Polyethersulfone Blended with Titanium Dioxide Nanoribbons/Multi-Wall Carbon Nanotubes for Strontium Removal from Water

Nanofiltration methods were used and evaluated for strontium removal from wastewater. The phase inversion method was used to create a variety of polyethersulfone (PES)/TiO2 nanoribbons (TNRs)–multi-walled carbon nanotubes (MWCNTs) membranes with varied ratios of TNR-MWCNT nanocomposite. The hydrothermal technique was applied to synthesize the nanocomposite (TNRs-MWCNTs), which was then followed by chemical vapor deposition (CVD). The synthesized membranes were characterized by scanning electron microscopy (SEM), transmission electron microscopy, and FTIR. TNR macrovoids are employed as a support for the MWCNT growth catalyst, resulting in a TNR-MWCNT network composite. The hydrophilicity, mechanical properties, porosity, filtration efficiency of the strontium-containing samples, water flux, and fouling tendency were used to assess the performance of the synthesized membranes. The effect of feed water temperature on water flux was investigated as well as its effect on salt rejection. As the temperature increased from 30 to 90 °C, the salt rejection decreased from 96.6 to 82% for the optimized 0.7 PES/TNR-MWCNT membrane, whereas the water flux increased to ≈150 kg/m2. h. Double successive filtration was evaluated for its high efficiency of 1000 ppm strontium removal, which reached 82.4%.


Introduction
Strontium is a rare element that only makes up 0.048% of the earth's surface [1]. It has oxidation states ranging from 0 to 2, and its most common forms are celestite SrSO 4 and strontianite SrCO 3 [2]. The separation of strontium from water may be required in a variety of situations, including drinking water and wastewater treatment. For the first, the Federal-Provincial-Territorial Committee on Drinking Water recommends a maximum allowed content of 7.0 mg/L in drinking water [3], while the US Environmental Protection Agency (EPA) set a health reference threshold of 1.5 mg/L [4]; strontium levels in stream water range from 0.001 to 13.6 mg/L [5]. For wastewater, as produced water accompanied by oil in petroleum production companies contains from a slightly low to a very high concentration of strontium, these strontium traces result from the dissolution of reservoir shale rocks containing mineral clay celestine (SrSO 4 ) [6] in produced water accompanied by oil in the form of free water or emulsion. Strontium concentrations in produced water in the Nanyishan oil-field brine of Qaidam, for example, reach 5364 mg L −1 [3], which is generally higher than the minimal concentration for industrial exploitation. In addition, a

Synthesis Titanium Nanoribbons (TNR)
Gradually, 10 g of TiO 2 powder was added to one liter of sodium hydroxide (concentration 10 Molar) with continuous stirring for five hours; then, the solution was heated in a Teflon-lined stainless steel autoclave for 170 • C for 24 h to form sodium titanate nanoribbons. After cooling at 15 • C, the produced powder was washed with dilute HCl (1 Molar); then, it was filtered and rinsed with distilled H 2 O and dried over 60 • C for 7 h for the formation of hydrogen titanium nanoribbons from sodium titanate nanoribbons; finally, it was calcined at 350 • C for five hours for dehydration into TiO 2 -B [22].

Synthesis of Catalyst TNR-MWCNT Nanocomposite Powder
The catalyst for the growth of carbon nanotubes is synthesized by dissolving the following chemicals in 100 mL of water: TNR, ferric nitrate, cobalt nitrate, and aluminum nitrate in the ratios of 10:20:20:50, respectively. Then, ammonia is added while stirring until pH 8 is reached and precipitation forms. After complete precipitation, the precipitated product was filtered and dried at 40 • C for 6 h; then, it was calcined at 450 • C for 6 h [23]. Tubular chemical vapor deposition (CVD) with functionalized TNRs was utilized for the growth of MWCNTs as a catalyst using C 2 H 4 (carbon source) and N 2 (carrier gas) with a ratio of 1:10 v/v at 700 • C for 50 min. Then, the product prepared was purified by adding the produced powder to 20 mL of nitric acid (8 molar) and sonicated for 5 h, and then, 20 mL of hydrochloric acid (5 molars) was added after sonication and refluxed at 120 • C for 3 h. Finally, the product is filtered and rinsed using distilled H 2 O and dried at 75 • C for 24 h [24].

Fabrication of PES/Nanocomposite Blended Membranes
Membranes were produced by the phase inversion method by adding nanocomposite powder (TNR-MWCNT) to 50 mL of DMF gradually [25]. The quantity of added nanocomposite powder was 0.1, 0.5, and 0.7 g to prepare the 0.1 PES/ TNR-MWCNT membrane, 0.5 PES/ TNR-MWCNT membrane, and 0.7 PES/ TNR-MWCNT membrane while stirring over 3 h; then, it was completed by DMF to 100 mL. After stirring for 6 h to form a homogeneous emulsion, 17.5 gm PES tablets were gradually added to the emulsion over an ultrasonic vibrator at 45 • C and stirred for 12 h. Then, they were left in the refrigerator for 12 h to expel air bubbles before being cast on a clean glass plate of 150 µ thickness using a thin film casting device. Afterwards, they were immersed in a water bath, and the casted membranes were removed and dried in the air [26].

Hydrophilicity of Synthesized Membrane
The contact angles of the membranes before and after adding the nanocomposite were measured using the sessile drop method to determine the nanocomposite addition effect on hydrophilicity.

Morphology of Synthesized Membranes
A field emission scanning electron microscope FE-SEM (Model FEG 250, Quanta, Tokyo, Japan) was used for characterization of the synthesized membranes and to define the effect of adding nanocomposite and its concentration on PES membranes, and the transmission electron microscope (TEM, Model 2010, JEOL, Tokyo, Japan) was used to scan the synthesized MWCNT-TNR nanocomposite.

Fourier Transform Infrared Examination
Pure PES membrane and PES blended with nanocomposite membranes were examined using FTIR (Bruker-Vertex 70, Bruker Optics Inc., Billerica, MA, USA).  To assess membrane tensile strength and elongation, a mechanical testing system  (INSTRON-5500R, INSTRON, Norwood, MA, USA) was employed. The gauge length and breadth of the dumbbell tensile specimens were 6.2 mm and 0.16 mm, respectively. After a membrane specimen had been inserted between the grips of the testing equipment, the tensile strength and elongation were determined. The measurements are within 5% of each other's accuracy.
Porosity was determined by immersing the manufactured membranes in distilled water. Then, the surplus water was filtered out using filter paper to determine the weight of the wet membrane (W2). After drying for 24 h at 85 • C, the membrane was weighted to determine W1 (dry membrane) [27]. Membrane porosity is calculated using Equation (1): Here, A is the membrane surface area (m 2 ), I is membrane thickness (m), and dw is water density (998 kg.m −3 ).

Filtration Efficiency
The water sample contains 100 ppm of strontium chloride, which was prepared by dissolving 0.1 gm of strontium chloride in 1 L of distilled water and stirring for 2 h. Then, filtration was carried out using a pressurized system as in Figure S1 (Supplementary Data). The system contains a peristaltic pump to suck a strontium solution sample and pump it to the membrane cell at a 45 psi discharge pressure. The membrane filter cell has an inlet, outlet, and vent that expels water not passed through the filter to maintain pressure in the required range.

Relationship between Concentration and Absorbance
Different concentrations of strontium were prepared. Then, the absorbance of its solution was measured using a spectrophotometer (Lambda 950 UV/VIS, PerkinElmer, Boston, MA, USA) to determine the relationship between absorbance and concentration.

Evaluation of Filtration Efficiency Using Different Membranes at Different Temperatures
Filtration of a sample (100 ppm Sr 2+ ) using a filtration apparatus with three different membranes (0.1, 0.5, 0.7) PES/TNR-MWCNT at temperatures of 30 • C, 45 • C, 60 • C, 75 • C, and 90 • C was performed, and absorbance was measured using a Perkin Elmer spectrophotometer to determine the filtration efficiency using the following equation: where C o is blank absorption and C is permeated solution absorption. In addition, water flux at a discharge pressure of 45 psi was measured by observing permeated quantity every 30 min, with a reduction in permeate quantity for the first 60 min, and then, a constant water flux was attained [28]. Water flux was calculated using the following equation: where J w is water flux (kg/m 2 .h), M is the permeate water weight (kg), A is the area of membrane (m 2 ), and ∆T is consumed time (h). To check if it can result in more filtration efficiency, double successive filtrations were carried out. The filtration efficiency was evaluated by the spectrophotometric measurements for the single filtration and double-successive filtration in a comparative manner.

Fouling Tendency
The fouling tendencies of membranes were investigated by a dynamic method. To utilize as an excellent fouling agent, a solution of 8000 mg/L milk powder was made. The first filtration using synthesized membranes and pure water was performed for 30 min at 45 psi, and water flux was recorded (J w1 ) Then, a synthesized solution was used to perform filtration at the same parameters of time and pressure using a solution of 8000 mg/L milk. Then, the membranes were washed using deionized water and repeated filtration using pure water while recording related water flux (J W2 ). The flux recovery ratio (F RR ) is obtained by [29]: F RR = Jw 2 /Jw 1 (4)

Hydrophilicity of Synthesized Membrane
The contact angles of the synthesized membranes were measured using the sessile drop method [30], which involved measuring the angle between the water droplet and the dry membrane surface immediately after a water droplet was dropped on the membrane surface, then repeating the process five times and recording the average of the readings. The contact angles for pure PES, 0.1 PES/TNR-MWCNT, 0.5 PES/TNR-MWCNT, and 0.7 PES/TNR-MWCNT membranes were 84, 81, 55, and 47, respectively [31]. This indicates that hydrophilicity increases by an increasing percentage of embedded nanocomposite in the PES matrix.

TNRs/MWCNTs Nanocomposite
Transmission electron microscopy (TEM) of a nanocomposite reveals wide, long, and straight titanium nanoribbons with nanopits dispersed across the surface [32] that serve as a substrate for the growth of multi-walled carbon nanotubes (MWCNTs). MWCNTs having inner and outer tube widths of 7 to 10 nm and 18 to 20 nm, respectively, and average nanoribbon widths of 15 to 150 nm are shown in Figure 1A. Figure 1B illustrates the distribution of nanopits on the TiO 2 surface, while Figure 1C illustrates the networking of MWCNTs and spongy nanoribbons.

Pure PES Membrane
SEM examination of a pure PES membrane reveals the development of a top layer that is denser and less porous, while the sub-layer has more micro pores, as shown in Figure 2A. The total thickness of the pure PES membrane ranges from 110 to 120 µm. The top view, as shown in Figure 2B, depicts the distribution of micro pits on the surface. The inset histogram showed the pore diameter distribution, whereas the average pore diameter is 1.51 µm. The difference in density and porosity between the top and sub-layers was caused by the solvent/non-solvent exchange mechanism, in which after the casted membrane was soaked in a distilled water bath, the water-immiscible polymer (PES) began to create repulsion forces with water particles, causing the coagulation and precipitation of membrane polymer particles. Simultaneously, the immiscibility of DMF in water caused the partial diffusion of DMF particles out of the casted membrane's polymer matrix, which was replaced by water particles, causing repulsion between polymer nuclei to the outer boundaries, resulting in a denser, less porous top layer and more porous sub-layer [33,34].

Pure PES Membrane
SEM examination of a pure PES membrane reveals the development of a top layer that is denser and less porous, while the sub-layer has more micro pores, as shown in Figure 2A. The total thickness of the pure PES membrane ranges from 110 to 120 µm. The top view, as shown in Figure 2B, depicts the distribution of micro pits on the surface. The inset histogram showed the pore diameter distribution, whereas the average pore diameter is 1.51 µm. The difference in density and porosity between the top and sub-layers was caused by the solvent/non-solvent exchange mechanism, in which after the casted membrane was soaked in a distilled water bath, the water-immiscible polymer (PES) began to create repulsion forces with water particles, causing the coagulation and precipitation of membrane polymer particles. Simultaneously, the immiscibility of DMF in water caused the partial diffusion of DMF particles out of the casted membrane's polymer matrix, which was replaced by water particles, causing repulsion between polymer nuclei to the outer boundaries, resulting in a denser, less porous top layer and more porous sub-layer [33,34].

Pure PES Membrane
SEM examination of a pure PES membrane reveals the development of a top layer that is denser and less porous, while the sub-layer has more micro pores, as shown in Figure 2A. The total thickness of the pure PES membrane ranges from 110 to 120 µm. The top view, as shown in Figure 2B, depicts the distribution of micro pits on the surface. The inset histogram showed the pore diameter distribution, whereas the average pore diameter is 1.51 µm. The difference in density and porosity between the top and sub-layers was caused by the solvent/non-solvent exchange mechanism, in which after the casted membrane was soaked in a distilled water bath, the water-immiscible polymer (PES) began to create repulsion forces with water particles, causing the coagulation and precipitation of membrane polymer particles. Simultaneously, the immiscibility of DMF in water caused the partial diffusion of DMF particles out of the casted membrane's polymer matrix, which was replaced by water particles, causing repulsion between polymer nuclei to the outer boundaries, resulting in a denser, less porous top layer and more porous sub-layer [33,34].  showing that the top layer is denser and less porous than the PES membrane, while sublayer macrovoids occur with an asymmetric distribution along the membrane thickness. The formation of macrovoids can be attributed to the hydrophilic nature of the added nanocomposite, which supports mass transfer rates of solvent and non-solvent [34], allowing water particles to pass through the hydrophobic polymer matrix faster, resulting in repulsion forces that cause macrovoids to form. Figure 3B shows MWCNTs grown on the TNRs macrovoids for the 0.1 PES/TNR-MWCNT membranes. The inset image showed the growth of CNT with a diameter of 20 nm on the surface of TNR. In cross-sectional SEM images for 0.5 PES/TNR-MWCNT, the nanocomposite distribution increases, resulting in the formation of more macrovoids that extend along with the membrane thickness, as shown in Figure 3C,D. This can be attributed to an increase in the hydrophilic nanocomposite, which increases solvent/non-solvent diffusion. In the case of 0.7 PES/TNR-MWCNT, Figure 3E shows that the high concentration of nanocomposite, which leads to high viscosity, resulted in the formation of multilayers. In addition, it resulted in the accumulation of nanocomposite at certain points across the PES matrix, as shown in Figure 3F. From Figure 3A,C,E, the average thickness of the PES/MWCNT-TNR membranes is 140 µm. The increase in skin layer thickness to 140 µm is attributed to a decrease in solvent/non-solvent diffusion due to the increased viscosity. A large number of pores of various sizes were produced.
SEM images for 0.5 PES/TNR-MWCNT, the nanocomposite distribution increases, resu ing in the formation of more macrovoids that extend along with the membrane thicknes as shown in Figure 3C,D. This can be attributed to an increase in the hydrophilic nan composite, which increases solvent/non-solvent diffusion. In the case of 0.7 PES/TN MWCNT, Figure 3E shows that the high concentration of nanocomposite, which leads high viscosity, resulted in the formation of multilayers. In addition, it resulted in the a cumulation of nanocomposite at certain points across the PES matrix, as shown in Figu 3F. From Figure 3A,C,E, the average thickness of the PES/MWCNT-TNR membranes 140 µm. The increase in skin layer thickness to 140 µm is attributed to a decrease in so vent/non-solvent diffusion due to the increased viscosity. A large number of pores of va ious sizes were produced.

FTIR Spectroscopy
As indicated in Figure 5A,B, FTIR was used to investigate pure PES, PES/TNR, an PES/TNR-MWCNT membranes. FTIR results reveal the creation of bonds between fun tional groups of nanocomposite and the PES matrix.

FTIR Spectroscopy
As indicated in Figure 5A,B, FTIR was used to investigate pure PES, PES/TNR, and PES/TNR-MWCNT membranes. FTIR results reveal the creation of bonds between functional groups of nanocomposite and the PES matrix.

FTIR Spectroscopy
As indicated in Figure 5A,B, FTIR was used to investigate pure PES, PES/TNR, and PES/TNR-MWCNT membranes. FTIR results reveal the creation of bonds between functional groups of nanocomposite and the PES matrix.  Table 1. The wider and lowerintensity beaks of the TNR-PES membrane in Figure 5A were caused by the TiO2 suppressing effect [35][36][37][38]. Meanwhile, Figure 5B shows new function groups for PES/TNR-MWCNT, as assigned in Table 1.
The   Table 1. The wider and lower-intensity beaks of the TNR-PES membrane in Figure 5A were caused by the TiO 2 suppressing effect [35][36][37][38]. Meanwhile, Figure 5B shows new function groups for PES/TNR-MWCNT, as assigned in Table 1. The

Mechanical Properties
The membrane's physical characteristics were improved because of the incorporation of the MWCNT-TNR nanocomposite in the PES membrane. From Figure 6, the greatest tensile strength of the membrane when mixed with 0.5 wt% MWCNT-TNR was 94 kg/cm 2 , which reduced to 59.2 kg/cm 2 when combined with 0.7 wt% but was still greater than that of pure PES (38 kg/cm 2 ). When the membrane was blended with 0.1 wt% MWCNT-TNR, the maximum elongation was 32 mm, as shown in Figure 6. The elongation was reduced to 15 mm at 0.7 wt% MWCNT-TNR because of the high viscosity of the cast solution and the coalescence effect [39]. The blended PES membranes are strengthened by their large surface area, high aspect ratio, and good interaction with MWCNT-TNR [40,41].
The porosity of the prepared membranes was determined using Equation (1). The porosity of PES membrane was 30%. The porosity of membranes decreased as the content of MWCNT-TNR increased. As the composite content increased to 0.1 wt%, 0.5 wt%, and 0.7 wt%, the porosity was reduced to 25%, 13%, and 6%, respectively. A high nanocomposite content cast solution becomes viscous, whereas a low nanocomposite content cast solution becomes less viscous. This delayed membrane development and made the structure denser [41]. As a result, the pore size may be decreased or even blocked. The included nanocomposite also functioned as a nucleation agent during the phase separation process. As a result, membrane nucleation and growth rates increased, making macroporous structure loosening simpler [41].

Evaluation of Membranes Filtration Efficiencies
The absorbance spectra of 100 ppm Sr 2+ aqueous standard solutions were measured using a UV/Vis Perkin Elmer spectrophotometer, which has maximum absorption at 289 nm. This solution will be used as a feed solution for salt rejection evaluation by the membranes. In addition, the absorbance of various strontium concentrations (50, 75, 100, 150, 250, and 500 ppm) was determined, and a plot between absorbance and concentration was drawn. From the plot, a linear relationship appears, which indicates that solution concentrations obey Beer's Lambert law, so we can apply this method for determining the change in strontium concentration in permeate solutions.
Beer's Lambert law: A = L C (5) where A, , L, and C refer to the absorbance, molar absorptivity, cuvette length, and solution concentration, respectively.  The porosity of the prepared membranes was determined using Equation (1). The porosity of PES membrane was 30%. The porosity of membranes decreased as the content of MWCNT-TNR increased. As the composite content increased to 0.1 wt %, 0.5 wt %, and 0.7 wt %, the porosity was reduced to 25%, 13%, and 6%, respectively. A high nanocomposite content cast solution becomes viscous, whereas a low nanocomposite content cast solution becomes less viscous. This delayed membrane development and made the structure denser [41]. As a result, the pore size may be decreased or even blocked. The included nanocomposite also functioned as a nucleation agent during the phase separation process. As a result, membrane nucleation and growth rates increased, making macroporous structure loosening simpler [41].

Evaluation of Membranes Filtration Efficiencies
The absorbance spectra of 100 ppm Sr 2+ aqueous standard solutions were measured using a UV/Vis Perkin Elmer spectrophotometer, which has maximum absorption at 289 nm. This solution will be used as a feed solution for salt rejection evaluation by the membranes. In addition, the absorbance of various strontium concentrations (50, 75, 100, 150, 250, and 500 ppm) was determined, and a plot between absorbance and concentration was drawn. From the plot, a linear relationship appears, which indicates that solution concentrations obey Beer's Lambert law, so we can apply this method for determining the change in strontium concentration in permeate solutions.
Beer's Lambert law: A = ϵ L C (5) where A, ϵ, L, and C refer to the absorbance, molar absorptivity, cuvette length, and solution concentration, respectively. The absorption spectrum of 100 ppm Sr 2+ aqueous standard solution and the plot of the absorbance at 289 nm and Sr 2+ concentration are shown in Figure 7A. The 0.1, 0.5, and 0.7 PES/TNR-MWCNT membranes were used for the filtration of 100 ppm Sr 2+ aqueous     The absorption spectrum of 100 ppm Sr 2+ aqueous standard solution and the plot of the absorbance at 289 nm and Sr 2+ concentration are shown in Figure 7A. The 0.1, 0.5, and 0.7 PES/TNR-MWCNT membranes were used for the filtration of 100 ppm Sr 2+ aqueous solutions at 30 • C, 45 • C, 60 • C, 75 • C, and 90 • C. The measured absorption spectra using a UV/Vis spectrophotometer in the range of 250 to 550 nm is shown in Figure 7B-D. From absorption curves appear the values of the maximum absorbance for each filtration, and the filtration efficiencies are determined as shown in Table 2. As a result of the analyses, it appears that increasing the filtration temperature of the 0.1, 0.5, and 0.7 PES/TNR-MWCNT membranes decreases the salt rejection ratio. At 30 • C, the 0.7 membrane has the maximum salt rejection compared to the 0.5 and 0.1 membranes, which resulted from the agglomeration of nanocomposite at membrane macrovoids obstructing strontium ions and increasing hydrophilicity. As the temperature rises, the water flux and net salt ion transfer both increase, resulting in the salt rejection drops. It can be shown that temperature enhances the convection, electro-migration, and diffusion of salt ion flux. Temperature causes changes in membrane structural characteristics, solvent viscosity, and solute diffusivity, which all cause these effects [42]. It appears that water flux increases as filtration temperature rises, indicating that membrane pores expand as temperature rises, as thermal expansion coefficients for the PES range of (1.4-2) 10 −4 k −1 [43], while the thermal expansion coefficients (CTE) for CNT and titanium oxide are (1.6-2.6) 10 −5 K −1 [44,45] and (8.4-11.8) 10 −6 K −1 [46], respectively. These differences in thermal expansion coefficient resulted in an increase in PES pores, whereas nanocomposites with lower CTE were unable to compensate, enabling increased water flux as the temperature increases. Figure 8A shows that the salt rejection decreases by increasing filtration temperatures in 0.1, 0.5, and 0.7 PES/TNR-MWCNT. Figure 8B shows that 0.1 PES/TNR-MWCNT has the highest water flux, while the decrease in water flux in 0.5 PES/TNR-MWCNT is attributed to the decrease in pores size of PES by filling nanocomposite. Whereas the slight increase in water flux of 0.7PES/TNR-MWCNT compared to 0.5 PES/TNR-MWCNT may be attributed to the formation of multilayers due to coagulation of nanocomposite [47,48]. which resulted from the agglomeration of nanocomposite at membrane macrovoids obstructing strontium ions and increasing hydrophilicity. As the temperature rises, the water flux and net salt ion transfer both increase, resulting in the salt rejection drops. It can be shown that temperature enhances the convection, electro-migration, and diffusion of salt ion flux. Temperature causes changes in membrane structural characteristics, solvent viscosity, and solute diffusivity, which all cause these effects [42]. It appears that water flux increases as filtration temperature rises, indicating that membrane pores expand as temperature rises, as thermal expansion coefficients for the PES range of (1.4-2) 10 −4 k −1 [43], while the thermal expansion coefficients (CTE) for CNT and titanium oxide are (1.6-2.6) 10 −5 K −1 [44,45] and (8.4-11.8) 10 −6 K −1 [46], respectively. These differences in therma expansion coefficient resulted in an increase in PES pores, whereas nanocomposites with lower CTE were unable to compensate, enabling increased water flux as the temperature increases. Figure 8A shows that the salt rejection decreases by increasing filtration temperatures in 0.1, 0.5, and 0.7 PES/TNR-MWCNT. Figure 8B shows that 0.1 PES/TNR-MWCNT has the highest water flux, while the decrease in water flux in 0.5 PES/TNR-MWCNT is attributed to the decrease in pores size of PES by filling nanocomposite Whereas the slight increase in water flux of 0.7PES/TNR-MWCNT compared to 0.5 PES/TNR-MWCNT may be attributed to the formation of multilayers due to coagulation of nanocomposite [47,48].

Successive Filtration of Highly Concentrated Sr 2+ Aqueous Solutions
The filtration efficiency of a sample of strontium with a concentration of 1000 ppm was evaluated using double successive filtration utilizing a 0.7 PES/TNR-MWCNT membrane and measuring the absorbance of spectra of permeate in the wavelength range of 300 to 1000 nm, as shown in Figure 9A. Figure 9B shows that the filtration efficiency is increased from 74.5% to 82.4% after the double filtration.

Membranes Fouling Tendencies and Impacts
The fouling tendency was investigated by measuring a flux recovery ratio (FRR) based on deionized water flux before (Jw1) and after (Jw2) filtration of a solution of 5000 ppm of milk powder as a fouling agent solution using synthesized membranes. The higher FRR value indicates a higher antifouling tendency and lower fouling. The FRR values for 0.1PES/TNR-MWCNT, 0.5PES/TNR-MWCNT, and 0.7PES/TNR-MWCNT were 72.7%, 78.7%, and 85.2%, respectively, indicating that the antifouling tendency increased by increasing the ratio of the embedded nanocomposite. The increasing antifouling tendency was attributed to the increased hydrophilicity due to the incorporated nanocomposite [48].
The industrial impact of this research can be concluded as follows. The hydrothermal technique of synthesizing TNR produced nanoribbons with nano pits on their surfaces, which were employed as catalyst support surfaces to grow CNTs on the surfaces of TNRs. There was no agglomeration of nanoparticles while utilizing tubular chemical vapor deposition (CVD) in the synthesis of CNT using C2H4 as the carbon source and N2 as the carrier gas. As a result, the combined approaches utilized in this study construct a TNR-MWCNT network and prevent agglomeration, which is one of the challenges in nanoparticle manufacture. Furthermore, several studies looked at nanofiltration for desalination, but just a few looked into removing strontium from water. Meanwhile, the threat of strontium contamination has increased as a result of high levels of strontium identified in wastewater from the oil and gas industry that exceed permitted limits. Almost Sr-contaminated wastewater is dumped in the sea or injected into wells, posing a risk of contamination of nearby underground water sources. Using our manufactured PES/MWCNT-TNR membrane saves space, increases efficiency, eliminates the need for chemicals, and increases water flow when compared to other standard techniques. Since the produced wastewater has a high temperature after separation from petroleum products in oil treatment stations at petroleum enterprises that utilize demulsifiers and heaters, the influence of feed water temperature on filtering efficacy is significant, as shown in this study. Furthermore, the simple fabrication of PES/MWCNT-TNR membranes will be beneficial in large-scale manufacturing for use in petroleum industries, as well as increasing tensile strength from 38 kg/cm 2 for pure PES membrane to 94 kg/cm 2 for 0.5 wt % PES/MWCNT-TNR membranes, which will improve membrane endurance in industrial applications. Meanwhile, this can alleviate the problem of generated water containing excessive levels of strontium, which is a health hazard. This sort of water treatment will be more beneficial at oil-producing off-shore sites where space is limited and the primary means of disposing

Membranes Fouling Tendencies and Impacts
The fouling tendency was investigated by measuring a flux recovery ratio (F RR ) based on deionized water flux before (J w1 ) and after (J w2 ) filtration of a solution of 5000 ppm of milk powder as a fouling agent solution using synthesized membranes. The higher F RR value indicates a higher antifouling tendency and lower fouling. The F RR values for 0.1PES/TNR-MWCNT, 0.5PES/TNR-MWCNT, and 0.7PES/TNR-MWCNT were 72.7%, 78.7%, and 85.2%, respectively, indicating that the antifouling tendency increased by increasing the ratio of the embedded nanocomposite. The increasing antifouling tendency was attributed to the increased hydrophilicity due to the incorporated nanocomposite [48].
The industrial impact of this research can be concluded as follows. The hydrothermal technique of synthesizing TNR produced nanoribbons with nano pits on their surfaces, which were employed as catalyst support surfaces to grow CNTs on the surfaces of TNRs. There was no agglomeration of nanoparticles while utilizing tubular chemical vapor deposition (CVD) in the synthesis of CNT using C 2 H 4 as the carbon source and N 2 as the carrier gas. As a result, the combined approaches utilized in this study construct a TNR-MWCNT network and prevent agglomeration, which is one of the challenges in nanoparticle manufacture. Furthermore, several studies looked at nanofiltration for desalination, but just a few looked into removing strontium from water. Meanwhile, the threat of strontium contamination has increased as a result of high levels of strontium identified in wastewater from the oil and gas industry that exceed permitted limits. Almost Sr-contaminated wastewater is dumped in the sea or injected into wells, posing a risk of contamination of nearby underground water sources. Using our manufactured PES/MWCNT-TNR membrane saves space, increases efficiency, eliminates the need for chemicals, and increases water flow when compared to other standard techniques. Since the produced wastewater has a high temperature after separation from petroleum products in oil treatment stations at petroleum enterprises that utilize demulsifiers and heaters, the influence of feed water temperature on filtering efficacy is significant, as shown in this study. Furthermore, the simple fabrication of PES/MWCNT-TNR membranes will be beneficial in large-scale manufacturing for use in petroleum industries, as well as increasing tensile strength from 38 kg/cm 2 for pure PES membrane to 94 kg/cm 2 for 0.5 wt% PES/MWCNT-TNR membranes, which will improve membrane endurance in industrial applications. Meanwhile, this can alleviate the problem of generated water containing excessive levels of strontium, which is a health hazard. This sort of water treatment will be more beneficial at oil-producing off-shore sites where space is limited and the primary means of disposing of generated water is by injecting it into disposal wells, where strontium will develop scales that obstruct well permeability and reduce water injection rate.

Conclusions
Membranes of PES blended with MWCNT-TNR were synthesized with different ratios of nanocomposite and characterized using TEM, SEM, and FTIR. TEM scanning shows the formation of a network between TNRs and MWCNTs. SEM showed that the top layer of formed membranes is dense and less porous while the sub-layer has macrovoids. SEM also showed the formation of macrovoids in the PES matrix after the incorporation of nanocomposite and the effect of nanocomposite ratio on macrovoid size and distribution, while FTIR scanning proved the formation of a bond between PES and TNR-MWCNT nanocomposite. The pore diameter and porosity are decreased by increasing the ratio of the incorporated nanocomposite. The tensile strength was enhanced from 38 kg/cm 2 for pure PES membrane to 94 kg/cm 2 for the 0.5 wt% PES/MWCNT-TNR membranes. The synthesized membranes are used for water filtration of strontium at different temperatures. Salt rejection has decreased as the temperature of the water has risen, while water flux increases by elevating the temperature due to the increasing diameter of PES matrix pores and changes in membrane structural characteristics, solvent viscosity, and solute diffusivity. At 30 • C, the highest filtration efficiency was 96.6% for the 0.7 wt% PES/TNR-MWCNT membrane, compared to 94.8% for the 0.5 wt% PES/TNR-MWCNT membrane and 94.1% for the 0.1 wt% PES/TNR-MWCNT membrane. The highest water flow was 84.4 kg/m 2 .h for the 0.1 wt% PES/TNR-MWCNT, while the lowest was 76.6 kg/m 2 .h for 0.7 wt% PES/TNR-MWCNT. The successive filtration of strontium increased filtration efficiency from 74.5% to 82.4%, which indicated that better efficiency could be achieved by adding multiple filtration steps. By increasing the incorporated ratio of nanocomposite, the antifouling properties of the membranes increased.