Abstract
In this study, a UV-assisted desalination system was prepared with a non-porous TiO2-doped Pebax 1657 membrane. The membranes were characterized and desalination tests were performed. The effects of UV light and TiO2 ratio on the performance of desalination were investigated. According to the results, TiO2 incorporation increased the membrane hydrophilicity, increased the membrane swelling values, enhanced the membrane flux and improved the salt rejection. Moreover, UV treatment has a positive effect on desalination performance. The best improvement was achieved in the results of 10 wt.% TiO2-doped membrane. It was found that the flux value of the UV-treated membrane having 10 wt.% TiO2 concentration was 8.2 kg/m2·h and the salt rejection value was 99.97%. It was found that the prepared membrane showed excellent desalination performance.
HIGHLIGHTS
Photocatalytic membrane reactor (PMR) was tested for desalination with/without UV.
In this study TiO2-doped Pebax 1657 membrane was used in PMR.
TiO2 nanotube incorporation enhanced the flux from 4.11 kg/m2·h to 8.2 kg/m2·h.
The conductivity of the pure water was improved from 206 μs to 21 μs by UV irradiation.
Graphical Abstract
INTRODUCTION
Photocatalytic oxidation of organic compounds for desalination and water treatment is one of the most interesting topics of recent years. The activity of this reaction depends on the process and the photocatalyst type. There are two main process configurations, as a process with free catalyst and as a process where the catalyst is immobilized on a support. The activity of the catalyst used in the free state is high, but after the water treatment, it is necessary to remove the catalyst to eliminate the poisoning effect of the catalyst. The use of photocatalytic membrane reactors (PMR) restricts the reduction of photocatalytic activity and ensures that the freshwater is not poisoned by the photocatalyst (Zheng et al. 2017; Argurio et al. 2018). Membrane usage prevents the passage of the catalyst into the purified water side. Additionally, the use of photocatalytic membrane reactors eliminates the use of further separation steps such as sedimentation–coagulation–flocculation, which will then be used to separate the photocatalyst. Furthermore, it reduces the total process volume and enables the reuse of the catalyst. The heat energy which is necessary for the reaction and separation can be provided by the UV in daylight. Therefore, the PMR system is defined as an energy-saving and cost-effective purification process (Mozia 2010; Molinari et al. 2017).
The photocatalytic membrane has been recently developed for a wide range of application (Zheng et al. 2017; Argurio et al. 2018; Ye et al. 2019). Biological treatment of organic content water, seawater desalination, and wastewater treatments are the best-known application areas of photocatalytic membranes. The mechanical properties and selective separation capability of membranes can be destroyed by the accumulation of organic pollutants in seawater. Therefore, photocatalytically eliminating organic pollutants may improve both the structural properties and separation performance of membranes (Yacou et al. 2015).
Membrane technology is an emerging science for advanced water treatment. Membranes are mostly used for water purification, wastewater treatment, and desalination purposes. The productivity and overall performances of membrane-based processes are directly related to the physicochemical properties of the membranes. Therefore, scientific studies have been focused on the development of high-performance membrane materials. Especially for water treatment, it is important to produce superior membranes having high water permeability, high rejection capacity, biofouling resistance, and process stability. For this purpose, new-generation composite membranes have been developed.
In this study, the aim is to develop a new material for UV-assisted desalination of seawater. For this purpose, titanium dioxide nanotube (TiO2)-doped polyether block amide (PEBA-Pebax 1657 (commercial name)) membrane has been produced. According to the authors' knowledge, TiO2–Pebax nanocomposite membrane is prepared here for the first time for the purpose of photocatalytic desalination. Different from the literature, the membrane is used as a non-porous form to remove water. Hence, the separation mechanism is based on water–membrane affinity. Therefore, any improvement by addition of TiO2 is much more important. TiO2 is the most widely used photocatalyst. TiO2 is a semiconducting material that destroys organic contaminants in water with the assistance of UV-light (Yacou et al. 2015; Solcova et al. 2016; Al Mayyahi 2018). However, because the energy band of TiO2 is large, it is not active in daylight. Since the UV content in daylight does not exceed 5%, a UV source is required for photocatalytic activation. In this study, the desalination system has been utilized by assisting UV light. The structural property of the composite membrane has been investigated by means of Fourier transform infrared spectroscopy and X-ray diffraction. The distribution of TiO2 particles within the membrane has been analysed using polarized electron microscopy (POM) and scanning electron microscopy (SEM). Swelling experiments have been performed to determine the water-uptake capacity of the TiO2 filled and unfilled Pebax 1657 membrane. The separation performance of membranes with/without TiO2 nanotubes was investigated to separate water seawater. Desalination tests have been conducted at room temperature with/without UV light. The effect of TiO2 concentration and UV light were investigated.
EXPERIMENT
Titanium dioxide nanotube preparation
One gram of commercial titanium dioxide powder (Acros, Thermo Fisher Scientific, USA) was mixed with 10 M sodium hydroxide solution. The mixture was subjected to an ultrasonic bath for 30 minutes. The mixture was left for hydrothermal treatment at 130 °C for 24 hours in a Teflon-lined autoclave. After filtration, the nanotube titanium dioxide was washed with hydrochloric acid and water. Nanotube titanium dioxide was dried at 110 °C for 16 hours and calcined at 500 °C for 2 hours.
Membrane preparation
Polyether block amide (Pebax 1657) was kindly supplied from Arkema, France, and 10 wt.% of Pebax-acetic acid solution was prepared and stirred until the solution became homogeneous. The determined amount of TiO2 nanotube particles was added to the Pebax solution according to the weight of the dry Pebax polymer. According to the Pebax content, TiO2 concentration in the membrane solvent was changed from 0 to 10 wt.%. Membranes were denoted according to the TiO2 content in the membrane. Pebax-0 represents the membrane without filler, Pebax-2.5T represents the membrane having 2.5 wt.% of TiO2 nanotube particles. Membrane solutions were cast onto a Teflon plate and allowed to dry at room temperature. Then, membranes were cured in a vacuum oven at 60 °C. The membrane was then immersed in a solution containing 2 vol.% of tolylene-2,4-diisocyanate-hexane solution to prevent the dissolving of the membrane in the water media.
Swelling experiment
Desalination
Characterization
The crystalline phases were characterized by means of X-ray diffraction (XRD) (Rigakku, Miniflex 2, Japan). The 2θ values were selected from 10° to 80° with a step size of 0.02 using Cu Kα radiation (λ = 0.15418 nm) at 45 kV/40 mA. Fourier transform infrared spectroscopy was used to determine the chemical structure of the membrane in the wavelength range from 650 cm−1 to 4,000 cm−1 (Perkin Elmer ATR). Surface hydrophilicity properties of the membranes were determined using contact angle measurements (Attension).
RESULTS
Characterization results of membranes
The prepared membranes were characterized in order to determine physical, chemical and morphological properties. In Figure 1, the XRD spectra of the prepared pure nanotube TiO2 and nanotube TiO2-doped membrane are given. The characteristic peaks of the anatase phase are obtained at 2θ 25.1°, 36.9°, 37.7°, 38.4°, 47.9°, 53.7° and 54.9° degrees. The characteristic peaks are determined at 12.3° and 21.4° for Pebax.
XRD patterns of Pebax–nanotube TiO2 membrane and pure nanotube TiO2.
In the literature, Pebax is defined as a semi-crystalline polymer which shows diffraction peaks at 2θ 14°, 17° and 26° (Sridhar et al. 2008). As seen from the figure, there are no crystalline phases obtained from nanotube-doped Pebax. This is due to the chemical interaction of Pebax with nanotube TiO2. Nanotubes affected the chemical bonds and an amorphous phase was formed.
The FTIR analysis of TiO2-nanoparticle-doped Pebax 1657 membrane is given in Figure 2. The bands at 1,107 and 1,733 cm−1 are attributed to the C─O─C and ─C═O stretching vibrations, respectively. Also, another two bands at 1,638 and 3,303 cm−1 are assigned to the presence of H─N─C═O and N─H groups, in the hard polyamide (PA) segment, respectively. In the FTIR spectrum of Pebax, it seems that the PA block of Pebax is significantly self-associated via hydrogen bonding.
Figure 3 shows FTIR spectra of commercial and nanotube TiO2 particles. The bands which are observed in the range of 3,500 to 3,200 cm−1 are related to the asymmetric and symmetric stretching vibrations of hydroxyl (-OH) groups. The Ti─OH stretching mode is observed at 1,633.58 cm−1 (Chougala et al. 2017) which is in agreement with the literature.
Figure 4 shows the distribution of TiO2 particles in the membrane. The first picture represents the Pebax 1657 membrane without filler, and the smooth, dense structure of the membrane can be clearly observed except for some impurity in the surface. The membrane has a non-porous structure. The POM images of the 2.5 wt.% TiO2-loaded membrane indicated that the distribution of the particles within the membrane is excellent. Homogeneous distribution of the additive into the membranes is very important in terms of providing the same physical and chemical properties in every region on the membrane. Homogeneous dispersion means uniform separation performance across the membrane. As is known, there are electrostatic interactions between particles, such as TiO2, which tend to connect with each other. In this study, the particles were separated from each other by the sonication process, and then the priming method was used to provide the homogeneous distribution (Nigiz 2018). Thus, agglomeration was prevented. Agglomeration within the membrane was rarely observed even when the concentration of TiO2 (wt.10%) was high. However, it is still possible to see accumulated TiO2 particles in a particular area. This is due to the orientation of the particles in the drying step of the membrane. During the casting–evaporation step, the membranes were first dried under room conditions. The TiO2 particles might be orientated at this stage.
Polarized electron microscopy images of the pristine and TiO2-loaded membranes.
In Figure 5, the cross-sectional views of the membranes are clearly seen. SEM images of 10 wt% TiO2-loaded membranes are given with different magnification. The uniform distribution of TiO2 is seen in Figure 5. The nanotube diameters vary from 10 to 23 nm and the nanotubes contribute to the properties of the membrane.
The contact angle between the membrane surface and water droplet is shown in Table 1. It is clear that the water contact angle decreases with increasing TiO2 ratio in the membrane. This means that the hydrophilicity of the membrane is enhanced with TiO2 addition. Owing to the hydrophilicity of TiO2 particles, the surface water affinity of the membrane is enhanced as also reported by Huang et al. (2017). According to the results obtained from the contact angle experiments, water flux is expected to increase as TiO2 ratio increases in the membrane matrix. There have also been many studies on the use of TiO2 in the membrane matrix to enhance the surface hydrophilicity and flux, as well (Du et al. 2017; Pan et al. 2019).
Contact angle of membranes with water droplets
. | Membrane code . | |||
---|---|---|---|---|
. | Pebax-0 . | Pebax-2.5T . | Pebax-5T . | Pebax-10T . |
Contact angle (°) | 68 | 57 | 49 | 38 |
. | Membrane code . | |||
---|---|---|---|---|
. | Pebax-0 . | Pebax-2.5T . | Pebax-5T . | Pebax-10T . |
Contact angle (°) | 68 | 57 | 49 | 38 |
Figure 6 shows the water affinity of the pristine and TiO2-nanotube-loaded membranes in terms of the swelling experiment. The increasing TiO2 nanotube ratio remarkably enhanced the water swelling ratio. The swelling values of the membrane increased from 72% to 129% when the TiO2 ratio was increased from 0 wt.% to 10 wt.%. This is due to the hydrophilic character of TiO2 nanotube particles, which was also confirmed by the results of the water contact angle experiments. It has been reported that TiO2 nanotube particles, and especially nano-sized TiO2 nanotubes, have an affinity to water (Bolis et al. 2012).
Desalination results in the system with/without UV lamps are given in Figure 7 as a function of flux and rejection. In order to investigate the effect of TiO2 nanotube addition, a desalination experiment on the pristine Pebax 1657 (Pebax-0) was also performed. As can be seen from the figure, increasing TiO2 ratio improved flux values significantly. In the pristine membrane, a 4.11 kg/m2·h flux value was obtained while 8.2 kg/m2·h flux was obtained with 10 wt.% TiO2-nanotube-loaded membrane when the separation occurred in the UV-assisted system. As was confirmed by the results of surface hydrophilicity and swelling degree, the water permeability was increased due to the hydrophilic character of the TiO2 nanotube particles. At the same time, it was also found that UV usage has little effect on the flux results. In the pristine membrane, 3.75 kg/m2·h flux was obtained in the system without using UV light and this value increased to 4.11 kg/m2·h when UV light was used. The increase in flux is related to the UV effect on organic compounds in seawater. By using UV light, TiO2 was activated and the degradation of organics contributed to preventing surface cake formation. Therefore, flux decrement due to the boundary layer problem could be prevented. This effect could become more effective in long-period separation experiments. Another reason could be related to the antimicrobial effect of TiO2 on bacteria, microbial and other substances in seawater. These substances also cause cake formation on the membrane surface. TiO2 is known as an anti-microbial chemical that prevents biological contamination on membrane surfaces.
The effect of TiO2 ratio on the salt rejection of the desalination process performed with/without UV light is seen in Figure 8. All Pebax-based membranes show a high salt rejection value above 99.78%. TiO2 incorporation into the membrane appears to increase salt rejection. The possible reason for these minor differences is that the TiO2 particles extend the tortuous pathway, increase the water affinity of the membrane, and increase the ion rejection by electrostatic interaction. The higher salt rejection values are seen in the system where the UV lamp is used. This should be attributed to the retention of ions by TiO2, which generates radicals by UV activation. The highest salt rejection value was obtained as 99.97% with 10 wt.% TiO2-loaded membrane. The conductivity of the permeated water was decreased from 206 μs to 21 μs by applying the UV light.
CONCLUSIONS
In this study, TiO2-nanotube-incorporated Pebax 1657 non-porous membranes were produced. The produced membranes were characterized and the desalination performance of the pristine and nanocomposite membranes was carried out at room temperature. General experimental results can be summarized as follows:
The compatibility of Pebax 1657 polymer and TiO2 nanotube particles was excellent and the TiO2 nanotube particles were homogeneously dispersed in the matrix.
TiO2 nanotube incorporation was found to increase both the surface and structural hydrophilicity of the membrane. The swelling values of the membrane increased from 72% to 129% when the TiO2 ratio was increased from 0 wt.% to 10 wt.%.
TiO2 nanotube addition enhanced the flux from 4.11 kg/m2·h to 8.2 kg/m2·h and the salt rejection from 99.90% to 99.97%.
Both the flux and salt rejection values were improved with assistance of UV light.
ACKNOWLEDGEMENTS
The authors would like to thank colleagues from the Polymer and Rubber Technology: Characterization Laboratory, Kocaeli University, for their support on use of characterization equipment. This work was supported by Kocaeli University Scientific Research Projects Coordination Unit, Project Number FBA-2019/1552.