Optimized preparation conditions of TiO₂ deposited on SiO₂ solid superacid nanotubes as filler materials

Abstract

The preparation conditions have an important impact on the structure and properties of TiO₂ deposited on SiO₂ solid superacid nanotubes (TSANs). In this paper, the preparation conditions of TSANs such as the molar ratio of Ti/Si, the concentration of the aqueous solution of sulphuric acid, the dipping time in the aqueous solution of sulphuric acid and the calcination temperature, etc., were investigated and optimized and the optimum preparation conditions were finally determined. TSANs were employed as novel filler materials to prepare TSANs/PSF composite membranes with micro reaction locations (MRLs) inside the channels and on the surface of the composite membranes. TSANs and TSANs/PSF composite membranes were characterized and studied by SEM, TEM, EDX, FT-IR, BET and the contact angle of water, etc. The results indicate that the thickness of the tube walls is 50 nm, the ratio of the length to the diameter of TSANs is approximately 23, and SO₄²⁻/TiO₂–SiO₂ solid superacid nanotubes were formed, as shown by the stretching vibration peak of S=O double bonds at 1300 cm⁻¹ in the FT-IR spectra. The contact angle decreased from 76.5° to 36.5° and the oil retention ratio of TSANs/PSF composite membranes reached 94.2%, which confirms the attractive hydrophilic and antifouling properties of TSANs/PSF composite membranes. Therefore, TSANs are desirable as novel filler materials for PSF membranes to improve their integrated properties.

---

1. Introduction

The strategy of doping inorganic oxide nanoparticles into polymer membranes to prepare organic–inorganic composite membranes is fascinating. Polysulfone (PSF) membranes have been extensively employed as microfiltration membranes, ultrafiltration membranes, gas separation membranes, and pervaporation membranes, because of their outstanding properties such as excellent mechanical strength, good chemical resistance and high thermal stability.^1–3 However, owing to the hydrophobic characteristics of PSF materials, conventional PSF membranes easily suffer from severe membrane fouling, which impairs membrane performance and shortens membrane life.^4–6 In order to overcome these drawbacks, various inorganic oxide nanoparticles such as TiO₂, SiO₂, and Al₂O₃ have been widely employed as filler materials for the synthesis of composite membranes to enhance their performance.^7–9 However, further enhancement of the integrated properties of polymer membranes has been limited, because these inorganic nanoparticles could neither be fully compatible with polymers along the direction of the polymer chains nor provide excellent anti-compaction capabilities to resist external impact forces, because they are not rod-shaped or tubal materials with a high ratio of length to diameter.

In recent years, with the discovery of carbon nanotubes (CNTs),¹⁰ CNTs have motivated many researchers to investigate their potential applications in advanced materials. CNTs are attractive membrane fillers with extraordinary mass transport channels, which have been studied by various research groups. Several studies have shown the successful application of CNTs in a polymer matrix.^11–13 Similarly, silica nanotubes (SNTs) have also excited special interest as filler materials because they have many advantages such as hollow structures,¹⁴ large specific surface areas,¹⁵ a high ratio of length to diameter,¹⁶ biocompatibility¹⁷ and high anti-compaction strength.¹⁸ Therefore, when SNTs are employed as filler materials in the synthesis of SNTs/PSF composite membranes, the composite membranes have better compatibility and good anti-compaction performance.

In addition, membrane fouling is a major obstacle to the widespread application of membrane technology. Via the analysis of membrane fouling mechanisms,^19,20 pollutants are mainly divided into three categories, namely, inorganic pollutants (metal oxides, etc.), microbes and organic pollutants (hydrocarbons and soluble oils, etc.).^21,22 However, the above methods simply dope nanomaterials with small sizes into polymer membranes to enhance their antifouling properties by physical interactions at the interface between aqueous solutions and the composite membranes, but without any chemical reactions. Therefore, in order to facilitate chemical reactions, it is important to design and research types of novel functional nanomaterials with solid superacid properties, which can form MRLs inside the channels and on the surface of membranes. Sulfated TiO₂ is well known as a solid superacid with 10 000 times the Hammett acidity of 100% H₂SO₄ (ref. 23) and can decompose inorganic pollutants such as metal oxides or inhibit their formation inside the channels and on the surface of membranes, extending the membrane lifespan.²⁴ Moreover, owing to the strong acidity of sulfated TiO₂ solid superacid and the inductive effect of S=O bonds, a large number of hydroxyl groups are produced on the surface of sulfated TiO₂ solid superacid, which can improve the hydrophilicity of polymers. Therefore, when sulfated inorganic oxide nanoparticles are used to fill polymer membranes, MRLs are formed inside the channels and on the surface of the membranes.

Based on the above-mentioned research and analysis, in order to investigate novel filler materials for composite membranes, TSANs can be fabricated via combining the advantages of SNTs and sulfated TiO₂ solid superacid. TSANs have enough hydroxyl active sites along their length direction to be compatible with PSF polymers along the direction of the PSF chains. Namely, O–H groups of TSANs interact with S=O bonds of PSF through S=O⋯H–O hydrogen bonds.²⁵ Therefore, when TSANs are doped into PSF to prepare a TSANs/PSF composite membrane, some of the hydroxyl active sites that exist on the surface and in the channels of the composite membrane can interact with water molecules through hydrogen bonds,²⁶ which can effectively inhibit organic pollutants (hydrocarbons and soluble oils, etc.) and microbes from passing through the composite membrane and enhance the hydrophilic properties of the membrane; alternatively, MRLs that are formed in the channels and on the surface of the polymer can decompose or inhibit the formation of inorganic pollutants, which will further enhance the integrated properties of the polymer membranes, such as hydrophilicity, antifouling and anti-compaction performance. Therefore, it can be concluded from the above analysis that the acidity of TSANs is important for forming MRLs in membranes. Moreover, the preparation conditions have an important impact on the structure and properties of TSANs such as their acidity. Therefore, the effect of the preparation conditions of TSANs on their acidity should be thoroughly investigated.

In this paper, the preparation conditions of TSANs such as the molar ratio of Ti/Si, the concentration of the aqueous solution of sulphuric acid, the dipping time in the aqueous solution of sulphuric acid and the calcination temperature, etc., have been investigated and optimized in detail, and the optimum preparation conditions are determined by the Hammett acidity (H₀) in the test method. TSANs and TSANs/PSF composite membranes have been studied by SEM, TEM, EDX, FT-IR, BET and the contact angle of water, etc. Moreover, the effect of doping with TSANs on the properties of polysulfone membranes has also been evaluated and observed.

2. Experimental

2.1. Materials and reagents

Concentrated sulfuric acid (H₂SO₄, AR grade, 98%) was purchased from Beijing Chemical Factory. Aqueous ammonia (NH₃, AR grade, 25%) and dehydrated ethanol were obtained from Tianjin Guangfu Fine Chemicals Co., Ltd. Tetraethyl orthosilicate (TEOS, AR grade, mass ratio of silicon dioxide of 28%) and tartaric acid (C₄H₄O₆, AR grade, 99.5%) were provided by Tianjin Jiangtian Chemical Co., Ltd and Tianjin No. 1 Chemical Reagent Factory, respectively. Tetra-n-butyl titanate (TBT, AR grade, 98.0%) was bought from Tianjin KemiO Chemical Reagent Co., Ltd. Polysulfone (PSF) was purchased from Dalian Polysulfone Co., Ltd, and its MW and polydispersity were 84 400 Da and 1.37, respectively. N,N-Dimethylacetamide (DMAC, AR grade, 94.6%) was bought from Tianjin KemiO Chemical Reagent Co., Ltd and polyethylene glycol with an average MW of 400 Da (PEG400, AR grade, 95%) was supplied by Tianjin Bodi Chemical Co., Ltd.

2.2. Preparation of TSANs

TSANs with a length of 8 μm and an inner diameter of around 300 nm were prepared in our laboratory. The preparation approach is shown schematically in Fig. 1.

Fig. 1 Schematic diagram of preparation of TSANs.

2.2.1. Preparation of SNTs. SNTs were prepared by a template-assisted route, which was similar to that in the literature.²⁷ The optimum preparation conditions of SNTs were determined to be as follows: the template is tartaric acid, the stirring rate is 300 rpm, the pre-reaction time is 15 min and the calcination temperature is 550 °C. The detailed procedures are as follows: typically, TEOS was added to absolute ethanol containing tartaric acid and deionized water to prepare a mixed solution, and the mixture was allowed to stand for a while. Subsequently, 28% NH₄OH was added dropwise into two reactors under stirring, respectively, and kept under stirring for 10 min, and then the solution was left to stand for 2 h. The white precipitate that was formed was washed with a large amount of water to remove colloidal aggregates, centrifuged and dried at 60 °C. Finally, the resulting particles were sintered at 550 °C to obtain SNTs.

All the SNTs samples employed in the subsequent investigation were prepared under the above preparation conditions.

2.2.2. Preparation of TSNs. SNTs were uniformly dispersed into anhydrous ethanol under stirring and intermittent ultrasound at 50 °C. Then, a moderate amount of TBT was added dropwise and slowly to the above solution under stirring to be dispersed uniformly. Subsequently, two drops of distilled water were added to the mixed solution and stirring was continued for 10 min. Then, the reaction was completed after a white precipitate was fully formed by allowing the solution to stand for 20 min. The white precipitate was repeatedly washed three times with distilled water, followed by being centrifuged to obtain TSNs.

2.2.3. Sulfation of TSNs. Finally, TSNs were immersed in an aqueous solution of sulphuric acid of an appropriate concentration for an appropriate time to be sulfated. TSANs were obtained after being calcined in a furnace for 4 h at a suitable calcination temperature.

2.3. Preparation of TSANs/PSF composite membranes

TSANs/PSF composite membranes were prepared by phase inversion methods.²⁸ Firstly, TSANs were added to DMAC in a 500 mL flask under stirring at 60 °C. Then, a moderate amount of PSF (the mass ratio of TSAN : PSF was 10 : 100 (g : g)) was added and dissolved sufficiently under stirring and intermittent ultrasound. Two hours later, PEG 400 was added to the mixture as a porogen to increase the yield of pores in the gelation process. Vigorous stirring was still needed to obtain a homogeneous solution. Then, the solution was kept for 24 h at 20 °C. The gel was cast onto a horizontal glass plate to form composite membranes (thickness of ca. 0.2 mm) using a glass blade. After being pre-evaporated for 10 s in air (25 °C and 60% relative humidity), the composite membranes were immersed in a water bath at 50 °C. TSANs/PSF composite membranes were obtained after gel films fell off the glass plate. Finally, the membranes were soaked in distilled water containing 1 wt% formaldehyde to avoid bacterial growth.

Pure PSF membranes, SiO₂/PSF and phosphorylated Zr-doped hybrid silica (SZP)/PSF composite membranes²⁸ were prepared by using the same procedures as mentioned above, in which TSANs were replaced with nothing, SiO₂ (the mass ratio of SiO₂ : PSF was 10 : 100 (g : g)) and SZP (the mass ratio of SZP : PSF was 10 : 100 (g : g)), respectively.

2.4. Characterization

The FT-IR spectra of SNTs, TSNs and TSANs samples were recorded with an Avatar 370 spectrometer (Thermo Nicolet Corporation, USA), using the potassium bromide pellet technique. The wavelength coverage was 4000–370 cm⁻¹ and the resolution was 0.1 cm⁻¹. TEM images of TSANs were recorded with a JEM-2100F II transmission electron microscope (JEOL Corporation, Japan) at an operating voltage of 100 kV. TEM samples were prepared by briefly ultrasonicating powders in absolute ethanol, followed by placing a drop of a suspension onto a carbon-coated copper grid. The grids were desiccated before measurement. Cross-section samples were obtained by freeze-fracturing in liquid nitrogen to obtain a clean cross-section and then sputtering with gold. The cross-sections of SNTs and TSANs/PSF composite membranes were observed under a Hitachi S4800 scanning electron microscope using an accelerating voltage of 5 kV. The specific surface areas and pore size distributions of TSANs were determined by the BET method using a Micromeritics Instrument Corp. Tristar 3000 specific surface area analyzer.

2.5. Measurement of decomposition rate of suspended solids

In order to prove that the TSANs/PSF composite membrane can decompose inorganic pollutants such as metal oxides, we investigated the decomposition rate of suspended solids in oily sewage using the composite membrane. The specific operational method and principle were as follows: a piece of microporous membrane was weighed accurately after vacuum drying to a constant weight to record its weight and was then put into a Buchner funnel. A certain volume of the fluid solution being tested was measured. A vacuum pump was kept open during the process of filtration, with a plastic turkey baster slowly dripping the fluid solution being tested onto the microporous membrane. Finally, the microporous membrane used in the process of filtration was dried to a constant weight to record its weight. The concentration of suspended solids in the fluid solution being tested was calculated as follows:where SS is the concentration of suspended solids in the fluid solution being tested (mg L⁻¹), A₁ is the weight of the fluid solution being tested before filtration (g), A₂ is the weight of the fluid solution being tested after filtration (g), and V is the volume of the fluid solution being tested (L). Therefore, the decomposition rate of suspended solids using the composite membrane could be calculated by the following formula:where D_SS is the decomposition rate of suspended solids in oily sewage (%), SS₀ is the initial concentration of suspended solids in oily sewage (mg L⁻¹), SS₁ is the remaining concentration of suspended solids in oily sewage (mg L⁻¹) after being adsorbed by the PSF membrane, and SS₂ is the remaining concentration of suspended solids in oily sewage (mg L⁻¹) after being adsorbed and decomposed by the TSANs/PSF composite membrane.

2.6. Membrane separation performance studies

2.6.1. Preparation of wastewater containing oil. Model oily wastewater (oil-in-water emulsion) was created from machine oil and distilled water with vigorous stirring at a speed of 300 rpm for 1 h until a homogeneous solution was obtained. The stability of the emulsion was observed visually over a period of 24 h and the mixture remained cloudy and turbid, which indicated that the oil was in an emulsified and soluble condition. After being treated by sand leaching and fiber filtration, oil in the model sewage was adsorbed and separated, which made the model oily wastewater become emulsions with different oil concentrations. The oil concentration in the permeate was determined by a UV spectrophotometer (SP-752 UV-visible spectrophotometer, Shanghai Spectrum Corporation) at 225 nm. The absorbance of the fluid being tested could be converted into the oil concentration.

2.6.2. Ultrafiltration experiments. Ultrafiltration experiments on oily wastewater were carried out with PSF membranes and the SiO₂/PSF composite membrane, SZP/PSF composite membrane and TSANs/PSF composite membrane, respectively. The membrane evaluation device was made in our laboratory. The operating conditions of the separation experiments were in an intermittent mode.

The oil retention of the membranes was measured after the above experiments, and the retention R was calculated by the following equation:

where R is the retention (%), C₁ is the oil concentration in the feed solution (mg L⁻¹) and C₂ is the oil concentration in the permeate (mg L⁻¹).

3. Results and discussion

3.1. Optimum preparation conditions of TSANs

SNTs and the silylation of SNTs during the preparation of TSANs had been studied in our previous investigation.²⁹ Therefore, in this paper the effect of the preparation conditions of TSANs on the acidity of TSANs was researched, such as the molar ratio of Ti/Si, different concentrations of the aqueous solution of sulphuric acid, the dipping time in the aqueous solution of sulphuric acid and the calcination temperature, etc. Based on the above-mentioned analysis, the reason for this is that the acidity of TSANs is of great significance for the formation of MRLs in membranes.

3.1.1. Effect of the molar ratio of Ti/Si on the acidity of TSANs. In order to investigate the effect of the molar ratio of Ti/Si on the acidity of TSANs, different molar ratios of Ti/Si (1 : 1, 1 : 2, 1 : 3, 1 : 4, and 1 : 5 (mol : mol)) were studied and observed. As shown in Fig. 2, when the molar ratio of Ti/Si is 1 : 3 the acidity of TSANs reaches a maximum value (−7.584), which can promote the formation of MRLs in membranes. This is because the formation of MRLs is related to the acidity of TSANs.³⁰ Therefore, the optimum molar ratio of Ti/Si in the process of the preparation of TSANs is 1 : 3.

Fig. 2 Effect of the molar ratio of Ti/Si on the acidity of TSANs.

3.1.2. Effect of concentration of aqueous solution of H₂SO₄ on the acidity of TSANs. In order to investigate the effect of the concentration of the aqueous solution of H₂SO₄ on the acidity of TSANs, different concentrations of an aqueous solution of H₂SO₄ (0.5 mol L⁻¹, 1.0 mol L⁻¹, 1.5 mol L⁻¹, 2.0 mol L⁻¹, and 2.5 mol L⁻¹) were studied when the molar ratio of Ti/Si was 1 : 3. As shown in Fig. 3, when the concentration of the aqueous solution of H₂SO₄ is 1.0 mol L⁻¹ the acidity of TSANs reaches a maximum value (−8.002). The reason for this is that when the concentration of the aqueous solution of H₂SO₄ is less than 1.0 mol L⁻¹, the acidity of TSANs is lower (−7.584) owing to the presence of fewer SO₄²⁻ groups; alternatively, when the concentration of the aqueous solution of H₂SO₄ is greater than 1.0 mol L⁻¹, the acidity of TSANs also becomes low, which is because many SO₄²⁻ groups may clog the pores of the metal oxide and reduce the specific surface area, which reduces the acidity of TSANs. Therefore, the optimum concentration of the aqueous solution of H₂SO₄ is 1.0 mol L⁻¹.

Fig. 3 Effect of concentration of aqueous solution of H₂SO₄ on the acidity of TSANs.

3.1.3. Effect of dipping time in aqueous solution of H₂SO₄ on the acidity of TSANs. The effect of the dipping time in the aqueous solution of H₂SO₄ on the acidity of TSANs was studied by changing the dipping time of TSANs that were prepared based on the optimized preparation conditions (the molar ratio of Ti/Si was 1 : 3 (mol : mol) and the concentration of the aqueous solution of H₂SO₄ was 1.0 mol L⁻¹). As shown in Fig. 4, comparing the different dipping times in the sulfation process, it can be observed that the acidity of TSANs increases gradually with an increase in the dipping time. However, when the dipping time exceeds 4 h, the acidity of TSANs reaches a steady state (−8.002). Therefore, the optimum dipping time for TSANs in an aqueous solution of H₂SO₄ is 4 h.

Fig. 4 Effect of dipping time in aqueous solution of H₂SO₄ on the acidity of TSANs.

3.1.4. Effect of calcination temperature on the acidity of TSANs. The effect of the calcination temperature after sulfation on the acidity of TSANs was studied by changing the calcination temperature of TSANs that were prepared based on the optimized preparation conditions (the molar ratio of Ti/Si was 1 : 3 (mol : mol), the concentration of the aqueous solution of H₂SO₄ was 1.0 mol L⁻¹ and the dipping time in the aqueous solution of sulphuric acid was 4 h). Fig. 5 shows the acidity of TSANs calcined at different temperatures for 4 h. As depicted in Fig. 5, when the calcination temperature is 600 °C the acidity of TSANs reaches a maximum value (−7.687). Amorphous TiO₂ can be converted into the tetragonal form (anatase type) with catalytic activities at 600 °C.³¹ Moreover, the tetragonal form of TiO₂ can combine with SO₄²⁻ to form the solid superacid. Therefore, a lower calcination temperature limits the formation of superacid centers. In contrast, a higher calcination temperature causes the loss and decomposition of SO₄²⁻,³² which therefore cannot form the solid superacid. Therefore, the optimum calcination temperature for TSANs is 600 °C.

Fig. 5 Effect of calcination temperature on the acidity of TSANs.

Based on the above discussion and analysis, the optimum preparation conditions of TSANs are confirmed below: the molar ratio of Ti/Si is 1 : 3 (mol : mol), the concentration of the aqueous solution of H₂SO₄ is 1.0 mol L⁻¹, the dipping time in the aqueous solution of H₂SO₄ is 4 h and the calcination temperature is 600 °C. TSANs prepared under the optimum preparation conditions exhibited high acidity in a Hammett acidity (H₀) test, which confirms that TSANs/PSF composite membranes formed by doping with TSANs contained MRLs inside the channels and on the surface of the membranes. The formation of MRLs further enhanced the integrated properties of PSF membranes such as hydrophilicity, antifouling and anti-compaction performance.

3.2. SEM analysis of SNTs

Fig. 6 shows SEM images of the cross-sections of SNTs synthesized under the above optimum preparation conditions. It can be seen from Fig. 6 that SNTs have a tubular structure, their walls are extremely smooth, and the nanotubes are translucent. In addition, the physical properties of SNTs are as follows: the inner diameter of SNTs is about 300–400 nm and the thickness of the tube walls is about 50 nm.

Fig. 6 SEM images of SNTs [magnification: (a) ×6000; (b) ×35 000].

3.3. TEM analysis of TSANs

Fig. 7 displays TEM images of TSANs synthesized under the optimum preparation conditions. As can be seen from Fig. 7, TSANs have tubular structural characteristics, the length of TSANs is approximately 8 μm, the inner diameter of TSANs is 350 nm, the thickness of the tube walls is 50 nm and the ratio of the length to the diameter of TSANs is approximately 23. Moreover, it can also be observed from Fig. 7 that the tube walls of TSANs are not as smooth as those of SNTs. The reason for this is that TiO₂ nanoparticles have been deposited on SiO₂ nanotubes.

Fig. 7 TEM images of TSANs [magnification: (a) ×1500; (b) ×5000].

3.4. EDX analysis of TSANs

Fig. 8 shows the EDX spectrum of TSANs prepared under the optimum preparation conditions. It can be seen from the EDX spectrum of TSANs that the presence of Si, Ti, O and S elements can be demonstrated. The presence of C and Cu elements can also be confirmed. This is because TSAN samples were prepared by briefly ultrasonicating powders in absolute ethanol, followed by placing a drop of a suspension onto a carbon-coated copper grid. The relative contents of the elements in TSANs are shown in Table 1. It can be shown from Table 1 that the molar ratio of Ti/Si is about 1 : 3 (mol : mol), which conforms to the optimum preparation conditions.

Fig. 8 EDX spectrum of TSANs.

Table 1 Relative contents of elements in TSANs

Element	Weight percentage (%)	Atomic percentage (%)
O K	36.17	44.25
Si K	20.25	14.16
C K	18.99	30.97
Ti K	10.85	4.42
Cu K	7.23	2.21
S K	6.51	3.99
Total	100.00	100

---

3.5. FT-IR analysis of TSANs

The infrared spectra of SNTs, TSNs and TSANs are shown in Fig. 9. It can be observed from the infrared spectra of the three kinds of particles that the absorption peaks that appear at 1100 cm⁻¹, 830 cm⁻¹ and 460 cm⁻¹ are the asymmetric stretching vibration peak, symmetric stretching vibration peak and bending vibration peak of Si–O–Si, respectively.^33–35 They indicate that the main component of these materials is silica. The peaks at both 3400 cm⁻¹ and 1640 cm⁻¹ (stretching vibration peak and bending vibration peak of O–H) in the second and third spectra are obviously larger than those in the first spectrum.^25,36 This indicates that there are abundant hydroxyl radicals on the surfaces of TSNs and TSANs, which means that the hydrophilicity of SNTs increased after the deposition of sulfated TiO₂. In the infrared spectrum of TSNs, the stretching vibration peak of Si–O–Ti appears at 950 cm⁻¹,³⁷ which confirms that TiO₂ has been deposited on SNTs successfully. As can be seen from (Fig. 9b), the absorption peak that appears at 1300 cm⁻¹ in the third spectrum corresponds to the stretching vibration of S=O bonds and the absorption peak near 1130 cm⁻¹ corresponds to the characteristic vibration of the Ti–O–S linkage, which indicates that a chemical linkage (Ti–O–S linkage) has been formed between SO₄²⁻ and TiO₂.^38,39 Therefore, the FT-IR results clearly indicate that sulphate has been grafted onto TiO₂ successfully and TSANs have been formed.

Fig. 9 FT-IR spectra of SNTs, TSNs and TSANs [(a) wavenumbers: 400–4000 cm⁻¹; (b) wavenumbers: 400–1400 cm⁻¹].

3.6. BET analysis of TSANs

The specific surface area and pore size distribution of TSANs synthesized under the above optimum preparation conditions were determined by the Brunauer–Emmett–Teller (BET) and Barrett–Joyner–Halenda (BJH) methods, respectively. The specific surface area of TSANs is 480.2 m² g⁻¹. Fig. 10 shows the nitrogen adsorption–desorption isotherm and pore size distribution of TSANs. It can be observed from Fig. 10 that TSANs display type IV adsorption characteristics and a hysteresis loop of type H1 appears owing to adsorption of nitrogen in the tubular pores of TSANs. As can be seen from Fig. 10, the nitrogen adsorption isotherm has two peaks, which indicates that TSANs have two types of pore size. Fig. 10 indicates that the pore size of TSANs exhibits a bimodal distribution; the pores with an average pore size of 7.4 nm correspond to mesopores in the tube walls of TSANs, whereas the pores with an average pore size of 350 nm correspond to the tubular inner pores of TSANs.

Fig. 10 Nitrogen adsorption–desorption isotherm and BJH pore size distribution of TSANs.

3.7. SEM studies of composite membrane

Fig. 11 shows SEM images of the cross-section of TSANs/PSF composite membranes prepared by doping TSANs into a PSF membrane. In (Fig. 11a), the TSANs/PSF composite membrane has an asymmetric structure with a compact skin layer and a porous support substrate. The skin layer can retain macromolecular substances owing to its compact form and let rinsing agents pass through, whereas the substrate, which has numerous finger-like cavities through the entire thickness of the membrane, functions as an anti-compaction supporting layer. It can be observed from (Fig. 11b) that TSANs are dispersed inside the PSF membrane uniformly and can form MRLs inside the channels and on the surface of the membrane. In addition, it can also be seen that TSANs and PSF polymer chains combine well, which increases the anti-compaction ability and hydrophilicity of PSF polymer membranes, enhancing the integrated properties of TSANs/PSF membranes.

Fig. 11 SEM images of the cross-section of TSANs/PSF composite membranes [magnification: (a) ×600; (b) ×2500].

3.8. Analysis of the hydrophilic properties of TSANs/PSF composite membranes

The relative hydrophilicity of a membrane surface can easily be determined by measurement of the water contact angle,⁴⁰ namely, a membrane with a lower water contact angle exhibits better hydrophilic properties. The contact angles between water droplets and PSF composite membranes doped with different inorganic particles were measured. It can be observed from Fig. 12 and 13 that the contact angle for the TSANs/PSF composite membrane is the lowest at 36.5°, which indicates that the hydrophilicity of the TSANs/PSF composite membrane is greater than that of the other three composite membranes. This is because sulfated TiO₂ solid superacid, with its high acidity, can form hydrophilic coatings on the surface of PSF membranes; in addition, in the length direction of TSANs, more hydroxyl radicals are supplied by S–OH, Ti–OH and many Lewis acid sites, which ensures the formation of many hydrogen bonds between the aqueous solution and TSANs. Therefore, TSANs/PSF composite membranes display stronger hydrophilic properties.

Fig. 12 Effect of different particles on membrane hydrophilicity.

Fig. 13 Contact angles of different membranes.

3.9. Studies of anti-compaction performance

Section S3 in the ESI† shows the method of measurement of membrane anti-compaction performance. Fig. 14 shows the effect of different doping materials on the membrane ultrafiltration water permeation flux of membranes (under a pressure of 0.10 MPa and at a temperature of 25 °C). In Fig. 14, it is observed that the membrane ultrafiltration water permeation flux of the TSANs/PSF composite membrane is greater than that of the PSF membrane, SiO₂/PSF composite membrane and SZP/PSF composite membrane. Moreover, the increment between membrane ultrafiltration water stable fluxes reaches a maximum of 424.67% for TSANs/PSF composite membranes, as the data indicate in Table 2, which indicates that the membrane ultrafiltration water stable flux of TSANs/PSF composite membranes is more than four times that of PSF membranes. This can be ascribed to the interaction between TSANs and PSF chains. On the one hand, SNTs with their tubular structure, large specific surface areas and high ratio of length to diameter could be fully compatible with polymers and the unique circular wall of nanometer-scale thickness of TSANs can function as a new energy-dissipating feature to resist the effects of external forces. On the other hand, the abundant hydroxyl radicals on the surface of TSANs can interact with PSF chains through hydrogen bonds. Thus, the anti-compaction properties of PSF membranes are significantly improved by doping with TSANs.

Fig. 14 Effect of different doping materials on the pure water permeation flux of membranes.

Table 2 Increments between pure water stable fluxes of membranes

Membrane	Pure water stable flux (L m^−2 h^−1)	Increment between pure water stable fluxes (%)
PSF membrane	77	0
SiO2/PSF composite membrane	161	109.09
SZP/PSF composite membrane	261	238.96
TSANs/PSF composite membrane	404	424.67

---

3.10. Effect of different doping materials on separation properties of membranes

In order to investigate the effect of different doping materials on the separation performance of membranes, ultrafiltration experiments on wastewater containing oil (oil concentration 71.46 mg L⁻¹, operating temperature 25 °C, operating pressure 0.1 MPa) were carried out with PSF membranes, SiO₂/PSF composite membranes, SZP/PSF composite membranes and TSANs/PSF composite membranes, respectively. Fig. 15 shows the effect of different doping materials on the permeate flux of membranes when treating wastewater containing oil. It can be observed from Fig. 15 that TSANs/PSF composite membranes exhibit a higher stable flux of 113 L m⁻² h⁻¹ than the other three membranes. Table 3 shows the oil retention rate of these four membranes. Among all the membranes, the TSANs/PSF composite membrane displays attractive oil retention properties and its oil retention rate reaches 94.2%. The above results indicate that the TSANs/PSF composite membrane has better separation performance. This is because the more strongly hydrophilic coatings inside the channels and on the surface of the membrane lead to the formation of a water layer on the membrane surface⁴¹ and more hydroxyl radicals can interact with water molecules through hydrogen bonds, which makes water molecules pass through the pore channels smoothly while most oil droplets and suspended solids are intercepted by the membrane. In addition, MRLs that are formed in the channels and on the surface of the TSANs/PSF composite membrane can decompose or inhibit the formation of inorganic pollutants. Moreover, there are many accessible active sites that are compatible with the PSF membrane along the direction of the polymer chains; in the diameter direction of TSANs there is a unique circular wall of nanometer-scale thickness, which can function as a new energy-dissipating feature to resist external impact forces. Thereby, TSANs/PSF composite membranes display better compatibility, tensile strength and anti-compaction performance. From the above research and analysis it can be concluded that TSANs with very high acidity, a tubular structure and a high ratio of length to diameter were first prepared and then doped into PSF to prepare TSANs/PSF composite membranes, making TSANs/PSF composite membranes have higher flux and attractive integrated properties, which are suitable in applications for treating wastewater.

Fig. 15 Effect of doping with different particles on membrane permeate flux.

Table 3 Effect of different particles on membrane oil retention

Membrane	Oil concentration in permeate (mg L^−1)	Oil retention rate (%)
PSF membrane	5.47	92.3
SiO2/PSF composite membrane	5.06	92.9
SZP/PSF composite membrane	4.61	93.5
TSANs/PSF composite membrane	4.15	94.2

---

3.11. Decomposition of inorganic pollutants by TSANs/PSF composite membrane

In order to prove that the TSANs/PSF composite membrane can decompose inorganic pollutants such as metal oxides, we investigated the decomposition rate of suspended solids in oily sewage by the composite membrane. The decomposition rate of suspended solids in oily sewage by the TSANs/PSF composite membrane was calculated by eqn (2) and is shown in Fig. 16. As indicated in Fig. 16, the decomposition rate of inorganic pollutants by the TSANs/PSF composite membrane reaches 15.53% when the decomposition experiment on inorganic pollutants has been carried out for 120 min. The reason for this is that TSANs/PSF composite membranes have many MRLs inside their channels and on their surface and a strong acid center of solid superacid is formed on the surface of TSANs, causing a low pH and hydrophilic coatings on the surface of PSF membranes. This makes it hard for inorganic pollutants to adhere to the surface of the membranes. Although some inorganic pollutants can be adsorbed on the surface or in the channels of the membranes, these pollutants would be decomposed by the strong acid center of TSANs. Therefore, in the process of treating oily sewage using TSANs/PSF composite membranes, inorganic pollutants can be partially decomposed, achieving the aim of improving the antifouling property of PSF membranes.

Fig. 16 Decomposition rate of inorganic pollutants by TSANs/PSF composite membrane.

3.12. Analysis of the mechanism of formation of MRLs inside the channels and on the surface of TSANs/PSF composite membranes

3.12.1. Mechanism of interaction between TSANs and PSF chains. Fig. 17 presents a scheme of the mechanism of the interaction between TSANs and PSF chains. As indicated in Fig. 17, TSANs have enough hydroxyl active sites along their length direction to be compatible with PSF polymers along the direction of the PSF chains. These hydroxyl groups can interact with PSF chains through hydrogen bonds. Namely, O–H groups of TSANs interact with S=O bonds of PSF through S=O⋯H–O hydrogen bonds. In addition, in the diameter direction of TSANs there is a unique circular wall of nanometer-scale thickness, which can function as a new energy-dissipating feature to resist external impact forces. This ensures that TSANs/PSF composite membranes have good anti-compaction properties.

Fig. 17 Mechanism of interaction between TSANs and PSF chains.

3.12.2. Scheme of formation of MRLs inside the channels and on the surface of TSANs/PSF composite membranes. Fig. 18 presents the scheme of formation of MRLs. The process of the degradation of inorganic pollutants by the TSANs/PSF composite membrane is shown in this figure. A strong acid center of solid superacid was formed on the surface of TSANs, causing a low pH and hydrophilic coatings on the surface of PSF membranes. This makes it hard for inorganic pollutants to adhere to the surface of the membranes. Although some inorganic pollutants can be adsorbed on the surface or in the channels of the membranes, these pollutants would be degraded by the strong acid center of TSANs or would not be formed around the strong acid center of TSANs. These results mean that TSANs/PSF composite membranes have good anti-inorganic fouling properties. Moreover, some of the hydroxyl active sites that exist on the surface and in the channels of the composite membrane can interact with water molecules through hydrogen bonds, which can effectively inhibit organic pollutants (hydrocarbons and soluble oils, etc.) and microbes from passing through the composite membrane and enhance the hydrophilic properties of the membrane.

Fig. 18 Scheme of formation of MRLs inside the channels and on the surface of TSANs/PSF composite membranes.

4. Conclusions

In this paper, the optimum preparation conditions of TSANs were investigated and determined, and then TSANs were doped into PSF to prepare a TSANs/PSF composite membrane. The optimum preparation conditions of TSANs are: the molar ratio of Ti/Si is 1 : 3 (mol : mol), the concentration of the aqueous solution of H₂SO₄ is 1.0 mol L⁻¹, the dipping time in the aqueous solution of H₂SO₄ is 4 h and the calcination temperature is 600 °C. The results of studies via SEM, TEM, EDX, FT-IR, BET and the contact angle of water indicate that TSANs with high acidity were successfully prepared and TSANs can be dispersed well in PSF membranes, making TSANs/PSF composite membranes not only have better compatibility but also contain many MRLs inside the channels and on the surface of the membrane. The formation of MRLs can further enhance the integrated properties of PSF membranes such as hydrophilicity, antifouling and anti-compaction performance. Therefore, TSANs are desirable as novel filler materials for polysulfone membranes.

Abbreviations

TSANs	TiO2 deposited on SiO2 solid superacid nanotubes
MRLs	Micro reaction locations
PSF	Polysulfone
CNTs	Carbon nanotubes
SNTs	Silica nanotubes
TSNs	TiO2 deposited on SiO2 nanotubes
SZP	Phosphorylated Zr-doped hybrid silica

Notation

H0	Hammett acidity
SS	Suspended solid in fluid solution being tested (mg L^−1)
A1	Weight of fluid solution being tested before filtration (g)
A2	Weight of fluid solution being tested after filtration (g)
V	Volume of fluid solution being tested (L)
DSS	Decomposition rate of suspended solids in oily sewage (%)
SS0	Initial concentration of suspended solids in oily sewage (mg L^−1)
SS1	Remaining concentration of suspended solids in oily sewage (mg L^−1)
R	Retention (%)
C1	Oil concentration in the feed solution (mg L^−1)
C2	Oil concentration in the permeate (mg L^−1)

Acknowledgements

This project is supported by National Natural Science Foundation of China (No. 21076143, 21676180), by the key technologies R & D program of Tianjin (15ZCZDSF00160), by the Basic Research of Tianjin Municipal Science and Technology Commission (13JCYBJC20100), by Tianjin Municipal Science and Technology Xinghai Program (No. KJXH2014-05), by State Key Laboratory of Chemical Engineering (No. SKL-ChE-15B03).

References

1. H. J. Song and C. K. Kim, J. Membr. Sci., 2013, 444, 318–326.
2. V. R. Pereira, A. M. Isloor, U. K. Bhat, A. F. Ismail, A. Obaid and H.-K. Fun, RSC Adv., 2015, 5, 53874–53885.
3. E. Eren, A. Sarihan, B. Eren, H. Gumus and F. O. Kocak, J. Membr. Sci., 2015, 475, 1–8.
4. H.-B. Dong, Y.-Y. Xu, Z. Yi and J.-L. Shi, Appl. Surf. Sci., 2009, 255, 8860–8866.
5. S. B. Teli, S. Molina, A. Sotto, E. G. Calvo and J. d. Abajob, Ind. Eng. Chem. Res., 2013, 52, 9470–9479.
6. G. Zhang, S. Lu, L. Zhang, Q. Meng, C. Shen and J. Zhang, J. Membr. Sci., 2013, 436, 163–173.
7. Y. Yang, H. Zhang, P. Wang, Q. Zheng and J. Li, J. Membr. Sci., 2007, 288, 231–238.
8. A. Ahmad, M. Majid and B. Ooi, Desalination, 2011, 268, 266–269.
9. Y. M. Mojtahedi, M. R. Mehrnia and M. Homayoonfal, Desalin. Water Treat., 2013, 51, 6736–6742.
10. S. Iijima, Nature, 1991, 354, 56–58.
11. Ihsanullah, A. M. Al Amer, T. Laoui, A. Abbas, N. Al-Aqeeli, F. Patel, M. Khraisheh, M. A. Atieh and N. Hilal, Mater. Des., 2016, 89, 549–558.
12. C.-M. Chang, H.-Y. Li, J.-Y. Lai and Y.-L. Liu, RSC Adv., 2013, 3, 12895–12904.
13. S. A. R. Hashmi, H. C. Prasad, R. Abishera, H. N. Bhargaw and A. Naik, Mater. Des., 2015, 67, 492–500.
14. H. J. Fan, U. Gösele and M. Zacharias, Small, 2007, 3, 1660–1671.
15. K.-S. Kim and S.-J. Park, Electrochim. Acta, 2012, 78, 147–153.
16. C. Gao, Z. Lu and Y. Yin, Langmuir, 2011, 27, 12201–12208.
17. N. Esman, A. Peled, R. Ben-Ishay, Y. Kapp-Barnea, I. Grigoriants and J.-P. Lellouche, J. Mater. Chem., 2012, 22, 2208–2214.
18. A. Nan, X. Bai, S. J. Son, S. B. Lee and H. Ghandehari, Nano Lett., 2008, 8, 2150–2154.
19. W. Gao, H. Liang, J. Ma, M. Han, Z.-l. Chen, Z.-s. Han and G.-b. Li, Desalination, 2011, 272, 1–8.
20. W. Zhang, W. Liang, G. Huang, J. Wei, L. Ding and M. Y. Jaffrin, RSC Adv., 2015, 5, 48484–48491.
21. C.-H. Yu, L.-C. Fang, S. K. Lateef, C.-H. Wu and C.-F. Lin, J. Hazard. Mater., 2010, 177, 1153–1158.
22. A. Sayari, S. Hamoudi and Y. Yang, Chem. Mater., 2005, 17, 212–216.
23. T. Jiang, Q. Zhao, M. Li and H. Yin, J. Hazard. Mater., 2008, 159, 204–209.
24. Z. Li, R. Wnetrzak, W. Kwapinski and J. J. Leahy, ACS Appl. Mater. Interfaces, 2012, 4, 4499–4505.
25. O. Lukin, A. Shivanyuk, V. V. Pirozhenko, I. F. Tsymbal and V. I. Kalchenko, J. Org. Chem., 1998, 63, 9510–9516.
26. L. Scatena, M. Brown and G. Richmond, Science, 2001, 292, 908–912.
27. F. Miyaji, S. A. Davis, J. P. H. Charmant and S. Mann, Chem. Mater., 1999, 11, 3021–3024.
28. Y. Zhang, Z. Jin, Y. Wang and P. Cui, J. Membr. Sci., 2010, 361, 113–119.
29. Y. Zhang, Y. Xu, Y. Lu, L. Zhao and L. Song, Nanotechnology, 2013, 24, 315701.
30. Y. Zhang and P. Liu, Chem. Eng. Sci., 2015, 135, 67–75.
31. H. Yang, R. Lu and L. Wang, Mater. Lett., 2003, 57, 1190–1196.
32. F. Zane, S. Melada, M. Signoretto and F. Pinna, Appl. Catal., A, 2006, 299, 137–144.
33. V. A. Zeitler and C. A. Brown, J. Phys. Chem., 1957, 61, 1174–1177.
34. K. Y. Jung and S. B. Park, Appl. Catal., B, 2000, 25, 249–256.
35. H. Xiong, S. Tang, H. Tang and P. Zou, Carbohydr. Polym., 2008, 71, 263–268.
36. L. Fang, X. Zu, Z. Li, S. Zhu, C. Liu, W. Zhou and L. Wang, J. Alloys Compd., 2008, 454, 261–267.
37. Y. Song, Y. Hasegawa, S. Yang and M. Sato, J. Mater. Sci., 1988, 23, 1911–1920.
38. L. Yan, W. Li, Z. Qi and S. Liu, J. Polym. Res., 2006, 13, 375–378.
39. D. S. Guo, Z. F. Ma, Q. Z. Jiang, H. H. Xu, Z. F. Ma and W. D. Ye, Catal. Lett., 2006, 107, 155–159.
40. F. Liu, B.-K. Zhu and Y.-Y. Xu, Appl. Surf. Sci., 2006, 253, 2096–2101.
41. B. Chakrabarty, A. Ghoshal and M. Purkait, Chem.–Eng. J., 2010, 165, 447–456.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra17110a

---