TiO2 decorated functionalized halloysite nanotubes (TiO2@HNTs) and photocatalytic PVC membranes synthesis, characterization and its application in water treatment

In this study photocatalyst, TiO2@HNTs were prepared by synthesizing TiO2 nanoparticles in situ on the functionalized halloysite nanotubes (HNTs) surface. Photocatalytic PVC membrane TiO2@HNTs M2 (2 wt.%) and TiO2@HNTs M3 (3 wt.%) were also prepared. Photocatalyst TiO2@HNTs and photocatalytic PVC membranes were used to study the photocatalytic activity against the methylene blue (MB) and rhodamine B (RB) dyes in UV batch reactor. The structure and morphology of photocatalyst and photocatalytic PVC membrane were characterized by fourier transform infrared spectroscopy (FT-IR), X-ray diffraction (XRD), scanning electron microscopy (SEM), energy dispersive X-ray (EDX), transmission electron microscopy (TEM), UV-Vis spectrophotometer and photoluminescence (PL). The PL study showed that the oxygen vacancies and surface hydroxyl groups present on the surface of TiO2@HNTs act as excellent traps for charge carrier, reducing the electron-hole recombination rate.TiO2@HNTs 2 (2 wt.%) and TiO2@HNTs 3 (3 wt.%) degraded MB dye up to 83.21%, 87.47% and RB dye up to 96.84% and 96.87%, respectively. TiO2@HNT photocatalyst proved to be stable during the three consecutive cycle of photocatalytic degradation of the RB dye. TiO2@HNTs M2 and TiO2@HNTs M3 degraded MB dye up to 27.19%, 42.37% and RB dye up to 30.78%, 32.76%, respectively. Photocatalytic degradation of both the dyes followed the first-order kinetic model. Degradation product analysis was done using the liquid chromatography–mass spectrometry (LC-MS) and the results showed that the dye degradation was initiated by demethylation of the molecule. MB and RB dye degradation reaction were tested by TBA and IPA as OH* and H+ scavengers respectively. Mechanism of photocatalytic activity of TiO2@HNTs and photocatalytic PVC membrane were also explained.


Result and Discussions
Characterization of tio 2 @HNts and photocatalytic pVC membranes. FT-IR spectra as shown in Fig. 1(a), confirmed the presence of functionalized HNTs after modification. When raw HNTs 48 were compared with the TiO 2 @HNTs, some new FTIR peaks were observed in the spectrum like stretching CH 2 vibration band around 2934 cm −1 , and the deformation CH 2 and Si-CH vibration at 1627 cm −1 and1507 cm −1 respectively . Apart from these peaks, broad peak of O-H stretching of water at 3433 cm −1 was also seen. These observations confirmed the presence of silane coupling agent 49 . Silane coupling agent is used because its main function is to ensure proper bonding between TiO 2 and HNTs [49][50][51][52] . If this functionalization is not done, due to the absence of chemical conjugation 53 , the TiO 2 may leach out during the course of the experiment because of the less adhesive nature of raw HNTs 54,55 . The TiO 2 @HNTs possess some significant signals, like distortions of aluminium-oxygen-silicon and silicon-oxygen-silicon bonds at 538 and 468 cm −1 respectively; and -OH groups of the inner hydroxyl groups at 909 cm −1 . Furthermore, for comparison, the broad stretching band of silicon-oxygen at about 1037 cm −1 shifts to about 1058 cm −1 , indicating hydrogen bonding between TiO 2 and HNTs 21 . In Fig. 1(b), ATR-FT-IR spectra of TiO 2 @HNTs M0 and TiO 2 @HNTs M3 with the principle bands of PVC were depicted. FT-IR spectrum of PVC reflects expected distinctive absorptions: 2970-2912 cm −1 attributed to stretching C-H of CHCl and stretching C-H of CH 2 group, 1435 cm −1 and 1427 cm −1 attributed to deformation wagging of CH 2 group, and 1331-1255 cm −1 shows the stretching of the C-H of CHCl groups. Stretching C-C (1092 cm −1 ), rocking CH 2 (811 cm −1 ), stretching C-Cl attributed to 688 cm −1 and 618 cm −1 respectively 56 . In Fig. 1b, the peaks of photocatalyst PVC membrane in the ATR-FT-IR results resembles the characteristic peaks belonging to both TiO 2 @HNTs M0 and TiO 2 @HNTs M3. Other characteristic absorption peaks for TiO 2 @HNTs in the spectrum of the membrane was not clearly identified because of the overlap with absorption peaks of PVC polymer or the IR beam might not be able to penetrate properly enough to get distinct peaks of HNTs 57 in FTIR spectra (Fig. 1b).
The results of XRD clearly shows the specific peaks as shown in Fig. 1(c) which compares the XRD results of HNTs and TiO 2 @HNTs synthesized. The peaks depicted for HNT sample can be translated into the characteristic peaks of halloysite shown in Fig. 1(c). Two fresh peaks, however, can be noticed at 2θ = 48° and 54.1° and a stronger peak at 2θ = 25.3° alongside with decline in the halloysite peaks due to the sol-gel method. Based on JCPDS 21-1272, all peaks related to TiO 2 properties can be indexed to the (101), (004), (200) and (105) planes of TiO 2 structure 58,59 . This verifies the successful preparation of TiO 2 @HNTs. Similar results of XRD pattern also described by the Ghanbari et al., in their study for the fabrication of high-performance thin film nanocomposite membranes 60 . Furthermore, in the EDX spectrum (Fig. 1d) confirms the loading of TiO 2 nanoparticles on the surface of HNTs with 24.42 wt%. Intense peaks of Ti and oxygen at 0.5 eV were observed which confirmed the presence of TiO 2 nanoparticles on the surface of HNTs.
Thermal behaviour of TiO 2 @HNTs and TiO 2 @HNTs/PVC blends was investigated from room temperature to 600 °C temperature range. The temperature was raised at the rate of 10 °C/min (Fig. 2). The raw HNTs showed a weight loss between 50 °C and 150 °C, which may be due to reduction in adsorbed water molecules 61 ; and structural dehydroxylation of structural Al-OH groups between 450-550 °C 62 . Decomposition of the (3-Aminopropyl) triethoxysilane causes an additional weight loss in TiO 2 @HNTs between 250 °C and 425 °C when compared with the raw HNTs 63 suggesting that the thermal stability, as well as the purity of the nanotubes, was high. PVC membrane (TiO 2 @HNTs M0-0 wt.%) and TiO 2 @HNTs membranes (TiO 2 @HNTs M2 and TiO 2 @HNTs M3) showed a different behaviour. The weight loss of the sample (around 60%) was mainly between temperature range 220 °C to 300 °C. The rest of the sample was thermally decomposed at 500 °C as shown in Fig. 2   Figure 1. FT-IR spectra of raw HNTs and TiO 2 @HNTs (a), raw PVC membrane (TiO 2 @HNTs M0) and photocatalytic PVC membrane (TiO 2 @HNTs M3) (b), XRD peaks of HNTs and TiO 2 @HNTs (c) and (d) EDX spectra of TiO 2 @HNTs photocatalyst.
www.nature.com/scientificreports www.nature.com/scientificreports/ FEG-SEM and TEM were done to analyse the structural and morphological characteristics of raw HNTs and TiO 2 @HNTs photocatalyst (Fig. 3). The TEM images of raw HNTs clearly show the hollow tubular structure with a diameter of 50-70 nm and length of 0.5-2 µm (Fig. 3c). The outer surface of HNTs was made up of silica and was surrounded by multi-walled aluminol layer. In Fig. 3b,d TiO 2 nanoparticles were seen to be randomly deposited on the surface of HNTs. Few aggregation or cluster of nanoparticles was observed in FEG-SEM and TEM analysis (Fig. 3b,d) which was formed when photocatalyst was separated from the solution during the fabrication process. The size of small particles and cluster size of this TiO 2 range from 10 nm to 80 nm. The presence of TiO 2 nanoparticles on the surface of HNTs confirms successful loading on HNTs.
All the membrane possessed typical asymmetric structure like that of ultrafiltration membranes and no distinct difference between the TiO 2 @HNTs M0 (Fig. 4a) and the TiO 2 @HNTs M2 and TiO 2 @HNTs M3 (Fig. 4b,c)  www.nature.com/scientificreports www.nature.com/scientificreports/ on the top surface was observed. Only large pores were observed on the membrane surface as shown in Fig. 4(b,c) when compared with TiO 2 @HNTs M0 (Fig. 4a). FEG-SEM images shows top surface and cross-sectional view of TiO 2 @HNTs M0 and TiO 2 @HNTs M2 and TiO 2 @HNTs M3 membranes. All membrane samples were homogeneous and asymmetric in nature with macro-voids and finger-like structures formed because of the high mutual diffusivity of DMAc and water. This porous structure indicates that TiO 2 @HNTs increases the membrane porosity without altering the membrane morphological structure 64 .
To investigate the recombination of the free carrier in HNTs, photoluminescence (PL) emission spectrum was analysed. One broad PL peak centred under excitation of 450 nm visible light irradiation was observed in Fig. 5a. The broad-band at 624 nm indicates that the light for excitation can initiate electron transition from the valence band (VB) to the conduction band (CB) in HNTs. As a result of this transition, the electron/hole pair can be generated which then further recombine radiatively to give broad and strong PL signal under 450 nm www.nature.com/scientificreports www.nature.com/scientificreports/ light irradiation. The PL emission spectrum of TiO 2 @HNTs was recorded at an excitation wavelength of 220 nm (high absorption region). The major peak at 361 nm (lower than the band edge emission) in PL emission spectra ( Fig. 5b) was due to the band to band transition. At longer wavelengths in the visible region, the emission peaks at 421, 445, 481, 535 and 556 nm reflect the surface state emissions, located within the band gap of TiO 2 @HNTs 65 . The oxygen vacancies and surface hydroxyl groups present on the surface of TiO 2 and HNTs acts as excellent traps for charge carrier, reducing the electron-hole recombination rate. Presence of oxygen vacancies from shallow trap state near the adsorption band edge act as efficient electron trap centres or colour centres. UV-Vis spectroscopy was used to study the light absorption properties. Figure 5c displays the UV-Vis absorption spectra of the prepared materials obtained. The band gap of pure TiO 2 is 3.2 eV and the absorption edge of the TiO 2 @HNTs material is around 220 nm corresponding to the band gap energy (Eg) of 3.66 eV (Fig. 5d). This difference is due to the addition of HNTs because of which the modified photocatalyst gets excited to produce more electron-hole pairs under light irradiation, resulting in higher photocatalytic activity 66,67 . photocatalytic activities of tio 2 @HNts photocatalyst and photocatalytic pVC membranes for the degradation of MB and RB dye solutions. In Fig. 6 images of MB and RB dye degradation under 120 min of UV irradiation can be seen. Initially, the colour of MB dye solution was blue and RB dye solution was pink. After 120 min, the colour of TiO 2 @HNTs 2 and 3 reaction mixture changed from blue to light blue in case of MB dye and from pink to colourless solution in case of RB dye. This decolourization was due to degradation of dye and was highest in the case of TiO 2 @HNTs 3. The chemical structure of MB and RB dye is shown in Fig. 7(a,b). To further confirm the degradation of dye, LC-MS chromatographic separation analysis was done. For the analysis, the dye solution was collected after 120 min of UV irradiation. As seen in Fig. 8(a-d), maximum absorption of MB dye in the visible region was at 664 nm and in the UV region two peaks located at 245 and 292 nm. The highest peak of MB dye at 664 nm is due to its centre benzene ring comprising sulphur and nitrogen whereas the two dimethylamine substituted aromatic ring exhibits its peak in the UV region at 245 and 292 nm. These high peaks at 664 nm gradually diminish with time in the presence of TiO 2 @HNTs 3 under UV irradiation. Similarly, for RB dye, the characteristic absorption peak at 554 nm as shown in Fig. 9(a-d) also degrades with time. The n → π transition of C=N, C=O groups in the aromatic ring of the RB dye structure (Fig. 7b) is responsible for the colour of solution and its absorbance at 554 nm. During the reaction process, when the dye structure was disrupted by the TiO 2 @HNTs 3 photocatalyst, the absorption intensity of RB dye decreased rapidly with change in colour from pink to a colourless solution. www.nature.com/scientificreports www.nature.com/scientificreports/   www.nature.com/scientificreports www.nature.com/scientificreports/ The shifts in the absorption spectra of both the dye solutions were studied during photocatalytic degradation analysis (Figs 8 and 9). The absorption peak of MB dye solution was at 664 nm which decreased gradually due to dye degradation and attained its lowest value at 120 min ( Fig. 8a,b). The reason for the decrease in absorption peaks was attributed to the fact that TiO 2 @HNTs photocatalyst cleaves the aromatic ring of the dye molecules and initiate its degradation 68 . In the case of RB dye solution, highest and characteristic peak was seen at 562 nm, which dipped quickly in the first 15 min of degradation and was lowest after 120 min ( Fig. 9a,b). The MB dye was degraded up to 83.21% and 87.47% (Fig. 8a,b) for TiO 2 @HNTs 2 and TiO 2 @HNTs 3 respectively. In the case of RB dye, 2 wt.% of photocatalyst was sufficient enough for degradation of 20 mg/L of RB dye and the degradation rate for both the 2 wt.% and 3 wt.% photocatalyst was 96.84% and 96.87% (Fig. 9a,b) respectively. Blue Shift in absorption peaks (λ max) was observed gradually with time which finally resulted in the respective lowest absorption peak of dye. The adsorption capacity of prepared TiO 2 @HNTs was also tested with MB and RB dye in dark condition at room temperature with 20 mg/L dye solution. The maximum dye adsorption efficiency of MB dye and RB dye were 17.5 mg/g and 4.8 mg/g for TiO 2 @HNTs 2 and 12.10 mg/g and 3.67 mg/g for TiO 2 @HNTs 3 respectively after 2 h. In the control experiment (HNTs 2 and HNTs 3), the removal of the MB dye was around 39.22% and 47.82% and RB dye was around 17.65% and 25.2% respectively after 2 h of irradiation. Also, there was a minor reduction in UV spectra even when TiO 2 @HNTs photocatalyst was added and both MB and RB dye solution was kept in dark. Thus it was confirmed that the degradation was due to TiO 2 nanoparticles in the presence of UV irradiation.
Though photocatalytic PVC membranes TiO 2 @HNTs M2 and TiO 2 @HNTs M3 both exhibited photocatalytic activity (as shown in Figs 8c,d and 9c,d for MB and RB dyes respectively), but TiO 2 @HNTs M3 membrane possessed more catalytic activity as compared to TiO 2 @HNTs M2 because with the increase in the TiO 2 @HNTs concentration the catalytic activity also enhances. However, it was observed that photocatalytic activity reduced in photocatalytic PVC membranes when compared with the same wt% of TiO 2 @HNTs alone. The MB dye was degraded up to 27.19%, 42.37% (Fig. 8c,d) and RB upto 30.78%, 32.76% (Fig. 9c,d) for TiO 2 @HNTs M2 and TiO 2 @HNTs M3 respectively. The reason behind this slow rate of degradation in both TiO2@HNTs M2 and TiO2@HNTs M3 can be due to the reduction of active sites of TiO 2 @HNTs photocatalyst during membrane preparation. During phase separation process, the increase in the amount of TiO 2 @HNTs photocatalyst increases its catalytic activity but its increase after 3 wt.% in membrane casting solution may result in photocatalyst agglomeration 69 . This agglomeration hampers the number of free catalytic sites for dye degradation and also delays the phase separation process during membrane preparation 46 . When the concentration of photocatalyst increases in the solution, the turbidity of the solution increases and light is scattered more due to which screening effect occurs hampering the specific activity of the catalyst and reducing the degradation rate [70][71][72] . Thus an optimal www.nature.com/scientificreports www.nature.com/scientificreports/ amount of photocatalyst must be used for photocatalytic degradation process for increased degradation and reduced inaccuracy.
The kinetics of photocatalytic reactions can be described using the first-order reaction for concentrations (20 mg/L) of MB and RB dye solutions. The rate constants (k) and the correlation coefficient (R 2 ) has been evaluated using linear regression curve of ln(C 0 /C) versus UV light irradiation time. First-order rate equations are as follows 73 : Here, k is the first order rate constant (min −1 ); [C] and [C 0 ] final and initial dye concentration in (mg/L), respectively.
The degradation efficiency has been calculated using 74 : where C 0 is the initial concentration of the dye and C is the concentration of the dyes after UV irradiation in the selected time interval. The correlation coefficient (R 2 ) was calculated to be nearly as high as R 2 ≈ 0.95-0.99 which reiterate the suitability of the first-order reaction listed in Table 1. MB and RB degradation rate are shown in Fig. 10c,d, where the TiO 2 @HNTs photocatalyst has notably improved the photocatalytic activity. During photocatalysis of RB dye by TiO 2 @HNTs 2 and 3, dye adsorption process occurred in two steps shown in Fig. 10d. At first stage, RB molecules diffuse from the aqueous solution to the external surface of TiO 2 @HNTs or the boundary layer diffusion of RB molecules. Secondly, gradual adsorption occurred until equilibrium was reached. The linear portion of the first stage did not pass through the origin, indicating the existence of a boundary layer resistance between TiO 2 @ HNTs photocatalyst and dye solution 75 Similar observation is reported in literature 73 . Furthermore, the detailed photocatalytic mechanism is described in detail in a later section.
The stability of the TiO 2 @HNTs photocatalyst was estimated by recycling the photocatalyst for degradation of RB dye under UV light irradiation for three times. The loss of the photocatalytic activity was negligible (approximately 6% after 3 repeated runs) (Fig. 11) which signifies the stability of TiO 2 @HNTs photocatalyst in terms of its non photo-corrosive nature during the photocatalytic degradation of the model dye molecules, which proved to be very important for its practical applications.
photocatalytic pVC membrane activity test with MB dye. To evaluate the photocatalytic property of the membrane surface, control (raw PVC membrane TiO 2 @HNTs M0), TiO 2 @HNTs M2 and TiO 2 @HNTs M3 samples were treated with UV light without stirring for 1 h (Fig. 12). After 1 h of irradiation TiO 2 @HNTs M2 and TiO 2 @HNTs M3, sample colour disappears from the membrane surface. On the contrary, the change in colour in control experiment was less after irradiation of UV light. The change in colour of membrane of the control sample was seen to be least when irradiated with UV light, which was contrary to the result of samples of TiO 2 @HNTs M2 and TiO 2 @HNTs M3. TiO 2 acts as a semiconductor and presence of UV light results in the formation of electrons and holes. These photo-generated electrons hence formed, reduces Ti (IV) cations to the Ti (III) state and the holes oxidise O 2 − * anions 76 . Simultaneously O 2 atoms are propelled out producing a set of O 2 vacancies on the surface. These vacancies were filled by water molecules present in the environment and adsorbed OH groups are formed on the surface, increasing the hydrophilicity of the surface 77 . Also, the radicals hence produced can degrade the dye molecules present around the membrane surface (detail mechanism explained in the mechanism www.nature.com/scientificreports www.nature.com/scientificreports/ section). For the reproducibility, a test procedure was repeated with five membrane pieces for each sample. MB dye showed better contrast while taking digital images and hence was preferred for photographic images.
When irradiated with UV light, the HNTs do not get excited, rather it acts as an electrical insulator and hence any charge generated on TiO 2 surface during UV irradiation cannot be transferred to HNTs. These electrostatic attraction and repulsion forces contribute together for an efficient movement and separation of e − and h + on TiO 2 1 . Apart from being a charge carrier separator, HNTs also enhance dye degradation. Due to the negatively charged surface of HNTs, the dye molecule (cationic in nature) are brought closer to the TiO 2 , increasing the  www.nature.com/scientificreports www.nature.com/scientificreports/ adsorption rate of the dye molecules. The detailed mechanism of the photocatalyst TiO 2 @HNTs and photocatalyst membrane are shown in Fig. 13. When TiO 2 @HNTs is irradiated by UV light, a photoelectron moves from valence band of TiO 2 @HNTs to the empty conduction band. This photon has the energy (hυ) equal to or greater than the band gap. Thus a hole is created in VB ( + h VB ) and an electron (e CB − ) in CB generated as shown in Fig. 13(a). These (h VB + ) then produce OH* radical after reacting with H 2 O. Now, these OH* radicals act as a potent oxidising agent and oxidise adsorbed organic molecules which are in near vicinity of TiO 2 @HNTs surface Fig. 13b. Simultaneously O 2 atoms are propelled out producing a set of O 2 vacancies on the surface. These vacancies are filled by water molecules present in the environment and results in the formation of adsorbed OH groups on the surface, increasing surface hydrophilicity 77 . These photo-generated electrons hence formed, reduces Ti +4 @HNTs cations to the Ti +3 @HNTs state and the holes oxidise O 2 − * anions. Also, the radicals hence produced can degrade the dye molecules present around the membrane surface as shown in Fig. 13c. The efficiency of the breakdown of these organic molecules depends upon their stability and structure. OH* radicals also degrade the pollutant present around it. The electrons in ( − e CB ) conduction band meanwhile reacts with O 2 and generate superoxide radicals (O 2 − *), which accelerates oxidation process (Fig. 13d) and also hinders any further e−/hole recombination  . Hence finally hydroxyl radical (OH * ) is formed, which is highly reactive in nature.
To understand the role and the involvement of active species in degradation process, control experiments were performed using scavengers for the photo-generated holes and free radicals. Photo-degradation of MB and RB dyes were investigated in the presence of UV light with the TiO 2 @HNTs photocatalyst to observe the role and the importance of degradation by free radicals. Isopropanol (IPA) and Tert-butyl alcohol (TBA) were used as H + and hydroxyl radical (OH*) scavenger respectively. In both reactions with MB and RB dye, H + showed no significant effect on photocatalytic degradation, while OH* free radical affected the MB and RB photocatalytic degradation reactions. Adding TBA (0.02 mmol) in the reaction, photocatalytic degradation decreased from 87.47% to 44% in case of MB dye while in case of RB dye degradation percentage decreases from 96.87% to 72%. This proved that OH* free radicals were generated during photodegradation of dye. Several researchers 78 have suggested that the OH* radical produced by the oxidation of water or OH − radicals by holes at the surface, diffuses towards the solution to oxidise the organic compound. H 2 O 2 and hydroxylated degradation products were formed during the reaction and the efficiency of degradation increases significantly when H 2 O 2 is formed in the presence of UV radiation. This was because of free hydroxyl radicals (which act as powerful oxidizing agent) generated by the dissociation of H 2 O 2 in the presence of UV irradiations. Moreover, a high concentration of hydroxyl peroxides itself acts as a scavenger which reduces the concentration of hydroxyl radicals and compound elimination efficiency. The generated hydroxyl radicals attack the MB and RB dye structure at different sites like un-saturation points etc. In several such attacks, the MB and RB dyes get converted into CO 2 and hetero-atoms which are further mineralized as mentioned in Supplementary File Tables S2 and S3. The combination of TiO 2 @ HNTs photocatalyst has both the advantages of being an efficient charge carrier separator and good absorbent for positively charged molecules 1 . The photocatalytic activity was also enhanced due to the absorptivity of HNTs and the crystalline TiO 2 nanoparticles which facilitate the interaction of dye and reactive TiO 2 @HNTs photocatalyst. Also, the agglomeration of TiO 2 nanoparticles was avoided by the homogeneous dispersion of TiO 2 nanoparticles on the HNTs surface.
The experimental data confirm that after 120 min of UV irradiation, the dye solution gets degraded with the formation of intermediate and end products. A mass spectroscopic (MS) study of the dye solution was also done to determine the intermediate and end products which were formed due to the cleavage of aromatic rings during the dye degradation process and eluted out at different retention time as per their mass and suggested structures. The difference in the concentration and composition of the products lead to many peaks with different intensities. Mass spectra and the possible structures of the dye degradation products are listed in Supplementary Table S2 and  Table S3 for MB and RB dye [79][80][81] . The demethylation cleavage has also been reported in the literature during the photocatalytic degradation 79,80,[82][83][84] .
For evaluation of the activity of a photocatalyst, commonly time dependence of the concentration loss of dye under UV irradiation is measured. However, there are many factors that govern the reaction rate and kinetics. These experimental conditions include the concentration of the photocatalyst, the surface area of the photocatalyst, the amount of the photocatalyst used in the experiment and the UV light intensity and more. Table 2 summarises some recently synthesized photocatalyst with their photocatalytic experimental data.

Conclusion
Utilization of naturally present HNTs as photocatalyst support is advantageous for the synthesis of TiO 2 @HNTs photocatalyst nanoparticles due to its size and shape dependent photocatalytic properties. In this study, TiO 2 @ HNTs photocatalyst and photocatalytic PVC membranes are synthesized. The prepared photocatalyst is stable and exhibits enhanced photocatalytic activity for the degradation of MB and RB dye solution under UV irradiation. The photocatalytic PVC membrane also exhibits similar photocatalytic activity against MB and RB dye but the degradation is slower as compared to the TiO 2 @HNTs photocatalyst of the same weight.
Due to the electrostatic interaction between TiO 2 and HNTs surface, the photocatalyst has more e − and h + pairs resulting in high photocatalytic activity. In the case of MB and RB dye which are positively charged, HNTs improved the supply and stability of the photo-generated charges and enhanced the absorption capability of the dye molecule on the photocatalyst. This was because of electrostatic attractive and repulsion forces originating www.nature.com/scientificreports www.nature.com/scientificreports/ from the negatively charged HNTs surface. The stability of the TiO 2 @HNTs photocatalyst is non photo-corrosive nature during the three consecutive cycle of photocatalytic degradation of the model dye molecules, which is very important for its practical applications. The MB and RB degradation catalysed by a TiO 2 @HNTs and photocatalyst PVC membrane followed the first-order kinetic model. Therefore, the capabilities of nano range TiO 2 @ HNTs photocatalyst to degrade dyes may be exploited for wastewater purification in various textile and chemical industries.
TiO 2 @HNTs photocatalyst preparation. For the preparation of photocatalyst, moisture was removed from the inner/outer surfaces of raw HNTs by drying them for 4 h at 400 °C. 30 ml of silane coupling agent was mixed with 100 ml of toluene and 10 g of dried HNTs was then added into the silane-toluene solution to make it functional 53 . The mixture was then stirred at 125 °C for 18 h. After this, the mixture of functionalized HNTs was centrifuged and then washed with isopropanol (3-4 times). Vacuum drying chamber was further used for drying of the pellet at 60 °C. To synthesize TiO 2 nanoparticles on functionalized HNTs surface, 1 g of silane HNTs was mixed with titanium (IV) isopropoxide (TIP) -ethanol solution (ratio 1:15) by dispersing it into the deionised water (pH value adjusted by adding HNO 3 or NH 4 OH). This solution was then vigorously stirred for sol-gel preparation. Hydrolysis reaction was initiated when TIP solution interacted with water molecules making the solution turbid and resulting in the increase of temperature to 60-70 °C for 18-20 h. When the peptization process was complete, stirring was stopped and centrifugation was done to retrieve TiO 2 @HNTs mixture, which was then subjected to vacuum drying chamber overnight at 65 °C 85,86 . At last, the calcination of TiO 2 @HNTs was done by heating it in a muffle furnace for 2 h (400 °C) 44 .
Photocatalytic PVC membranes. The photocatalytic PVC membranes were prepared based on the principle of classical phase inversion method. The mixture containing TiO 2 @HNTs photocatalyst (2 wt.% and 3 wt.% by weight of PVC) and DMAc solvent (85 wt.% by weight of the solution) was stirred at 600 rpm for 1 h to get the photocatalyst dispersed properly in the solvent. For the formation of pores, PVP (1 wt.% by weight of the solution) and PVC (14 wt.% by weight of the solution) polymer were added into the mixture (shown in Table 3). For uniform dispersion, the mixture was vigorously stirred for 12 h. After stirring, a homogeneous casting solution thus obtained was then degassed (room temperature) and poured on a glass slide with the help of membrane applicator (thickness 150 µm). The glass plate was then dipped immediately in a pure water bath for 12 h (room temperature) for the proper phase inversion process. Two photocatalytic PVC membranes (Table S4) were prepared based on the weight% of TiO 2 @HNTs photocatalyst added in the membrane casting solution.
Thermo-gravimetric analysis (TGA). Samples were heated from room temperature to 600 °C at the rate of 10 °C min −1 under flowing nitrogen using a Diamond TG/DTA (Perkin Elmer, USA) instrument.
X-ray diffraction (XRD). HNTs and TiO 2 @HNTs Photocatalyst powder samples were put into the sample collector for X-ray diffraction analysis with PANalytical, The Netherlands, scan rate 2 degrees/min. XRD peaks were recorded in the reflection mode in the angular range of 10-80° with (2 theta) angle.
Transmission electron microscopy (TEM). The morphological characteristics of raw HNT and TiO 2 @HNTs were studied by a CM 200 transmission electron microscope (Philips). The samples were dispersed in deionized water, and then the suspended particles were transferred to a copper grid.
Photoluminescence spectra. At room temperature, the photoluminescence spectra were recorded with a Cary Eclipse fluorospectrometer using 220 and 450 nm Ar + laser as excitation source.
Band gap energy. For determining the absorption coefficient, optical energy gap (Eg) and nature of transitions involved, the optical absorbance spectra of TiO 2 @HNTs (at room temperature) were studied. The thickness of the quartz cuvette (t), the optical absorption coefficient (α) was determined from the measurement of wavelength (λ). Generally,the absorption coefficient (α) was related to photon energy (hν) by known equation 87,88 : where, β signifies a constant known as band tailing parameter, Eg: energy of the optical band gap and n: power of the transition.
To convert the absorption spectra, in place of α, the kubelka-Munk function was used to eliminate any tailing contribution from UV spectra. The following function was applied to convert the absorption spectra: where R, the reflectance E g values were estimated from plot of (F(R) hν) 2 versus energy by extrapolating the linear part.
Liquid chromatography-mass spectrometry (LC-MS). The chromatographic experiments with LCMS system were carried out on an Agilent 1290 Infinity UHPLC System, 1260 infinity Nano HPLC with Chipcube, 6550 iFunnel Q-TOFs (Agilent Technologies, USA) with a Column, binary pump and an autosampler. Acetonitrile was used as mobile phase solvent. The mass spectrometer was equipped with an electrospray ionization (ESI) source. The mass range was from 50 to 1000 m/z. Degradation products were monitored by LC-MS. Measurement conditions are listed in Supplementary Table S1.
Methods. Photo-catalytic reaction experiments. The possible photocatalytic activities of the TiO 2 @HNTs 2, TiO 2 @HNTs 3, TiO 2 @HNTs M2 and TiO 2 @HNTs M3 membranes were examined by assessing the MB and RB dye degradation which was prepared in an aqueous medium. UV batch reactor was used to carry out the photocatalytic reaction at a low pressure of 125 W UV lamp (254 nm) and Photon flux (Φ) = 1.69 × 10 20 s −1 m −2 with continuous stirring. The photoreactor was initially filled with 100 mL of a 20 mg/L aqueous dye solution along with different weight percentage of a photocatalyst for the process and also with a different weight percentage of photocatalytic PVC membranes. The different concentration and combination used in the reactorare mentioned below. A UV−visible spectrophotometer (HACH, DR 6000, USA) was used to record the absorption magnitude of the dye solution regularly and similar steps was performed for both the dyes, after regular time intervals, at wavelengths of 664 nm for MB and 562 nm for RB dye respectively. Similarly, the photocatalytic efficiency of photocatalytic PVC membranes (TiO 2 @HNTs M2 and TiO 2 @HNTs M3) were also assessed. Initially, the membranes were dried and cut into small pieces and then dispersed into the aqueous dye solution for 1 h under constant stirring, after which dye solution irradiated with UV light, the concentration was analysed by a UV−visible spectrophotometer at regular intervals.
For comparative analysis, two different weight percentage of TiO 2 @HNTs photocatalyst (accordingly photocatalyst added in membrane casting solution) and photocatalytic PVC membranes were used for this study under following conditions -(i) MB and RB dye solution exposed to UV light in the absence of HNTs, TiO 2 @HNTs photocatalyst and photocatalytic PVC membrane (adsorption of dyes). (ii) MB and RB dye solution irradiated with UV light with only HNTs. (denoted as HNTs 2 and HNTs 3 i.e 2 wt.% and 3 wt.%) (iii) MB and RB dye solution were irradiated with UV light with only TiO 2 @HNTs photocatalyst (denoted as TiO 2 @HNTs 2 and TiO 2 @HNTs 3, a similar weight of photocatalyst added in membrane casting solution).
In addition, we have also taken a higher concentration of photocatalyst in the reaction. (iv) MB and RB dye solution irradiated with UV light with photocatalytic PVC membranes (denoted as TiO 2 @ HNTs M2 and TiO 2 @HNTs M3 i.e 2 wt.% and 3 wt.% photocatalyst added in membrane casting solution as mentioned in membrane preparation section) (v) Scavenger effects on the MB and RB dye degradation reaction were also tested by using TBA and IPA (with 0.2 mmol solutions) as hydroxyl radical (OH * ) and H + scavengers respectively. where, q e (mg/g): the amount of dye adsorbed, C i and C f (mg/L): the concentrations of dye at initial and equilibrium respectively, V (L): the volume of the solution and M (g): the mass of dry TiO 2 @HNTs photocatalyst used.
Photocatalytic PVC membrane activity test with MB dye. Photocatalytic PVC membrane TiO 2 @HNTs M2 and TiO 2 @HNTs M3 pieces (1 cm 2 per piece) with control membrane were dipped into 20 ml aqueous dye solution of MB dye and kept in dark for1 h. Membrane pieces were then positioned in open petri dishes individually and kept in UV light (Philips 15 W UV light). Digital images of these membranes were captured after keeping them on petri plates in dark and after exposure to UV light.