membrane performance, such as morphology and the permeability of the membranes, were analysed. The addition of the PANI–TiO₂ nanocomposite enhanced the hydrophilicity and antifouling ability of the prepared membranes. The polysulfone hollow fiber membranes were examined for their dye rejection of Reactive Black 5 and Reactive Orange 16. The results indicated that the polysulfone hollow fibers containing 1.0 wt% of PANI–TiO₂ fabricated using a 5 cm air gap can be used as a potential candidate for industrial dye rejection and showed a maximum rejection of 81.5% and 96.5% for Reactive Black 5 and Reactive Orange 16, respectively.

---

1. Introduction

Membrane separation methods using hollow fiber membranes are among the emerging technologies developed over the past 3–4 decades.¹ Hollow fiber membranes are used for several applications, such as wastewater treatment, dye rejection and gas separation, due to their selectivity and good productivity.^1,2 Hollow fiber membranes have been preferred over other membrane configurations due to their beneficial properties such as high membrane surface area to module volume, mechanical property and easy handling.² The immersion precipitation process provides asymmetric hollow fiber membranes, which are used for separation processes and exhibit characteristic features more advantageous than flat sheet membranes.³ The main advantages of hollow fibers over flat sheet membranes are their high productivity and self supporting nature. The high productivity is due to the increased surface area to volume ratio of the hollow fibers. Since the hollow fibers are self supporting, it simplifies the fabrication of the permeation cell whereas, the flat sheet membranes require spacers or porous supports during permeation.^4,5

Polysulfone (PSf) is one among the excellent polymeric materials used for hollow fiber membrane fabrication.⁶ The properties of hollow fiber membranes can be tailored through surface modification techniques such as polymer coating, photografting and plasma treatment.⁷ Hollow fiber membranes with additives are fabricated in order to have an essential membrane morphology and desired properties.² Adding inorganic nanoparticles into the polymer dope solution when preparing nanocomposite hollow fiber membranes is an interesting method used to improve the separation properties due to the synergistic effects of both the nanoparticles and organic membranes. The preparation of polyaniline (PANI) composites with inorganic nanoparticles has gained interest in recent years to improve the performance.⁸ The PANI–TiO₂ nanocomposites exhibit properties as a combination of PANI and TiO₂, which are otherwise difficult to achieve with the individual components.⁸ Pereira et al. prepared polysulfone ultrafiltration flat sheet membranes using PANI–TiO₂ as an additive. Polyethylene 1000 was added as a pore former. The properties of the membranes, such as porosity, water uptake and antifouling ability, improved with the addition of PANI–TiO₂. However, the membranes showed a decrease in these properties upon adding 1.5 wt% of PANI–TiO₂.⁹ Teli et al. prepared PSf ultrafiltration flat sheet membranes with PANI coated TiO₂ nanoparticles. The hydrophilicity of the membranes increased with an increase in the concentration of the PANI–TiO₂ nanoparticles (0–1.5 wt%). The membranes exhibited higher porosity and more finger-like projections with PANI–TiO₂. The results indicated 1.0 wt% of PANI–TiO₂ showed excellent properties.

Dyes have been extensively used in the textile industry and are one of the dangerous sources of water contamination.^8,10 Reactive dyes are usually found in higher concentrations in textile industry effluents due to their low degree of fixation to fibers such as cotton and viscose.¹¹ TiO₂ nanoparticles are widely used in wastewater treatment because of their low toxicity, stability and photocatalytic activity.¹² However TiO₂ nanotubes are preferred over TiO₂ nanoparticles as they provide a higher surface area.^13,14 As a result, a higher PANI coating on TiO₂ can be achieved when TiO₂ nanotubes are used. Recently, there have been various reports where PANI–TiO₂ nanocomposites were effectively used for dye removal. The presence of PANI improved the dye removal efficiency of TiO₂. Liu et al.¹² prepared PANI coated TiO₂/SiO₂ nanofiber membranes, which were used for the degradation of methyl orange dye. The membranes showed enhanced photocatalytic activity under visible light due to the synergetic effects of TiO₂ and PANI. Debnath et al.¹⁵ synthesized a PANI coated TiO₂ nanocomposite, which was used as a catalyst for the photodegradation of Eosin Yellow (EY) and naphthol blue black (NBB) dyes. The catalyst showed 99.85% and 99.74% degradation for the EY and NBB dyes, respectively under optimum conditions. Jeong et al.¹⁶ prepared PANI–TiO₂ nanocomposites, which were used for their antibacterial activity and as a photocatalyst for the degradation of Methylene Blue (MB) dye under visible light irradiation. The synergy between PANI and TiO₂ contribute towards the higher charge separation, which in turn enhance the degradation of MB. Razak et al.¹⁷ immobilized PANI coated TiO₂ on a glass plate using PVC as an adhesive. The presence of PANI on TiO₂ enhanced the photocatalytic activity for the removal of Reactive Red 4 dye.

In our previous research work,⁹ PANI coated TiO₂ was used as an additive to prepare polysulfone flat sheet membranes along with PEG 1000 as a pore former. Keeping in view the benefits of hollow fiber membranes over flat sheet membranes and also the enhancement in membrane performance upon the addition of PANI–TiO₂, the same was incorporated into the polysulfone hollow fiber membranes. However, in our current work the PEG 1000 pore former was not incorporated and only the effect of PANI–TiO₂ on the polysulfone hollow fiber membranes was studied. Polyaniline also serves as an effective additive for polysulfone membranes. Kajekar et al.¹⁸ incorporated PANI nanofibers into polysulfone hollow fibers and used these membranes for the rejection of Reactive Red dye, where the presence of PANI significantly improved the dye rejection.

In view of the encouraging observations described beforehand, polysulfone nanocomposite hollow fiber membranes were prepared with PANI-coated TiO₂ nanotubes as an additive. Hollow fibers were fabricated with different concentrations of PANI–TiO₂ nanocomposites and their effect on the performance of the membranes was analyzed. The air gap distance was also varied during the membrane fabrication process and its effect on the membrane performance was evaluated. The hollow fiber nanocomposite membranes were also investigated for the rejection of the reactive dyes, Reactive Black 5 and Reactive Orange 16.

2. Experimental

2.1 Materials

Polysulfone (PSf Udel-P3500), TiO₂ nanoparticles and aniline (99.5%) were purchased from Sigma Aldrich Co. Bangalore, India. Bovine Serum Albumin (BSA), Reactive Black 5 and Reactive Orange 16 were purchased from Sigma Aldrich Co. Malaysia. N-Methyl-2-pyrrolidone (NMP) was purchased from Merck Malaysia, Ltd. Hydrochloric acid (HCl) was procured from Merck India, Ltd. Ammonium peroxydisulfate (APS) was obtained from the Central Drug House (CDH), New Delhi.

2.2 Preparation of the PANI coated TiO₂ nanotubes

PANI coated TiO₂ nanotubes were synthesized as reported in our earlier work.⁹ Initially TiO₂ nanotubes were prepared from TiO₂ nanoparticles and then the TiO₂ nanotubes obtained were surface coated with PANI. In brief, TiO₂ nanotubes were coated with PANI via chemical oxidative polymerization.¹⁹ 2.5 mL of aniline was added to 500 mL of 2 M HCl solution. To this solution, 2.5 g of TiO₂ nanotubes dispersed in 25 mL of water was also added. The solution was stirred for 2 h. Polymerization of aniline was carried out upon the addition of 5.7 g of ammonium peroxydisulfate (APS) to the suspension. The solution was centrifuged to isolate the product followed by washing with HCl and water. The product was dried for 24 h at 80 °C to yield 6.0 g of PANI coated TiO₂.

2.3 Preparation of the membrane

PSf, which was used as the base polymer for the hollow fibers, was initially dried in a hot air oven at 50 °C for 24 h to remove any moisture present. The PANI coated TiO₂ nanotubes were dispersed in a desired amount of NMP taken in a Scott bottle. The TiO₂–PANI nanocomposites were dispersed in NMP by sonication for 30 min at frequency of 40 kHz using an ultrasonic bath (Ultrasonic Cleaner DC-150H). The mixture was then stirred at 27 °C with an overhead stirrer and the PSf polymer was added to the solution. The solution was stirred until complete dissolution of the polymer was achieved.²⁰ The obtained homogenous dope solutions were degassed using the ultrasonic bath to remove the trapped air bubbles.²¹ The neat PSf dope solution, which did not contain any PANI–TiO₂ nanocomposite, was prepared in a similar way without the addition of PANI–TiO₂. The compositions of the dope solutions are given in Table 1. The viscosity of the dope solutions was recorded using a basic viscometer (Model: EW-98965-40, COLE PARMER, 20–2 million centipoises). The hollow fiber membranes were prepared via dry–wet spinning methods. The homogenous spinning dope solutions and bore fluid were extruded through the spinneret to obtain the hollow fiber membranes. The spinning conditions are given in ESI S1.† The hollow fiber membranes obtained were labeled as per the codes given in Table 1. The hollow fibers were immersed in water for 1 day to remove the residual solvent. Then, the fibers were post-treated using an aqueous solution of glycerol (glycerol : water; 10 : 90 wt%) for 24 h to prevent the pores from collapsing.²² Finally, the hollow fiber membranes were air dried for 2 days at room temperature before further use.

Table 1 The composition of the dope solutions

Membrane code	Polysulfone (wt%)	PANI–TiO2 (wt%)	Air gap between the spinneret and the bath	Flow rate (mL min^−1) (dope/bore liquid)
P-A	20	0	5 cm	5/5
P-B	20	0	10 cm	5/5
P-0.5-A	20	0.5	5 cm	5/5
P-0.5-B	20	0.5	10 cm	5/5
P-1.0-A	20	1.0	5 cm	5/5
P-1.0-B	20	1.0	10 cm	5/5

---

3. Characterization

3.1 Characterization of the PANI coated TiO₂ nanotubes

The TiO₂ nanotubes were coated with PANI by performing an oxidative polymerization reaction on the surface of the TiO₂ nanotubes. Initially, the TiO₂ nanotubes were dispersed in a solution containing aniline in HCl. In the presence of HCl, the monomer aniline is converted to an anilinium cation, which gets adsorbed on the surface of the TiO₂ nanotubes. Polymerisation of the anilinium cations results in the PANI coating on the TiO₂ nanotubes.^9,23

The PANI coating on the TiO₂ nanotubes was studied by X-ray diffraction (XRD) and Fourier transform infrared spectroscopy (FTIR). The PANI coated TiO₂ nanotubes were also characterized by transmission electron microscopy (TEM). A SHIMADZU ATR-FTIR spectrophotometer was used to record the FTIR spectra. XRD measurements were recorded using a Rigaku Miniflex 600 diffractometer with Cu radiation. The morphology of the PANI coated TiO₂ nanotubes was observed using transmission electron microscopy (JEOL JEM-2100).

3.2 Characterization of the membranes

3.2.1 Membrane morphology. The hollow fiber membrane morphology was inspected by observing the cross-sectional images under scanning electron microscopy (SEM, TM 3000, Hitachi). The membrane samples were immersed in liquid nitrogen, frozen and then broken. The membrane samples were placed on a metal holder and sputtered with gold under vacuum for electron conductivity.

3.2.2 Porosity and pore size of the membranes. The overall porosity (ε) of the hollow fibers was evaluated by a gravimetry method,²⁴ which is given by the following equation:
where, w₁ and w₂ are the wet weights and dry weights of the membranes, respectively, ‘A’ is the effective membrane area (m²), ‘l’ is the thickness of the membrane (m) and d_w is the density of water.

The pore size (r_m) of the membranes was calculated using the Guerout–Elford–Ferry equation given below:

where ‘ε’ is the porosity of the membranes, ‘η’ is the viscosity of water, ‘l’ is the membrane thickness (m), ‘Q’ is the volume of pure water permeating through the membrane per unit time (m³ s⁻¹), ‘A’ is the effective membrane area (m²) and ΔP is the operating pressure (0.2 MPa).⁹

3.2.3 Water uptake study. The water uptake ability of the hollow fibers was determined by immersing the hollow fibers in distilled water for 5 h. Then, the membranes were removed from the water and the wet weight was noted immediately after blotting the surface water. The hollow fiber membranes were then dried and the dry weight was noted. The water uptake capacity of the hollow fiber membranes was calculated using the relationship:
where, W_w and W_d are the wet weight and dry weight of the hollow fibers, respectively.

3.2.4 The contact angle of the membranes. The contact angle between water and external surface of the hollow fibers was measured using a contact angle goniometer (Model: OCA 15 EC, Dataphysics). To reduce the experimental error, for each sample, the values were taken as an average of 10 contact angles measured at different locations on the hollow fibers.

3.2.5 Atomic force microscopy of the membranes. The AFM studies on the hollow fibers were conducted using SPM atomic force microscopy. AFM analysis was conducted using antimony doped silicon cantilever, which had a force constant of 20–80 N m⁻¹. The hollow fibers were cut into smaller pieces and were fixed horizontal on the metal disk using double-sided tape.²⁵ The samples were imaged at a scan rate of 1 Hz and scan size of 3 μm × 3 μm. The difference in the surface roughness was assessed in terms of roughness parameters such as average roughness (R_a) and root mean square roughness (R_a).^25,26

3.2.6 Pure water flux measurements. The pure water flux of membranes was measured using a bench scale cross flow filtration set up. The membrane modules for the filtration studies were prepared by potting the membranes with the help of epoxy resin and hardener.¹⁸ Five membranes with a length of 16 cm were assembled into each module. Distilled water was circulated through the set up. The membranes were compacted at a transmembrane pressure (TMP) of 3 bar for 30 min prior to further measurements. Then, the pure water (J_w1) was collected on the lumen side at a TMP of 2 bar.
where, Q is the pure water collected (L) for time Δt (h) and A is the surface area of the membrane (m²).

3.2.7 Antifouling studies. Fouling experiments were carried out using BSA as a model foulant. The BSA solution was prepared by dissolving BSA in distilled water at a concentration of 800 ppm. The BSA solution was filtered through the membranes. The BSA flux was collected to evaluate the % BSA rejection of the membranes. After BSA filtration, the membranes were thoroughly rinsed and washed with water and then the water flux (J_w2) of the membranes was measured. The fouling resistance of the membranes was calculated using the following formula:

The % BSA rejection of the membranes was evaluated using the equation:

where, C_p (mg mL⁻¹) is the concentration of BSA in the permeate and C_f (mg mL⁻¹) is the concentration of BSA in the feed solution.

3.2.8 Dye rejection. The dye removal performance of the nanocomposite hollow fibers was studied using Reactive Black-5 and Reactive Orange-16 dyes. The aqueous solutions of RB-5 and RO-16 were prepared at a concentration of 10 ppm. The molecular structures of RB-5 and RO-16 are given in the ESI S3.†² The dye rejection studies of the hollow fiber membranes were carried out on the same cross-flow filtration set up, under similar conditions¹⁰ as those used for the permeation studies. The concentration of the dyes in the aqueous solutions were investigated using an UV-Vis spectrophotometer (HACH, Model: DR 5000) at the respective wavelength of the maximum absorption for each dye.²⁷ The dye rejection by the membranes was investigated by means of % rejection, given by the formula:
where, C_i and C_f are the initial and final concentrations of the dyes in the aqueous solutions, respectively.

4. Results and discussion

4.1 Characteristics of the PANI coated TiO₂ nanotubes

The characteristics of the PANI coated TiO₂ nanotubes confirmed the PANI coating on TiO₂. The FTIR spectra of PANI, TiO₂ and PANI–TiO₂ nanocomposite are shown in the ESI S4.†

The X-ray diffraction patterns of PANI, TiO₂ and PANI–TiO₂ nanocomposite were also studied and compared (ESI S5†).

Fig. 1 shows the TEM images of the PANI coated TiO₂. The diameters of the PANI coated TiO₂ nanotubes were in the range of 40–50 nm.

Fig. 1 TEM images of the PANI coated TiO₂ nanotubes (A and B).

4.2 Membrane morphology

The SEM cross-sectional images of the hollow fiber membranes are shown in Fig. 2–4. It was found that finger-like projections were formed beneath inner and outer surfaces of the hollow fibers with a thin spongy layer sandwiched between them.²⁸ The addition of the PANI–TiO₂ nanotubes and variation in spinning parameters like the air gap distance brought about a change in the membrane morphology.²⁹ It was observed that the pristine polysulfone hollow fibers showed a more porous nature in the skin layer with elongated pores whereas the porous nature in the skin layer was decreased upon the addition of PANI–TiO₂.

Fig. 2 SEM cross-sectional images of P-A and P-B ((A) and (B) are the magnified images of (a) and (b), respectively).

Fig. 3 SEM cross-sectional images of P-0.5-A and P-0.5-B ((C) and (D) are the magnified images of (c) and (d), respectively).

Fig. 4 SEM cross-sectional images of P-1.0-A and P-1.0-B ((E), (F) and (G) are the magnified images of (e), (f) and (g), respectively).

The viscosity of the polymer dope solution increased with an increase in the PANI–TiO₂ content in the dope solution as seen in the ESI S2.† The viscosity of the dope solution affects the rheological properties during liquid–liquid demixing, which in turn affects the membrane morphology. During the phase inversion process, an increase in the dope solution viscosity decreases the mass transfer rate, which prevents the formation of a macroporous structure. Hence, the length of the finger-like projections decreased.³⁰

A difference in morphology beneath the outer surface of the hollow fibers was also observed when the air gap distance was varied for same composition of the dope solutions. With an increase in the air gap distance, the depth of the finger-like projections was reduced. The outer surface appeared to be denser with a porous nature as the air gap distance was increased from 5 cm to 10 cm. This was more prominently visible in the case of the P-1.0-A and P-1.0-B membranes. Similar results were also observed and reported by Sengur et al. and Shi et al.^29,31

During the dry–wet spinning method, the hollow fiber undergoes two coagulation paths. The first is a “non convective” type in the air gap region, then a rapid exchange of solvents takes place as the fiber enters the coagulation bath. As the air gap distance was increased, the external surface of the fiber was in more contact with air and undergoes a humid-induced phase separation for a longer time and then, a delayed phase inversion in the coagulation bath. Whereas, when the air gap was reduced, the time delay causing the phase inversion process was reduced, since the fiber enters the coagulation bath faster, as a result, more and longer finger-like projections are produced.³²

At a higher content of PANI–TiO₂ in the hollow fibers, i.e., at 1.0 wt%, agglomeration of PANI–TiO₂ was observed as shown in Fig. 4(g and G). The agglomeration of PANI–TiO₂ in the membrane was further confirmed by elemental mapping and the EDX spectrum (Fig. 5 and ESI S8†). However, a lower content of PANI–TiO₂ did not show any agglomeration.

Fig. 5 Elemental mapping of the P-1.0-B membrane.

4.3 Porosity, pore size and water uptake of the membranes

The porosity and pore size of the membranes is shown in Table 2. The porosity of the hollow fibers decreased with the inclusion of the PANI–TiO₂ nanocomposites. This is because the viscosity of the polymer solutions increased upon the addition of PANI–TiO₂, where the out-diffusion of solvent was more favored than the in-diffusion of the non-solvent, leading to a decreased porosity.²⁹ However at 1.0 wt% of PANI–TiO₂, the porosity increased. This increase was attributed to the hydrophilicity of the PANI–TiO₂ membranes at 1.0 wt% where the presence of a higher concentration of PANI–TiO₂ will attract more water flow into the membrane, leading to a higher porosity.³⁰

Table 2 The porosity and pore size of the membranes

Membranes	Porosity (%)	Pore size (nm)
P-A	38.94	5.45
P-B	35.8	8.37
P-0.5-A	34.38	4.12
P-0.5-B	28.61	6.19
P-1.0-A	41.47	1.88
P-1.0-B	37.97	3.57

---

The water uptake of the membranes is shown in Fig. 6. The membrane water uptake was dependent on the porosity of the membranes. The water uptake decreases and then increases for the 1.0 wt% of PANI–TiO₂ membrane. The water uptake values are consistent with the porosity values, where the porosity was found to first decrease and then increase.

Fig. 6 Water uptake by the membranes.

4.4 Contact angle of the membranes

The contact angle of the hollow fiber membranes is shown in Fig. 7. The neat hollow fiber membranes P-A and P-B, showed a contact angle of 84.54° and 82.27°, respectively, indicating that the membranes were less hydrophilic in nature. The contact angle of the hollow fiber membranes P-0.5-A and P-0.5-B, reduced, exhibiting a contact angle of 78.60° for the P-0.5-A membrane and 79.26° for the P-0.5-B membrane. The membranes P-1.0-A and P-1.0-B, which were composed of 1.0 wt% of PANI–TiO₂ showed an even lower contact angle of 73.05° and 72.28°, respectively. This reduction in the contact angle was attributed to the addition of the PANI–TiO₂ nanotubes, which are hydrophilic in nature. Hence, the PANI–TiO₂ nanocomposites reduced the hydrophobicity of the polysulfone membranes. The variation in air gap did not affect the contact angle of the membranes. The membranes with same composition but prepared with different air gaps showed nearly the same contact angle values.

Fig. 7 The contact angle of the membranes.

4.5 Surface roughness of the membranes

The surface topography of the hollow fiber membranes was examined through AFM. The 3-dimensional and 2-dimensional scans of the hollow fibers P-A, P-0.5-A and P-1.0-A are shown in Fig. 8. It was observed that the surface roughness of the membranes increased upon an increase in the amount of PANI–TiO₂ added. This may be due to the reason that neat PSf, which does not contain any PANI–TiO₂ resulted in a smooth surface. Whereas in the case of the P-0.5-A and P-1.0-A membranes, the presence of PANI–TiO₂ on the surface of the hollow fibers gave rise to nodules or lumps on surface, which may be due to the presence of agglomerated PANI–TiO₂, contributing to an increase in the surface roughness. The trend was further supported by the surface roughness parameters measured as the mean roughness (R_a) and root mean surface roughness (R_q) given in Table 3.

Fig. 8 Three-dimensional scans of (A) P-A (B) P-0.5-A (C) P-1.0-A and two-dimensional scans of (a) P-A (b) P-0.5-A (c) P-1.0-A.

Table 3 The surface roughness parameters of the membranes

Membranes	Ra (nm)	Rq (nm)
P-A	10.7	12.5
P-0.5-A	13.8	16.9
P-1.0-A	30.0	37.1

---

4.6 Permeation studies

Fig. 9 displays the PWF of the membranes measured at 0.1 MPa. There are two main trends observed in the pure water flux values of the membranes. The first trend was that the membranes fabricated with a higher residence time in air exhibited a higher PWF. As reported by Sengur et al.²⁹ membranes with a higher air residence time have a greater permeation rate. The pore sizes of the membranes fabricated with a higher air gap were greater than those of the membranes prepared with a lower air gap (Table 2). Therefore, the membranes P–B, P-0.5-B and P-1.0-B exhibited a higher flux than the corresponding P-A, P-0.5-A and P-1.0-A membranes.

Fig. 9 Pure water flux of the membranes.

The second trend was that the PWF of the membranes decreased as the PANI–TiO₂ content in the hollow fibers increased because the porosity and pore size of the membranes showed a decrease upon increasing the PANI–TiO₂ content in the hollow fiber membranes.

4.7 Antifouling properties

The flux of the membranes during BSA filtration is given in Fig. 10. The figure also shows the flux of the membranes for water before and after the BSA filtration. The antifouling ability of the membranes was evaluated by calculating the FRR value for the hollow fiber membranes. The pristine membranes exhibited lower FRR values whereas the hollow fibers containing PANI–TiO₂ showed better % FRR values. Membrane P-1.0-B showed the highest flux recovery ratio of 95% (Fig. 11). The higher FRR value indicates that in the presence of PANI–TiO₂, the BSA molecules did not remain adsorbed on the membrane surface after BSA filtration. Thus, the membranes containing PANI–TiO₂ exhibited antifouling nature.

Fig. 10 Flux of the membranes during BSA filtration.

Fig. 11 The flux recovery ratio (FRR) of the membranes.

The % BSA rejection of the membranes was also evaluated. The hollow fibers containing PANI–TiO₂ showed 99% rejection for BSA molecules whereas the pristine membranes P-A and P–B exhibited a rejection of 95% and 93.6%, respectively (Fig. 12). This is because the nanocomposite hollow fibers have pore sizes smaller than that of the BSA molecules, resulting in almost complete rejection.

Fig. 12 The % BSA rejection by the membranes.

4.8 Dye rejection by the membranes

The dye rejection by the hollow fiber membranes is shown in Fig. 13 and 14, the P-1.0-A membrane showed the highest rejection of 81.5% for Reactive Black 5 and a maximum of 96.5% for Reactive Orange 16. The pristine hollow fiber membranes, P-A and P-B showed the least rejection for both dyes.

Fig. 13 Dye rejection by the membranes for Reactive Black 5.

Fig. 14 Dye rejection by the membranes for Reactive Orange 16.

The rejection of dyes by the hollow fibers increases with an increase in the PANI–TiO₂ composition in the membranes. The PANI–TiO₂ incorporated hollow fibers have NH⁺ centers from PANI due to which there is an electrostatic interaction between the polyaniline and dye anions.³³ Reactive Black 5 and Reactive Orange 16 are anionic dyes, which dissociate in aqueous solutions to give sulfonic acid groups. As a result of this, the dyes are adsorbed on the membranes. Therefore, the higher the PANI–TiO₂ content in the membranes, the higher the rejection will be. Also, it was observed that the ‘A’ membranes (P-A, P-0.5-A and P-1.0-A) showed a higher rejection than their corresponding ‘B’ membranes (P-B, P-0.5-B and P-1.0-B) for both dyes. This is because the membranes P-B, P-0.5-B and P-1.0-B have pore sizes greater than their corresponding ‘A’ membranes (Table 2). The digital images of the dye permeate after filtration of the Reactive Black 5 and Reactive Orange 16 dyes through the hollow fiber membranes are given in the ESI S9.†

5. Conclusions

Polysulfone hollow fiber membranes containing different compositions of PANI–TiO₂ nanocomposites were fabricated. The air gap distance was also varied during the fabrication of each composition. The PANI–TiO₂ content and air gap distance affected the morphology and permeation property of the membranes. With a higher content of PANI–TiO₂ nanocomposite and higher air gap distance, the finger-like projections in the membranes decreased. The antifouling ability and hydrophilicity of the membranes improved upon the addition of PANI–TiO₂. The P-1.0-A membrane showed a maximum rejection of 81.5% and 96.5% for the Reactive Black 5 and Reactive Orange 16 dyes, respectively. Thus, the PANI–TiO₂ nanocomposite PSf hollow fiber membranes can be used as potential candidates for dye removal from aqueous solutions by suitably varying the air gap distance and PANI–TiO₂ content.

Acknowledgements

A. M. I. is thankful to the Director of the NITK (National Institute of Technology, Karnataka), Surathkal, India for providing the required research facilities. The authors also thank the Vision group on Science & Technology, Government of Karnataka, India for the CESEM award.

References

1. C. Feng, K. Khulbe, T. Matsuura and A. Ismail, Sep. Purif. Technol., 2013, 111, 43–71.
2. Y. Chen, X. Hu, X. Hu, S. Zhang and Y. Zhang, Mater. Lett., 2015, 143, 315–318.
3. H. Tsai, H. Chen, K. Lee and J. Lai, Desalination, 2006, 193, 129–136.
4. K. C. Khulbe, C. Feng and T. Matsuura, Synthetic polymeric membranes: characterization by atomic force microscopy, Springer Science & Business Media, 2007.
5. M. Chanda and S. K. Roy, Industrial polymers, specialty polymers, and their applications, CRC Press, 2008.
6. F. Korminouri, M. Rahbari-Sisakht, T. Matsuura and A. Ismail, Chem. Eng. J., 2015, 264, 453–461.
7. J. Macanas, L. Ouyang, M. L. Bruening, M. Muñoz, J.-C. Remigy and J.-F. Lahitte, Catal. Today, 2010, 156, 181–186.
8. M. A. Salem, A. F. Al-Ghonemiy and A. B. Zaki, Appl. Catal., B, 2009, 91, 59–66.
9. V. R. Pereira, A. M. Isloor, A. Al Ahmed and A. Ismail, New J. Chem., 2015, 39, 703–712.
10. X. Wei, X. Kong, C. Sun and J. Chen, Chem. Eng. J., 2013, 223, 172–182.
11. M. Mondal and S. De, Chem. Eng. J., 2016, 285, 304–318.
12. Z. Liu, Y.-E. Miao, M. Liu, Q. Ding, W. W. Tjiu, X. Cui and T. Liu, J. Colloid Interface Sci., 2014, 424, 49–55.
13. W. Liu, J. Gao, F. Zhang and G. Zhang, Mater. Trans., 2007, 48, 2464–2466.
14. P. Roy, S. Berger and P. Schmuki, Angew. Chem., Int. Ed., 2011, 50, 2904–2939.
15. S. Debnath, N. Ballav, H. Nyoni, A. Maity and K. Pillay, Appl. Catal., B, 2015, 163, 330–342.
16. W.-H. Jeong, T. Amna, Y.-M. Ha, M. S. Hassan, H.-C. Kim and M.-S. Khil, Chem. Eng. J., 2014, 246, 204–210.
17. S. Razak, M. Nawi and K. Haitham, Appl. Surf. Sci., 2014, 319, 90–98.
18. A. J. Kajekar, B. Dodamani, A. M. Isloor, Z. A. Karim, N. B. Cheer, A. Ismail and S. J. Shilton, Desalination, 2015, 365, 117–125.
19. B. G. Lokesh, K. S. V. K. Rao, K. M. Reddy, K. C. Rao and P. S. Rao, Desalination, 2008, 233, 166–172.
20. A. Zulhairun and A. Ismail, J. Membr. Sci., 2014, 468, 20–30.
21. H. Dzinun, M. H. D. Othman, A. F. Ismail, M. H. Puteh, M. A. Rahman and J. Jaafar, Chem. Eng. J., 2015, 269, 255–261.
22. J. Ren and J. R. McCutcheon, Desalination, 2015, 372, 67–74.
23. S. B. Teli, S. Molina, A. Sotto, E. G. a. Calvo and J. d. Abajob, Ind. Eng. Chem. Res., 2013, 52, 9470–9479.
24. R. Kumar, A. Ismail, M. Kassim and A. M. Isloor, Desalination, 2013, 317, 108–115.
25. K. Khulbe, C. Feng, T. Matsuura, G. Kapantaidakis, M. Wessling and G. Koops, J. Membr. Sci., 2003, 226, 63–73.
26. S. Zhuang, Y. Li, M. Zuo, X. Tan, B. Meng, N. Yang and S. Liu, Sep. Purif. Technol., 2014, 132, 712–718.
27. A. Akbari, S. Desclaux, J. Rouch and J. Remigy, J. Membr. Sci., 2007, 297, 243–252.
28. Y. Zhang and R. Wang, J. Membr. Sci., 2014, 452, 379–389.
29. R. Sengur, C.-F. de Lannoy, T. Turken, M. Wiesner and I. Koyuncu, Desalination, 2015, 359, 123–140.
30. H. Dzinun, M. H. D. Othman, A. Ismail, M. H. Puteh, M. A. Rahman and J. Jaafar, J. Membr. Sci., 2015, 479, 123–131.
31. H. Shi, F. Liu and L. Xue, J. Membr. Sci., 2013, 437, 205–215.
32. T. S. Chung and X. Hu, J. Appl. Polym. Sci., 1997, 66, 1067–1077.
33. S. M. Ahmed, F. I. El-Dib, N. S. El-Gendy, W. M. Sayed and M. El-Khodary, Arabian J. Chem., 2012 DOI:10.1016/j.arabjc.2012.04.049.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra18682c

---