PVDF mixed matrix nano-filtration membranes integrated with 1D-PANI/TiO₂ NFs for oil–water emulsion separation

Abstract

The treatment of directly discharged oily waste water is difficult because of colloidal stability and the deformable nature of emulsified oil. In this work, one dimensional (1D) PANI/TiO₂ nanofibers (NFs) were incorporated into polyvinylidene fluoride (PVDF) Mixed Matrix Membranes (MMMs) for the removal of oil and water droplets from oily waste water. The size of the 1D PANI/TiO₂ (polyaniline/titanium dioxide) NFs ranged between 60 to 75 nm, as identified using transmission electron microscopy (TEM). Atomic force microscopy (AFM) analysis also revealed the significant decrease in the surface roughness of the membranes from 186 to 36 nm following the addition of 1D PANI/TiO₂ NFs. Due to the hydrophilic property of the 1D PANI/TiO₂ NFs, the contact angle, pure water flux and antifouling properties were increased for 1D PANI/TiO₂ NF MMMs compared to the neat PVDF membrane. The pure water flux increased from 80 to 132 L m⁻² h⁻¹, which clearly indicates the impact of addition of the 1D PANI/TiO₂ NFs. Among the MMMs, PT-4 exhibited the maximum oil rejection of 99% at a 5 bar operating pressure. Hence, the incorporation of PANI/TiO₂ NFs in MMMs is proposed to enhance the oil rejection and fouling resistance of MMMs.

---

1. Introduction

An enormous volume of oily waste water is generated every day in domestic life as well as in industrial activities such as, petrochemical, chemical, textile, leather, food, and steel processing and metal finishing. The direct discharge of oily waste water causes terrible environmental pollution, waste of resources and health problems.¹ Conventional treatment strategies chiefly comprise gravity separation and skimming, de-emulsification, air flotation, coagulation, and sedimentation in a centrifugal field and in hydrocyclones.^2–5 These methods have several intrinsic disadvantages such as, low efficiency, high cost, operational difficulties, corrosion and recontamination problems.⁶

Nowadays, membrane separation processes such as, ultrafiltration (UF), nanofiltration (NF) and reverse osmosis (RO) are gaining popularity as the most reliable methods for oil–water separation.^7,8 UF has been validated as an effective pretreatment step prior to NF and RO in the treatment of oily water owing to its suitable pore sizes (generally in the range of 2–50 nm) and its ability to remove emulsified oil droplets without any de-emulsification steps.^9,10 However, fouling is a major problem which occurs in NF membranes either due to the deposition of oily droplets on the surface of the membrane or blocking of the membrane pores by oil droplets, resulting in a significant decline in membrane flux.¹¹ Recently, mixed matrix membranes, (MMMs) containing nanoparticles mixed with a polymeric membrane have demonstrated an improvement in performance in terms of permeability, solute rejection and better anti-fouling properties.^12,13 Polyvinylidene fluoride (PVDF) exhibits excellent properties such as chemical resistance, thermal stability, high organic selectivity, as well as good mechanical and membrane separation properties. Among the various inorganic nanomaterials used in membranes such as, Al₂O₃, ZrO₂, ZnO and Fe₂O₃, titania (TiO₂) owns several advantages including super hydrophilicity, higher specific surface area, photocatalytic activity, bactericidal and self-cleaning properties and robustness in both thermal and chemical environments.¹⁴

Membrane modification influences the shape, size and dispersion of nano materials. Recently, one dimensional (1D) titanium dioxide (TiO₂) nanostructures (nanotubes, nanorods and nanofibers) have drawn both theoretical and technological interest.^15,16 In recent years, nanohybrid materials with at least two different nanomaterials have gained interest, where the benefits of the two different particles could be combined to gain an improvement in the properties.¹⁷ Besides inorganic nanoparticles, certain functional groups of organic nanomaterials have demonstrated novel performances due to their larger surface areas than bulk particles.

PANI is a distinguished polymer possessing versatile advantages such as, ease of preparation, high conductivity, chemical stability, low cost and good separation characteristics.¹⁸ PANI nanofibers possess a high surface energy and hydrophilic properties, because of which they are used to achieve super hydrophilic surfaces.¹⁹ PANI has been synthesized and demonstrated for the formation of membranes by a non-solvent induced phase inversion.²⁰ The increase in hydrophilicity by addition of polyaniline into polysulfone membranes has improved the performance and lowered fouling.²¹ The functionalization of TiO₂ with PANI appears to be a key solution to access an enhanced hydrophilicity and alter the surface properties for oil/water separation with better antifouling properties. Hence, TiO₂ is considered as one of the best materials to be functionalized with PANI.²² Teli et al.²³ developed polysulfone PANI/TiO₂ ultrafiltration nanocomposite membranes, which revealed an improved hydrophilicity of the membrane surface and a better antifouling property owing to the addition of PANI/TiO₂.

Hence, in this study, PVDF membranes impregnated with 1D PANI/TiO₂ NFs were synthesized by a chemical oxidation method. The membranes were characterized using thermogravimetric analysis, mechanical strength, contact angle, BET analysis, surface topography, and membrane morphology analyses. The outcome of the addition of 1D PANI/TiO₂ NFs on the performance properties of MMMs, such as the permeability, hydrophilicity, pore volume and antifouling properties were investigated. To the best of our knowledge, there is a lack of investigation of the effect of the size and dispersion of one-dimensional nanoparticles on the efficacy of modification of NF membranes. Hence, in this work, for the first time, 1D PANI/TiO₂ NFs have been employed as hydrophilic additives to improve the performance of PVDF membranes for oil–water separation.

2. Materials and methods

2.1 Materials

Commercial PVDF polymer pellets (Kynar® 740) were supplied by Arkema Inc., PA, USA. Aniline (99% pure) and sodium dodecyl sulphate (SDS) were purchased from Merck, India. Ammonium peroxydisulfate (APS) and hydrochloric acid were procured from Finer, India. Titanium butoxide (TNBT, 99%) was purchased from Alfa aesar, India. N-Methyl pyrrolidone, polyvinyl pyrrolidone (PVP, M.W = 1 300 000) and bovine serum albumin (BSA) (66.5 kDa) were procured from M/s. Merck chemical, India limited.

2.2 Fabrication of 1D TiO₂ NFs

1.6 g of TNBT was dissolved in 15 mL of ethanol under continuous stirring for 10 h. Then, 1.2 g of PVP was mixed with ethanol with stirring. The above composite solutions were fabricated by using an electrospinning method. The solution was constantly introduced through a stainless steel needle at a flow rate of 0.3 mL h⁻¹ onto a Si wafer (2 cm × 2 cm) placed on a grounded substrate at a high DC voltage (13 kV) and at a fixed distance of 10 cm between the needle and the grounded substrate. The PVP/TNBT NFs on the Si wafer were sintered at 300 °C for 5 h.

2.3 Synthesis of 1D PANI/TiO₂ NFs

1D PANI/TiO₂ NF nanoparticles were prepared by an in situ chemical oxidation polymerization method with aniline using ammonium peroxydisulfate (APS) as an oxidant in the presence of colloidal 1D TiO₂ NFs at 0–5 °C in air. In a typical procedure, the TiO₂ particles were suspended in 1 M HCl solution and sonicated for 1 h to reduce the aggregation of 1D TiO₂ NFs. Aniline (0.1 M) was then mixed with the sonicated colloidal 1D TiO₂ NFs, followed by further sonication for 30 min. An equal molar ratio of APS was dissolved in 1 M HCl and added dropwise to a solution containing the aniline monomer under constant stirring. The mixture was then allowed to polymerize under stirring for 5 h at room temperature. Consequently, the reaction mixture was vacuum filtered, washed several times with ethanol and water and dried at 80 °C for 24 h.

Fig. 1 is the schematic representation of PANI/TiO₂. When the oxygen molecules are adsorbed onto the surface layer of the TiO₂ NFs, they generate O₂⁻, O⁻ and O²⁻ ions which create a positive charge on the surface of the TiO₂ grains by extracting electrons from the conduction band of TiO₂.^24–27 When TiO₂ is added to the aniline solution with APS as an oxidant, anilinium cations are formed from the aniline monomers and display an electrostatic interaction with the adsorbed anions on the surface of TiO₂. When the aniline monomers polymerize, PANI partially covers the surface of TiO₂.

Fig. 1 Scheme of the prepared PANI/TiO₂ nano composite.

2.4 Characterization of 1D PANI/TiO₂ NFs

The prepared PANI/TiO₂ composites were characterized by an X-ray diffractometer (XRD, Rigaku Corporation, Japan) with Cu Kα radiation. Fourier transform infrared (FTIR) spectra were recorded on SHIMADZU ATIR-FTIR mode. A transmission electron microscope (FEI Technai F20 S, USA) was used to analyze the morphology of the PANI/TiO₂ composites.

2.5 Preparation of PVDF–PANI/TiO₂ NF MMMs

The PVDF polymer pellets were dried to remove the moisture content in a vacuum oven for 24 h at 80 °C. The phase inversion method was used to synthesize composite membranes. PANI/TiO₂ NFs were mixed with NMP solvent and sonicated for 1 hour to ensure a uniform dispersion of the material. Then, PVDF polymer was added to the casting dope and stirred continuously for 12 h. This resulted in a homogenous solution and the dope solution was cast on a smooth glass plate with the help of thin film applicators, followed by evaporation. The resultant thin film was then immersed in a non-solvent water coagulation bath for 24 h at 10 °C. The obtained membranes were preserved for further studies. The composition of the prepared membrane is provided in Table 1.

Table 1 Composition and mechanical strength of the prepared membranes

Membrane description	PVDF (wt%)	PANI TiO2 (wt%)	NMP (wt%)	Elongation break (mm)	Tensile stress at break (MPa)
a Average value.
PT-1	17.5	0	82.5	1.2 (0.42)^a	6.4 (1.6)^a
PT-2	17.48	0.02	82.5	1.9 (0.18)^a	5.9 (2.5)^a
PT-3	17.1	0.4	82.5	2.3 (0.29)^a	4.8 (2.1)^a
PT-4	16.5	1.0	82.5	2.9 (0.34)^a	4.3 (1.8)^a

---

2.6 Characterization of PVDF–PANI/TiO₂ NF MMMs

2.6.1 Surface roughness analysis of PVDF–PANI/TiO₂ NF MMMs. The surface of the prepared membranes was imaged in tapping mode using atomic force microscopy (NT-MDT modular AFM, NTEGRA Prima, Ireland). To compare the roughness of PVDF and PVDF–PANI/TiO₂ NF MMMs, high resolution three dimensional images were captured. The images were then processed using Nova-NTEGRA AFM software with a sample scan area of 30 μm × 30 μm so as to compute the average roughness (R_a) and root mean square roughness (R_q). The hydrophilicity of the synthesized PVDF–PANI/TiO₂ NF MMMs can be inferred from the measurement of the membrane roughness, which varies linearly with the membrane permeability.

2.6.2 Thermal behavior of PVDF–PANI/TiO₂ NF MMMs. Thermal properties of neat PVDF and 1D PANI/TiO₂NF MMMs were analyzed using a thermogravimetric analyser (TGA) coupled with a differential thermogravimetric analyser (DTGA) (Model TGA 4000, Perkin Elmer, USA). 10 to 20 mg of the membrane sample was dried in a vacuum oven at 180 °C overnight to remove moisture before analysis. The thermogram was then recorded at a temperature range of 35 to 800 °C at a heating rate of 10 °C min⁻¹ under nitrogen gas. The thermal decomposition and corresponding loss in wt% were thus examined and compared. TGA measurements were performed by maintaining 10 min of isothermal conditions at the final stage along with a similar temperature range for DTGA.

2.6.3 Mechanical strength of PVDF–PANI/TiO₂ NF MMMs. The mechanical strength of the prepared membranes was determined using a uni-axial mechanical testing machine (Instron, Canton MA). The tensile strength and elongation ratio of the neat PVDF and 1D PANI/TiO₂ NF MMMs were studied; the obtained values for each sample were tested five times and then averaged for reliability.

2.6.4 BET surface area and pore volume of PVDF–PANI/TiO₂ NF MMMs. The BET surface area and t-plot micropore volume of the prepared MMMs were analyzed. Nitrogen adsorption–desorption isotherms were measured by a micromeritics Gemini V, USA. Brunauer Emmett Teller (BET) surface area analyzer and the t-plot pore volume was used in order to characterize the nano-scale textural properties of the membrane surface.

2.6.5 Contact angle measurement of PVDF–PANI/TiO₂ NF MMMs. The hydrophilicity of the membranes was determined by measuring the water contact angle of the membranes by the sessile droplet method using a dynamic contact angle analyzer instrument. A deionized water droplet was placed on each sample at different locations and the average contact angle was taken into account.

2.6.6 Morphology of PVDF–PANI/TiO₂ NF MMMs. The morphology of the prepared PVDF–PANI/TiO₂ NF MMMs was studied using a Field Emission Scanning Electron Microscope, (FESEM) Carl Zeiss Microscopy Ltd, UK & SIGMA. The morphology of the membrane cross-sections was studied before gold sputtering for the imparting conductivity.

2.6.7 Pure water flux of PVDF–PANI/TiO₂ NF MMMs. The pure water flux (PWF) measurements were carried out in a dead-end ultrafiltration unit consisting of a 400 mL batch type stirred cell (model Cell XFUF076, Millipore USA) fitted with a Teflon coated magnetic paddle. An effective membrane area of 5 cm² was considered for permeation analyses. The membranes were immersed in water for 24 h prior to the permeation experiments. The membranes were then subjected to compaction at 0.5 MPa transmembrane pressures (TMP) for 30 min, followed by a reduced pressure of 0.2 MPa TMP. The time dependent pure water flux of the sample membranes was measured at 0.2 MPa TMP at 25 °C at every 1 min time interval. The pure water flux of the membranes was calculated using eqn (1):where J_w1 is the pure water flux expressed in L m⁻² h⁻¹, Q indicates the quantity of pure water collected (L) in time Δt (h), and A denotes the effective membrane area (m²).

2.6.8 Antifouling study of PVDF–PANI/TiO₂ NF MMMs. Bovine serum albumin (BSA) was used as a model protein to investigate the antifouling nature of PVDF–PANI/TiO₂ NF MMMs using an UF stirred dead end flow cell (model Cell XFUF076, Millipore, USA). A 0.8 g L⁻¹ aqueous solution of BSA was placed in a filtration cell following the initial pure water flux J_w1 (L m⁻² h⁻¹) measurement. BSA solution was then filtered through the membrane samples for 30 min at 0.2 MPa TMP. Then, the fouled membranes were washed with water for 20 min and the flux J_w2 (L m⁻² h⁻¹) was determined again. To evaluate the resistance of the membranes to fouling, the flux recovery ratio (FRR) was calculated by the following eqn (2):

2.7 Treatment of oil in water emulsion of PVDF–PANI/TiO₂ NF MMMs

0.1 g of castrol oil and 0.05 g of SDS were added to 500 mL of deionized water and stirred at 500 rpm and room temperature for 5 hours to obtain a steady oil in water emulsion. The membranes were cut to sizes of 3.8 cm and the permeation of the oil–water emulsion was performed using a dead end ultrafiltration cell at 5 bar operating pressure. Both the retentate and permeate were collected and the concentration of oil in water was measured by UV-Vis spectroscopy at 270 nm.²⁸ The rejection of oil was determined by the following eqn (3):where C_f and C_p are the concentrations of oil in the feed and permeate.

3. Results and discussion

3.1 Size and crystallite analysis of 1D PANI/TiO₂ NFs

X-Ray diffraction patterns of the PANI/TiO₂ nanocomposite, compared with pure TiO₂ and PANI/TiO₂ are shown in Fig. 2. Both anatase and brookite phases are present in the XRD diffraction pattern of the crystalline structure of TiO₂. The TiO₂ anatase (JCPDS no. 21-1272) and brookite (JCPDS no. 35-0088) have some common peaks. The 1D PANI/TiO₂ NF diffraction peaks at 2θ values of 20.5°, 21.2°, 24°, 28.4° and 29.2° correspond to the characteristic peaks observed in PANI,^29,30 and also TiO₂ peaks are observed which are confirmed in the 1D PANI/TiO₂ NFs. The FTIR spectra of TiO₂ and PANI–TiO₂ NFs were obtained in the range of 500–3700 cm⁻¹. From Fig. 3, TiO₂ shows a broad peak at around 3400 cm⁻¹ and a small peak at 1638 cm⁻¹ which is due to the stretching vibration of the hydroxyl group (–OH) and also the adsorbed water present.³¹ A broad peak is observed at 600 cm⁻¹ ranging from 1000 to 550 cm⁻¹. The broad peak at 600 cm⁻¹ is due to the Ti–O–Ti bonds. The peaks at 1577 and 1485 cm⁻¹ corresponded to the quinoid and benzenoid rings of PANI. The peaks at 1297 to 1233 cm⁻¹ correspond to the C–N stretching mode of the benzoid ring. The peaks at 1120 to 1098 cm⁻¹ are assigned to a plane bending vibration of the C–H mode which is found during protonation.³² The TiO₂ strong absorption around 670 cm⁻¹ is due to Ti–O stretching, this peak is weak in the 1D PANI/TiO₂ NFs due to the presence of PANI.³³ The 1D-PANI/TiO₂ NFs were synthesized by a chemical oxidation method and the TEM images are shown in Fig. 4. The average size of the PANI/TiO₂ was 60–75 nm in the exhibited NF structure.

Fig. 2 1D PANI/TiO₂ NF XRD diagram.

Fig. 3 FTIR spectra of PANI/TiO₂ and TiO₂.

Fig. 4 TEM images of 1D PANI/TiO₂ NFs.

3.2 Surface topography of PVDF–PANI/TiO₂ NF MMMs

The topography of the prepared membranes was studied by using an AFM instrument in tapping mode and the three dimensional images of PANI/TiO₂ NF MMMs are shown in Fig. 5. The surface roughness of the MMMs was evaluated on different parts of the same membrane and the mean value was considered. The brightest areas in the AFM images represent the nodules of the MMMs surface and the dark regions elucidate valleys or membrane pores. The surface roughness of the membrane was decreased by the addition of PANI/TiO₂ NFs, as illustrated in Table 2. The decrease in roughness is because of the dispersion of PANI/TiO₂ NFs causing a reduction of the surface pore size on the polymer matrix.^34,35 Consequently, the PT-4 membrane with the highest concentration of PANI/TiO₂ NFs showed the lowest roughness of 27 nm while the maximum roughness of 186 nm was observed for the PT-1 membrane.

Fig. 5 Three dimensional AFM images of PVDF–PANI/TiO₂ NF MMMs.

Table 2 Surface roughness, thermal stability, BET surface area and t-plot pore volume of PVDF–PANI/TiO₂ MMMs

Membrane description	Average roughness (nm)	Rms roughness (nm)	Td (°C)	Weight loss (%)	BET surface area, m^2 g^−1	t-Plot pore volume (pore volume between 1.7 and 80 nm) cm^3 g^−1
PT-1	186 ± 18	225 ± 07	438	97	0.8930	0.005861
PT-2	55 ± 09	57 ± 11	456	84	1.7021	0.003379
PT-3	37 ± 12	52 ± 08	458	72	4.9334	0.006260
PT-4	27 ± 14	34 ± 15	459	63	5.6330	0.004534

---

3.3 Thermal behavior of PVDF–PANI/TiO₂ NF MMMs

The thermal behaviour of the thermogravimetric analysis (TGA) and differential TGA (DTGA) of neat PVDF and the MMMs is shown in Fig. 6 and 7 and values are indicated in Table 2. When the membranes were heated from 35 °C to 800 °C, the neat PVDF membrane exhibited a significant weight loss compared to the MMMs. The thermal decomposition of the MMMs increased from 438 °C, 456 °C, 458 °C and 459 °C, respectively for membranes PT-1, PT-2, PT-3 and PT-4. The weight loss% of the MMMs following the addition of PANI/TiO₂ NFs is shown in Table 2. The addition of PANI/TiO₂ NFs in the MMMs has resulted in a loss in the thermal decomposition temperatures for PT-3 and PT-4. PT-3 and PT-4 have a little more weight loss then the PT-1 membrane, which is due to the loss of HCl doped in the PANI polymer chain.³⁶ The addition of PANI/TiO₂ NFs in MMMs in the above TGA values also conformed with the DTGA analysis.

Fig. 6 Thermogravimetric analysis of PVDF–1D PANI/TiO₂ NF MMMs.

Fig. 7 DTGA of PVDF–PANI/TiO₂ NF MMMs.

3.4 Mechanical strength of PVDF–PANI/TiO₂ NF MMMs

Table 1 shows the elongation ratio and tensile strength for neat PVDF and 1D PANI/TiO₂ NF MMMs. Pristine PVDF membrane hold a higher tensile strength of 6.4 Mpa and it is decreased up to 4.3 Mpa for PT-4 MMMs. It is interesting to note that tensile strength decreases with loading concentration of PANI/TiO₂ NFs. It indicates that the PANI/TiO₂ NFs have blended effectively with the PVDF membrane. Interestingly, the elongation ratio decreased with the addition of PVDF–PANI/TiO₂ NF membranes. It shows that the surface porosity was improved due to the addition of hydrophilic PANI/TiO₂ material, which moved towards the membrane surface during membrane formation. It causes an increase of the surface pores on the membrane surface. This study indicates that PANI/TiO₂ has a good miscible property with the PVDF membrane.

3.5 BET surface area and pore volume of PVDF–PANI/TiO₂ NF MMMs

The membrane polymer surface area and t-plot pore volume was studied using nitrogen adsorption and desorption measurements. The results are reported in Table 2. In terms of the t-plot a pore volume in the range of 1.7–80 nm was considered for the mesopore structure. The surface area of PT-1, PT-2, PT-3 and PT-4 was 0.8930 m² g⁻¹, 1.7021 m² g⁻¹, 4.9334 m² g⁻¹ and 5.6330 m² g⁻¹ respectively. With the addition of 1D-PANI/TiO₂ NFs in PT-2, PT-3 and PT-4 MMMs the surface area is increased. Therefore, the surface area increased as a result of the addition of 1D-PANI/TiO₂ NFs in the MMMs. The t-plot pore volume of MMMs (0.005861 to 0.004534 cm³ g⁻¹) is increased by the addition of the 1D-PANI/TiO₂ NFs.

3.6 Contact angle of PVDF–PANI/TiO₂ NF MMMs

The contact angles of the prepared membranes were determined to deduce the hydrophilicity and wettability as shown in Fig. 8. PT-1 had the highest contact angle because of its hydrophobic nature due to the absence of 1D-PANI/TiO₂ NFs. The contact angle decreased with increasing 1D-PANI/TiO₂ NF concentration in the MMMs, which confirms that the 1D-PANI/TiO₂ NFs could enhance the hydrophilicity of the membranes. The presence of TiO₂ particles and PVDF usually create hydroxyl groups on the membrane surface. These polar groups can in turn interact with water molecules through van der Waals’ forces and hydrogen bonds.^37–39 Thus, it could mean that the water molecule is more favorably adsorbed then permeated through the membrane. In this study, the membrane PT-4 with the highest concentration of 1D-PANI/TiO₂ NFs will clearly have the highest hydrophilicity compared to the other membranes. Such an increase in hydrophilicity is beneficial to reduce membrane fouling, especially organic fouling.⁴⁰ The hydrophilic and more wettable surface layer typically yields a better water permeability and fouling resistance in water purification applications.⁴¹

Fig. 8 Contact angle images of PVDF–PANI/TiO₂ NF MMMs.

3.7 SEM analysis of PVDF–PANI/TiO₂ NF MMMs

The top surface and cross sectional morphology of the PVDF membrane and MMMs investigated by field emission scanning electron microscopy are shown in Fig. 9 and 10, respectively. A significant difference in the pore sizes among PT-1, PT-2, PT-3 and PT-4 membranes was observed and the fine pore structure was visualized for the membrane cast with 1D-PANI/TiO₂ NFs as compared to the other membranes. Fig. 9 clearly shows the distribution of the nanocomposite filler in the PVDF membrane surface. In addition, it can be seen that the introduction of nanoparticles into the PVDF solution has increased the pore size and number of pores on the membrane top surface. The PANI/TiO₂ nanoparticle surface can promote the PVDF network formation and formation of pores, which is caused by the retort between the amine groups in PANI on the TiO₂ nanoparticle surface and the PVDF. The PANI/TiO₂ nanoparticle surface resulted from the formed covalent bonds (the stable balance of attractive forces between atoms) between the PVDF and the amine groups in PANI.⁴² Furthermore, the increased pore size resulted from the membrane surface between the nanomaterial and PVDF, which provided an enhancement of the pores in the surface for the PVDF polymer chain segments. In Fig. 10, neat PVDF (PT-1) shows an upper surface with open membrane voids and bottom surface without pores due to the absence of 1D-PANI/TiO₂ NFs. However, PT-2, PT-3 and PT-4 MMMs exhibit formation of an upper macrovoid layer with open pores and the presence of the pores in the bottom surface is also due to the addition of 1D-PANI/TiO₂ NFs. From these results, it can be proposed that 1D-PANI/TiO₂ NFs can act as pore forming agents due to their hindrance effect in the interphase polymer solvent during the membrane formation process.⁴³ Moreover, the addition of 1D-PANI/TiO₂ NFs was done by blending and hence, nanoparticles were almost stably embedded in the membranes. This could have resulted in the formation of pores and opening of pores in the bottom surface and an increase in the membrane surface pore size and porosity.

Fig. 9 Top surface of PVDF–PANI/TiO₂ NF MMMs.

Fig. 10 Cross sectional image of PVDF–PANI/TiO₂ NF MMMs.

3.8 Pure water flux of PVDF–PANI/TiO₂ NF MMMs

The pure water flux (PWF) of the prepared membranes is provided in Fig. 11. PVDF–PANI/TiO₂ MMMs were found to have a higher flux compared to the neat PVDF membrane due to the addition of 1D-PANI/TiO₂ NFs. The presence of 1D-PANI/TiO₂ NFs in MMMs has endowed a better hydrophilicity, higher porosity and larger surface area, and resulted in void formation. The inorganic or organic additives are generally used to produce macrovoids, improve pore connectivity, and increase surface hydrophilicity.⁴⁴ PT-2, PT-3 and PT-4 membranes showed better PWF due to the addition of 1D-PANI/TiO₂ NFs, which enhanced the hydrophilicity of the membranes. The water flux of the neat PVDF membrane was about 88 L m⁻² h⁻¹, while the water flux of the PT-4 membrane was as high as 132 L m⁻² h⁻¹. In general, an inorganic material such as TiO₂ is used as a bulk material. The flux enhancement can be easily attributed to the interaction between the well-distributed nanoparticles and an enhanced interaction is likely to be expected for the polymer/nanoparticle combinations.⁴⁵

Fig. 11 Permeate flux of pure water, BSA and oil for PVDF–PANI/TiO₂ NF MMMs.

3.9 Membrane performance

3.9.1 Permeate flux of PVDF–PANI/TiO₂ NF MMMs. Fig. 11 shows the changes in permeate flux of BSA and the oil/water solution through the neat PVDF and 1D-PANI/TiO₂ NF MMMs. The permeate flux of the BSA solution was higher for all the membranes compared to the permeate flux of the oil/water solution. The reason for the lower flux of the oil/water solution was the deposition of an oil layer on the membrane surface or the membrane pore blocking by oil droplets.⁴⁶ The permeate flux also increased in the order PT-4 > PT-3 > PT-2 > and PT-1 and followed the addition of 1D-PANI/TiO₂ NFs. The density of the oil is higher than that of water, and thus oil settles below water and forms a barrier on the membrane surface preventing water permeation. Overall, the hydrophilic and porous properties of 1D-PANI/TiO₂ NF incorporated PVDF membranes can facilitate the selective permeation of water from oil/water mixtures.

3.9.2 Flux recovery ratio of PVDF–PANI/TiO₂ NF MMMs. Fig. 12 shows the flux recovery ratio (FRR) value for the BSA and oil–water filtration. Among the prepared MMMs, FRR values was the lowest for the PT-1 membrane. The FRR ratio also increased in the order PT-4 > PT-3 > PT-2 > and PT-1 by addition of 1D-PANI/TiO₂ NFs. The PT-4 membrane showed an excellent flux recovery ratio and better flux rate. The interaction between the membrane surface and protein molecules was weakened due to increased addition of 1D-PANI/TiO₂ NFs and the fouling resistance of the membrane was also enhanced. The permeate flux and oil rejection of the prepared membranes with respect to time were calculated by performing UF in the dead end mode, at different oil concentrations. Moreover, the PT-4 membrane performance was evaluated at different concentrations of oil–water and the flux pattern is shown in Fig. 13. The excellent performance of the membrane in eliminating oil droplets from water is primarily due to the addition of hydrophilic nanomaterials. In general, when the average oil droplet size is smaller than the mean porous radius of the membrane, an excellent separation performance is promised. With the addition of 1D-PANI/TiO₂ NFs, the PT-4 membrane had become highly hydrophilic with a reduced pore plugging as well as oil layer deposition on the membrane surface.

Fig. 12 Flux recovery ratio of BSA and oil for PVDF–PANI/TiO₂ NF MMMs.

Fig. 13 Different oil concentrations for PVDF–PANI/TiO₂ NF MMMs.

3.10 Oil rejection of PVDF–PANI/TiO₂ NF MMMs

The oil rejection by the prepared membranes is shown in Fig. 14. With the increased addition of 1D-PANI/TiO₂ NFs in MMMs, the rejection also increased. The PT-4 MMM presented an excellent antifouling phenomenon due to the addition of the highest concentration of 1D-PANI/TiO₂ NFs to the membrane surface. The results obtained clarify the role of the hydrophilic nature of the 1D-PANI/TiO₂ NFs in enhancing the fouling resistance of the membrane. The wetting behavior of water on the membranes purely depends on the surface properties of the membranes. In the case of the PT-4 membrane, the highest oil/water permeate flux was observed, thus indicating an excellent hydrophilic property due to the addition of the 1D-PANI/TiO₂ NFs. The hydrophilic property ensures a water layer formation on the membrane surface and avoids direct contact between the oil and membrane surfaces during oil/water rejection.⁴⁷ It could have greatly supported the flux recovery of the PT-4 membrane and its antifouling performance. The oil/water solution made contact with the membrane surface with an alteration of oil and some amount of oil and water was then permeated. 1D-PANI/TiO₂ NF incorporated membranes have a detectable adhesion due to the high hydrophilicity during the permeation process, thus 99% rejection was observed.

Fig. 14 Oil rejection of the PVDF–PANI/TiO₂ NF MMMs.

4. Conclusion

This study reports a novel approach of including PVDF–1D PANI/TiO₂ NFs on nanofiltration membranes, which were then investigated for the separation of oil from synthetic oil-in-water emulsions. The synthesized 1D-PANI TiO₂ NFs had an average size of 60–75 nm and were successfully incorporated in the PVDF MMMs. The existence of 1D-PANI/TiO₂ NFs on the PVDF membranes was confirmed by BET analysis, mechanical strength, contact angle measurements, surface topography and thermogravimetric analyses. The membrane morphology was thus inferred to be altered with the addition of 1D-PANI/TiO₂ NFs. The addition of the PANI/TiO₂ NFs into the dope solution was found to play an important role in increasing the membrane hydrophilicity and enhancing the pure water flux and oil rejection. The PT-4 membrane, with the maximum addition of 1D-PANI/TiO₂ NFs in the MMMs, showed 99% oil rejection and an excellent antifouling behavior. Hence, the addition of 1D-PANI/TiO₂ NFs is proposed to improve the fouling resistance of MMMs for oil water emulsions.

References

1. P. Kajitvichyanukul, Y. T. Hung and L. K. Wang, Membrane and desalination technologies, in Handbook of Environmental Engineering, ed. L. K. Wang, J. P. Chen, Y.-T. Hung and N. K. Shammas, The Humana Press Inc., New York, 2011, vol. 13, p. 639.
2. M. Cheryan and N. Rajagopalan, J. Membr. Sci., 1998, 151, 13–28.
3. A. A. Al-Shamrani, A. James and H. Xiao, Colloids Surf., A, 2002, 209, 15–26.
4. A. Cambiella, J. M. Bentio, C. Pazos and J. Coca, Chem. Eng. Res. Des., 2006, 84, 69–76.
5. K. A. Hashmi, H. A. Hamza and J. C. Wilson, Miner. Eng., 2004, 17, 643–649.
6. A. Rezvanpour, R. Roostaazad, M. Hesampour, M. Nystrom and C. Ghotbi, J. Hazard. Mater., 2009, 16, 1216–1224.
7. R. J. Peterson, J. Membr. Sci., 1993, 83, 81–150.
8. B. D. Freeman and I. Pinnau, ACS Symp. Ser., 2004, 876, 1–23.
9. C. V. Vedavyasan, Desalination, 2007, 203, 296–299.
10. Y. Yang, H. Wang, J. Li, B. He, T. Wang and S. Liao, Environ. Sci. Technol., 2012, 46, 6815–6821.
11. N. Hilal, O. O. Ogunbiyi, N. J. Miles and R. Nigmatullin, Sep. Sci. Technol., 2005, 1957, 1957.
12. A. Sotto, J. Kim, J. M. Arsuaga, G. del Rosario, A. Martinez, D. Nam, P. Luis and B. Van der Bruggen, J. Mater. Chem. A, 2014, 2, 7054.
13. R. Saranya, G. Arthanareeswaran, S. Sakthivelu and P. Manohar, Ind. Eng. Chem. Res., 2012, 51, 4942.
14. X. Chen and S. S. Mao, Chem. Rev., 2007, 107, 2891–2959.
15. P. Poudel and Q. Q. Qiao, Nanoscale, 2012, 4, 2826.
16. P. D. Cozzoli and A. Kornowski, J. Am. Chem. Soc., 2003, 125, 14539.
17. H. Wu, B. Tang and P. Wu, J. Membr. Sci., 2014, 451, 94–102.
18. S. B. Teli, S. Molina, E. G. Calvo, A. E. Lozano and J. de Abajo, Desalination, 2012, 299, 113–122.
19. Z. Fan, Z. Wang, N. Sun, J. Wang and S. Wang, J. Membr. Sci., 2008, 320, 363–371.
20. G. R. Guillen, T. P. Farrell, R. B. Kaner and E. M. V. Hoek, J. Mater. Chem., 2010, 20, 4621–4628.
21. S. J. Oh, N. Kim and Y. T. Lee, J. Membr. Sci., 2009, 345, 13–20.
22. Y. Li, Y. Yu, L. Wu and J. Zhi, Appl. Surf. Sci., 2013, 273, 135–143.
23. S. B. Teli, S. Molina, E. G. Calvo, A. E. Lozano and J. de Abajo, Desalination, 2012, 299, 113–122.
24. C. Xiaobo and S. S. Mao, Chem. Rev., 2007, 107, 2891–2906.
25. U. Diebold, Surf. Sci. Rep., 2003, 229, 53–229.
26. R. H. Gonc-alves, W. H. Schreiner and E. R. Leite, Langmuir, 2010, 26(14), 11657–11662.
27. A. Katoch, M. Burkhart, T. Hwang and S. S. Kim, Chem. Eng. J., 2012, 192, 262–268.
28. T. Rajasekhar, M. Trinadh, P. V. Babu, A. V. Sesha Sainath and A. V. R. Reddy, J. Membr. Sci., 2015, 481, 82–93.
29. B. A. Bhanvase and S. H. Sonawane, Chem. Eng. J., 2009, 156, 177–183.
30. J. Jiang and L. Hong Ai, Appl. Phys. A: Mater. Sci. Process., 2008, 2, 341–344.
31. F. Cheng, S. M. Sajedin, S. M. Kelly, A. F. Lee and A. Kornherr, Carbohydr. Polym., 2014, 114, 246–252.
32. E. T. Kang, K. G. Neoh and K. L. Tan, Prog. Polym. Sci., 1998, 23, 277–324.
33. J. C. Xu, W. M. Liu and H. L. Li, Mater. Sci. Eng., C, 2005, 25, 444–447.
34. N. Ghaemi, S. S. Madaeni, A. Alizadeh, H. Rajabi and P. Daraei, J. Membr. Sci., 2011, 382, 135–147.
35. V. Vatanpour, S. S. Madaeni, R. Moradian, S. Zinadini and B. Astinchap, J. Membr. Sci., 2011, 375, 284–294.
36. Z. F. Fan, Z. Wang, N. Sun, J. X. Wang and S. C. Wang, J. Membr. Sci., 2008, 320, 363–371.
37. S. Pourjafar, A. Rahimpour and M. Jahanshahi, J. Ind. Eng. Chem., 2012, 18, 1398–1405.
38. S. Y. Kwak, S. H. Kim and S. S. Kim, Environ. Sci. Technol., 2001, 35, 2388–2394.
39. Y. Yang, H. Zhang and P. Wang, J. Membr. Sci., 2007, 288, 231–238.
40. B. V. Bruggen, M. Mänttäri and M. Nyström, Sep. Purif. Technol., 2008, 63, 251–263.
41. E. M. Vrijenhoek, S. Hong and M. Elimelech, J. Membr. Sci., 2001, 188, 115.
42. J. Jang, J. Bae and K. Lee, Polymer, 2005, 46, 3677.
43. J. María Arsuaga, A. Sotto, G. del Rosario, A. Martínez, S. Molina, S. B. Teli and J. de Abajo, J. Membr. Sci., 2013, 428, 131.
44. H. Susanto and M. Ulbricht, J. Membr. Sci., 2009, 327, 125–135.
45. J. Kim and B. Van der Bruggen, Environ. Pollut., 2010, 158, 2335–2349.
46. N. Hilal, O. O. Ogunbiyi, N. J. Miles and R. Nigmatullin, Sep. Sci. Technol., 2005, 1957, 1957–2005.
47. W. Zhang, Y. Zhu, X. Liu, D. Wang, J. Li, L. Jiang and J. Jin, Angew. Chem., Int. Ed., 2014, 53, 856–860.

---