Abstract
TiO2 nanotubes were synthesized by alkaline hydrothermal treatment of TiO2 nanoparticles with a controlled proportion of anatase and rutile. Tailoring of TiO2 phases was achieved by adjusting the pH and type of acid used in the hydrolysis of titanium isopropoxide (first step in the sol-gel synthesis). The anatase proportion in the precursor nanoparticles was in the 3–100% range. Tube-like nanostructures were obtained with an anatase percentage of 18 or higher while flake-like shapes were obtained when rutile was dominant in the seed. After annealing at 400°C for 2 h, a fraction of nanotubes was conserved in all the samples but, depending on the anatase/rutile ratio in the starting material, spherical and rod-shaped structures were also observed. The photocatalytic activity of 1D nanostructures was evaluated by measuring the deactivation of E. coli in stirred water in the dark and under UV-A/B irradiation. Results show that in addition to the bactericidal activity of TiO2 under UV-A illumination, under dark conditions, the decrease in bacteria viability is ascribed to mechanical stress due to stirring.
1. Introduction
TiO2 nanomaterials are well-studied and commonly used photocatalysts for the degradation of organics, water splitting, and solar cells, among others [1–4]. In the last years, several approaches were explored to increase the photoefficiency of TiO2, with the modification of the particle morphology and dimensionality being one of the newest [5]. One-dimensional (1D) nanostructures such as nanotubes, nanorods, nanowires, and nanobelts have attracted great attention because of their unique properties that may be beneficial for photocatalysis: (i) enhanced light absorption due to the high length/diameter ratio, (ii) rapid and long-distance electron transport capability, (iii) large specific surface area, and (iv) ion exchange ability [6]. Hydrothermal treatment of TiO2 particles in alkaline solutions is one of the simplest and cheapest techniques to produce 1D-layered titanate structures. The hydrothermal synthesis of TiO2 nanotubes involves several steps where the structure of the TiO2 precursor changes completely.
Results obtained in our laboratories show that 1D TiO2 nanostructures display photocatalytic activity for dye degradation [7]. Although its performance as a photocatalyst is not as good as other industrially produced TiO2, this form of TiO2 can be easily recuperated from the solution.
Furthermore, since 1985 when Matsunaga et al. [8] published the first report of the photocatalytic biocide effects of TiO2 under metal halide lamp irradiation, there has been increasing interest in photocatalytic disinfection. Use of TiO2 nanoparticles in suspension is an efficient method for decontamination due to the large surface area of catalysts available to perform the reaction. It has, however, some drawbacks before its scaling at the industrial level; for example, the necessity of removing the catalyst from the solution after decontamination using filtration increases the cost and time of the cleaning process [9].
In this sense, considering the advantages that our 1D TiO2 nanostructures are more easily filterable than nanoparticles and can be easily removed from solutions—in addition to the fact that the efficiency of 1D TiO2 as bactericide under UV-A irradiation was only briefly explored—in this work, we assess the photocatalytic activity of 1D TiO2, obtained from nanoparticles with a controlled proportion of anatase and rutile made by the sol-gel method, for E. coli ATCC 25922 in water.
2. Materials and Methods
2.1. Materials
Titanium isopropoxide purity 98%, hydrochloric acid fuming 37%, nitric acid 65%, and pure sodium hydroxide pellets were purchased from Merck. All reagents were used as received.
2.2. Synthesis of TiO2 Nanostructures
TiO2 nanoparticles (TiO2 NPs) were synthesized by the sol-gel method (SG). Titanium isopropoxide was added drop by drop to vigorously stirred HNO3 or HCl solutions at pH 0.5, 0.8, and 1.0. Suspensions were heated at 70°C for 2 h, autocleaved in a stainless-steel chamber at 220°C for 12 h, washed by centrifugation, and dried at 60°C.
1D TiO2 nanostructures were synthesized by hydrothermal treatment of 1 g TiO2 NPs obtained by the sol-gel method in 40 mL of 10 M NaOH at 130°C for 24 h. After hydrothermal treatment, the obtained white powder was vacuum filtered, washed with HCl solution for ionic exchange, and then washed with distilled water until a neutral pH was reached. Finally, the samples were annealed at 400°C for 2 h to crystallize the material.
The obtained nanostructures were characterized by X-ray diffraction (XRD) in a Rigaku diffractometer using Cu Kα radiation (). The morphology was studied by field emission scanning electron microscopy (FE-SEM SUPRA 40, Carl Zeiss) and high-resolution transmission electron microscopy (HRTEM) using a JEOL JEM-2010F transmission electron microscope operating at 200 kV. TEM samples were prepared by dispersing a small amount of the sample in ethanol with the help of an ultrasonic bath. Small droplets of the freshly prepared dispersion were placed onto a copper grid covered with carbon to improve the conduction of the electrons.
2.3. Assessment of Photocatalytic Activity of TiO2 Nanoparticles and 1D TiO2 Nanostructures against Escherichia coli in Water
The photocatalytic activity for water disinfection was tested using E. coli ATCC 25922. Experiments were performed in a batch reactor, with illumination from above using an Ultra-Vitalux 300 W lamp (30 W/m2) and, under dark conditions, containing 100 mL aqueous solution with 107 CFU/mL bacteria. 1.0 mL aliquots were collected after 0, 20, 40, and 60 min irradiation. Aliquots were diluted 1 : 10 with sterile water to fit in the range 10–500 CFU/mL. 1.0 mL samples of the final dilutions were vacuum filtered through a sterile filter; this results in all bacteria present in the water being retained on the filter. Finally, the filters were placed onto a paper pad soaked in “membrane lauryl sulphate broth” (Oxoid MM0615), which feeds E. coli bacteria but inhibits the growth of any other bacteria. The bacterial concentration was determined by counting after 18 h incubation at 37°C.
TiO2 nanoparticle samples were codified with C or N (for samples made with HCl and HNO3, resp.) accompanied by 0.5, 0.8, and 1.0, depending on the pH used in the sol-gel synthesis. In a similar way, 1D TiO2 nanostructures were codified adding 1D to the nanoparticle code (e.g., N0.5 refers to TiO2 nanoparticles obtained with HNO3 in pH 0.5, and N0.5-1D refers to a one-dimensional TiO2 nanostructure obtained for N0.5) resulting in twelve samples.
2.4. Assessment of Stirring in Bacteria Viability
In order to evaluate the mechanical stirring effect in bacteria viability, E. coli ATCC 25922 were tested in the dark under stirring (100 rpm) and without stirring at room temperature (20°C). 1.0 mL aliquots were collected after 0, 20, 40, and 60 min stirring.
3. Results and Discussion
3.1. 1D TiO2 Nanostructures Prepared from Seeds Presenting Tailored TiO2 Crystalline Phases
Figure 1 shows FE-SEM images of TiO2 nanoparticles obtained by the sol-gel method (SG-TiO2 NPs) using HNO3 and HCl as catalysts in the acid hydrolysis reaction of titanium isopropoxide (, 0.8, and 1). At pH 1, regardless of the acid, the images show spherical nanoparticles with average diameters of about 15 nm and 13 nm, with HNO3 and HCl, respectively. Polyhedral structures (60–100 nm) were observed when the pH decreased to 0.8, and octahedral structures with edges of about 140 nm were obtained with HCl at pH 0.5.
XRD patterns (Figures 2(a) and 2(b)) show that the crystalline structures correspond mainly to anatase when acidic solutions with pH 1 were used with both catalysts. Both anatase and rutile were observed with acidic solutions at pH 0.8 and 0.5; a small amount of brookite was detected in most cases. The amount of rutile increased as the pH decreased, and it was the dominant phase when HCl at pH 0.5 was used. The small peak for brookite disappears in this case.
(a)
(b)
The average crystallite size for anatase and rutile (Danatase and Drutile, resp.), the anatase content, Ap, estimated with the Spurr-Myers equation [10] from the main diffraction peaks, and the pH of the acidic solution (HNO3 and HCl) are shown in Table 1.
Figure 3 shows the morphology, by FE-SEM, of the 1D nanostructures obtained after alkaline hydrothermal treatment of the sol-gel TiO2 nanoparticles presented in Figure 1 and Table 1. In the case of SG-TiO2 synthesized with HNO3, the particles displayed a tube-like shape, with an average diameter of 11 ± 1 nm in different anatase contents within 18% to 100%, respectively. On the other hand, using as precursor SG-TiO2 synthesized with HCl acid, flake-like particles were identified together with tube-like structures. The proportion of flake-shaped particles increased as the pH decreased.
After the annealing process at 400°C for 2 h, TEM images (Figure 4) show nanotube structures in all the samples; depending on the seed material, some spherical and rod-shaped structures were also present. It can be seen that a sintering-like process took place during the annealing and that, as a consequence, bundles of tube-like structures and cracked structures were produced.
Tube-like structures seemed to be best conserved when obtained from TiO2 nanoparticles with 56% of anatase, synthesized with HNO3. When seed material with lower anatase content (~18%) was employed, large and irregular particles measuring about 80 nm were accompanying the nanotube structures. These might be rutile seed aggregates that could not react in the hydrothermal treatment because of their large particle size. In contrast, needle-like shapes and nanotubes turning to nanorods were observed when anatase-rutile TiO2 nanoparticles synthesized with HCl were used as seed materials. It must be mentioned that, because of the lack of homogeneity in the samples, it is difficult to represent the final TiO2 structure in a single TEM image; the pictures shown represent the most typical structure in each sample.
The XRD analysis of the samples after hydrothermal treatment (Figure 5(a)) shows that the crystalline structure of the seed material changed and displayed peaks around , 24.5, 28.4, and 48.3°. These peaks represent the diffraction of sodium titanate with the chemical formula Na2TinO2n+1 (, 6, and 9). This is observed for samples where the anatase content in the seed material was higher than 55%. In other cases, rutile was also present as shown by the reflection peaks around , 36.1, 41.5, 54, and 56°, corresponding to the (110), (101), (111), (211), and (220) planes in agreement with JCPDS No. 21-1276. This confirms that part of the rutile seeds could remain unreacted after the hydrothermal treatment. After the acid treatment, the features corresponding to titanates almost disappeared, leaving those of the rutile TiO2 polymorph (not shown).
(a)
(b)
After the annealing process (Figure 5(b)), a mix of anatase and rutile was observed for samples whose seed had a rutile content larger than 60%. Only peaks corresponding to anatase TiO2 were observed for samples with anatase higher than 56% in seed. This suggests that when rutile was the dominant phase in the seed material, a portion of it remained unreacted, probably because of the large crystallite size of rutile (~28 nm), compared with the anatase crystallite size (~9 nm). The conditions of the hydrothermal treatment seem insufficient to carry out the dissolution-precipitation process that would be involved in the transformation of TiO2 to sodium titanate, followed by proton exchange to produce hydrogen titanate and, finally, crystallization to anatase after thermal treatment. On the other hand, it can be noted that the seed material obtained with HCl produced samples with the best crystallinity, as the X-ray reflections were well defined, compared to those obtained with HNO3.
3.2. Assessment of the Photocatalytic Activity of TiO2 Nanostructures for Escherichia coli in Water
Bacteria viability under the stirring process was determined by colony counting after 24 h of incubation. The results showed that the stirring process affects in 1, 3, and 5% (gradually for 20, 40, and 60 minutes, resp.). The assays without stirring were not performed because it was not possible to obtain a homogeneous bacteria distribution.
The effect of stirring in the presence of SG-TiO2 nanoparticles and their corresponding 1D TiO2 nanostructures against E. coli was evaluated in the dark and under UV-A/B irradiation. As shown in Figure 6, considering that the initial E. coli concentration was 1 × 107 CFU/mL, the presence of TiO2 nanoparticles and nanotubes under stirring conditions in the dark produced a diminution of bacteria viability of around two orders of magnitude (105 CFU/mL). It is ascribed to mechanical stress produced by the stirring process.
(a)
(b)
Also, as is reported in other studies [11], the photolysis is present in our experiments. It contributes to a decrease in bacteria viability at three orders of magnitude. The bactericidal activity of TiO2 nanostructures is similar to the photolysis in consequence; the catalyst plus irradiation can decrease bacteria viability until five orders of magnitude. However, no major difference was observed for the bactericidal effect of nanoparticles and one-dimensional TiO2 nanostructures. The latter have an important advantage since 1D TiO2 nanostructures can be easily removed from solutions.
4. Conclusions
In summary, TiO2 anatase 1D nanostructures, with different shapes such as tube- and rod-like shapes, were synthesized by hydrothermal treatment of seeds controlling the anatase-rutile proportion. The synthesized 1D TiO2 nanostructure was effectively used for photocatalytic abatement of E. coli in water. Although the 1D TiO2 nanostructures have a similar photocatalytic activity than the nanoparticles have, the use of one-dimensional TiO2 nanostructures has an important advantage since the 1D TiO2 nanostructure can be easily removed from solutions and could be reusable avoiding the necessity of use filtration that increases the cost and time of the cleaning process.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
Acknowledgments
This work was partially supported by the Innovate Peru Project C.133-PNICP-PIAP-2015 and the Concytec Project 223-2015-FONDECYT-DE. The authors are also grateful to the Microscopy Centre of FCEyN, Universidad de Buenos Aires, and Central Microscopy Laboratory of the Institute of Physics (UNAM). Roberto J. Candal is a member of CONICET.