Introduction

Metal oxides are useful for industrial applications such as gas sensors, batteries, and artificial photosynthesis1,2,3,4,5,6,7,8,9,10,11,12. For example, SnO2 micropatterns were fabricated on a polymer film1 and a silicon substrate2 using a patterned octadecyltrimethoxysilane self-assembled monolayer template. The sensitivity of a hydrogen gas sensor with an SnO2 micropattern increased linearly with increasing SnO2 crystallinity. SnO2 could also be used to fabricate a lightweight flexible gas sensor on a polymer film. A porous iron vanadate (FeVO4) thick film consisting of disordered nanorods was fabricated by a hydrothermal method for gas penetration and permeation12. The probed I–V behavior and ultraviolet–visible spectroscopy measurements confirmed the semiconducting nature of triclinic FeVO4 (Eg = 2.72 eV) and revealed that the activation energy for electric conduction was 0.46 eV. The best gas sensitivity of 0.29 ± 0.01 (m =  − 3.4 ± 0.1) was obtained at optimal working temperature of 250 °C. SnO2 nanowires were fabricated for an H2S gas sensor3. They were functionalized with copper particles during chemical vapor deposition. A CuO@SnO2 p–n heterojunction was fabricated by oxidizing Cu to CuO. The fabricated sensor demonstrated high sensitivity and selectivity for H2S gas. VO2 nanobelts or nanoparticles were fabricated by a hydrothermal method for thermochromic devices4. Nanoparticles showed high phase transition enthalpy (ΔH = 32.4 J/g) and VO2-PET composite films showed optical switching characteristics (Tlum = 33.5%, ΔTsol = 16.0%). A dye-sensitized solar cell was prepared with a ZnO@TiO2 core shell nanorod array via a low-temperature solution method5. High-aspect-ratio ZnO nanorod arrays were covered with a TiO2 shell. The TiO2 shell increased the short circuit current (from 4.2 to 5.2 mA/cm2), open circuit voltage (from 0.6 to 0.8 V), fill factor (from 42.8 to 73.02%), and overall cell efficiency (from 1.1 to 3.03%). ZnO nanorods were grown on a highly conductive sandwich-like seed layer (ZnO seed layer/Ag nanowires/ZnO seed layer) and modified with α-Fe2O3 nanoparticles6. ZnO nanorods showed high potential for Ca2+ sensing in water or serum samples. They can be applied to drinking and irrigation water as well as to clinical analysis. A sensitive and selective sunlight-driven photoelectrochemical sensor was developed for the direct detection and reduction of chromium(VI)11. It was based on single-crystal rutile TiO2 nanorods decorated with gold nanoparticles. It showed high sensitivity under solar simulator illumination. Evaluations of real water samples showed excellent anti-interference and recovery capabilities. Metal oxide nanomaterials and a biosensor fabricated using them were reviewed7. Nanomaterial deposition on conductive electrodes is a crucial step for achieving high performance, and a simple, stable, and reproducible method is considered essential for depositing nanomaterials for fabricating a biosensor.

Many devices have reaction sites on the metal oxide surface, and therefore, they need to have a large surface area. In addition, studies are developing new devices by using the characteristic surface structure of metal oxides. Controlling the size, shape, and crystallinity of the metal oxide is known to greatly affect the device characteristics, and new metal oxide materials are being actively developed. There is also a strong need to control the shape and even the exposed crystal plane after the metal oxide material is crystallized. Especially, nanostructured TiO2 films with high surface area and high photocatalytic activity are strongly required for photocatalysts and related devices. Performance of the nanostructured TiO2 films is strongly affected by the size, shape, crystallinity, and the exposed crystal plane. The development of high performance nanostructured TiO2 films is a powerful demonstration of precise control of the size, shape, crystallinity, and the exposed crystal plane.

Metal oxide nano-/microstructures have also been synthesized in animals and plants. The size, shape, crystallinity, and surface structure of metal oxides in aqueous solutions were controlled at room temperature and atmospheric pressure. They are called “Bio-mineralization” and have created a new academic field, “Bio-inspired mineralization”. In the bio-inspired mineralization, we learn from nature and aim to develop novel materials that are not in nature.

This study investigates the bio-inspired mineralization of nanostructured TiO2. Nanostructured TiO2 was formed on polymer or inorganic films in aqueous solutions. Further, its surface area and photocatalytic properties were investigated.

Experimental

Ammonium hexafluorotitanate ([NH4]2TiF6) (Morita Chemical Industries Co., Ltd., FW: 15, purity 96.0%) and boric acid (H3BO3) (Kishida Chemical Co., Ltd., FW: 61, purity 99.5%) were used as received. Nanostructured TiO2 films were crystallized on polyethylene terephthalate (PET) films or fluorine-doped tin oxide (FTO) films in an aqueous solution containing ammonium hexafluorotitanate (50 mM) and boric acid (150 mM) at 50 °C for 24 h.

The morphology of the nanostructured TiO2 film on the PET film was observed using a field-emission scanning electron microscope (FE-SEM; JSM-6335F, JEOL Ltd.) at 20 kV. Nitrogen adsorption–desorption isotherms were obtained using Autosorb-1 (Quantachrome Instruments). Nanostructured TiO2 films on PET films were outgassed at 110 °C under 10–2 mmHg for more than 6 h prior to measurement. The specific surface area was calculated by the Brunauer–Emmett–Teller (BET) method using adsorption isotherms. The pore size distribution was calculated by the Barrett–Joyner–Halenda (BJH) method using desorption isotherms. The photocatalytic properties of the nanostructured TiO2 on FTO films were evaluated at the Kanagawa Academy of Science and Technology (KAST), Japan, based on Japanese Industrial Standards (JIS).

Results and discussion

Bio-inspired mineralization of nanostructured TiO2

Ten sheets of PET films (50 mm long × 50 mm wide × 0.1 mm thick) were pasted on glass plates with polytetrafluoroethylene tapes. Ammonium hexafluorotitanate (10.31 g) and boric acid (9.321 g) were dissolved in 1,000 mL of distilled hot water13. The concentrations of ammonium hexafluorotitanate and boric acid were 50 and 150 mM, respectively. PET films were exposed to vacuum-ultraviolet light for 20 min in air using a low-pressure mercury lamp (PL16-110, SEN Lights Co.). They were immersed in the solutions at 50 °C for 24 h. The titanium oxide-formed surface of the PET films faced obliquely downward. Therefore, even if homogeneously nucleated titanium oxide particles formed in the solution settle, they are less likely to be deposited on the PET film. Further, the PET film is less likely to be bent or deformed in the solution because it is attached to the glass plate. Moreover, titanium oxide is formed only on the PET film surface because the back surface of the film is in close contact with the glass plate. Thereafter, the glass plates were removed from the solution. The PET films were peeled off from the glass plates, washed with running water, and dried under a strong air flow.

Nanostructured TiO2 with high surface area

PET films were uniformly covered with nanostructured TiO2 (Fig. 1a). The coatings had an uneven surface structure (Fig. 1b). The needle-like surface structures had size of ~ 5–10 nm (Fig. 1c). Each needle had nanosized surface asperities.

Figure 1
figure 1

(a) FE-SEM micrograph of nanostructured TiO2 on a PET film. The nanostructured TiO2 was formed in an aqueous solution using liquid-phase crystal growth. (b), (c) Magnified FE-SEM micrographs of (a).

Full size image

The mass of the PET film was measured before and after immersion in the aqueous solution. The mass of nanostructured TiO2 was calculated from the difference in mass before and after immersion. The composite was cut to obtain rectangular pieces with dimensions of ~ 3 mm × 10 mm. All of these pieces were filled in a glass sample holder for gas adsorption measurements. The gas adsorption amount of the composite can be measured with almost no influence of the PET film, and the measured gas adsorption amount of the composite can be regarded as the gas adsorption amount of nanostructured TiO2.

Nanostructured TiO2 showed nitrogen adsorption–desorption isotherms (Fig. 2a). The pore size distribution of nanostructured TiO2 was calculated from the nitrogen desorption isotherm by the BJH method (Fig. 2b). Nanostructured TiO2 had inside pores and/or interparticle spaces of ~ 2–100 nm. The result did not indicate whether nanostructured TiO2 has pores of 1 nm or less because such pores cannot be estimated by the BJH method. The BET specific surface area was calculated to be 503.6 m2/g (Fig. 2c). The total surface area of nanostructured TiO2 was calculated by multiplying this value by the mass of nanostructured TiO2. The ratio of the surface area to the substrate projected area was calculated to around 284 times that of a bare PET film by dividing the total surface area of nanostructured TiO2 by the projected area of substrates (25,000 mm2, 50 mm × 50 mm × 10 sheets). The total surface area is not affected by the error of the weight of the nanostructured TiO2 in this calculation method, and therefore, it can be determined accurately. To the best of the authors’ knowledge, the TiO2 film surface area was the highest ever reported. Nanostructured TiO2 in this study was considered to be different from general nanostructured TiO2 reported elsewhere. Micro-, nano-, and subnanosized surface roughness and inside pores contributed to increased nitrogen adsorption. Surface crystal defects such as kinks, terraces, and secondary nuclei were also considered to contribute to increased nitrogen adsorption.

Figure 2
figure 2

(a) N2 adsorption–desorption isotherm of nanostructured TiO2 on a PET film. (b) Pore size distribution calculated from N2 adsorption data of nanostructured TiO2 on a PET film using BJH equation. (c) BET surface area of nanostructured TiO2 on a PET film. Inset: ratio of surface area to substrate projected area.

Full size image

Nanostructured TiO2 with high photocatalytic activity

FTO films were exposed to vacuum-ultraviolet light for 20 min. They were immersed in the solutions containing ammonium hexafluorotitanate (50 mM) and boric acid (150 mM) at 50 °C for 24 h. The photocatalytic properties of the nanostructured TiO2 on FTO films were evaluated based on JIS R 1701-2: 2008 (acetaldehyde removal performance) and JIS R 1701-1: 2010 (nitrogen oxide (NOx) removal performance). Acetaldehyde and NOx serve as evaluation indices for indoor and outdoor photocatalytic activity, respectively. Materials with high photocatalytic activity for acetaldehyde and NOx can be used in building interiors and exteriors, respectively. The photocatalytic characteristics were evaluated using two FTO films (50 mm × 26 mm × 1.1 mm in thickness) (AGC Fabritech Co., Ltd., TOC substrate (DU film)). The FTO layer (0.1 mm in thickness) was formed on a glass substrate (0.1 mm in thickness). The size of the test piece was around half of the JIS-specified size (50 mm × 100 mm). The measured values excluding regeneration efficiency for NOx removal characteristics were thus doubled.

In general, light irradiation on a photocatalytic material such as TiO2 generates electrons and holes on the surface (Fig. 3). Oxygen and water in the air react with electrons and holes, respectively. These reactions produce hydroxy (OH) radicals and superoxide ions on the titanium dioxide surface. OH radicals have strong oxidizing power and remove electrons from acetaldehyde molecules. Acetaldehyde molecules lose electrons and break bonds. They were converted to CO2 and/or H2O that were released to the atmosphere.

Figure 3
figure 3

Schematic image of photocatalytic activity.

Full size image

Acetaldehyde was removed from the sample at the rate of 2.80 μmol/h, and its removal ratio was 20.6% (Table 1). This rate was similar to that of commercial products certified by the Photocatalysis Industry Association of Japan14 (Fig. 4). Further, acetaldehyde was converted to CO2 at the rate of 4.56 μmol/h, and its conversion ratio was 16.8%.

Table 1 Photocatalytic properties of nanostructured TiO2.
Full size table
Figure 4
figure 4

Comparison of acetaldehyde decomposition characteristics with those of commercial products certified by the Photocatalysis Industry Association of Japan.

Full size image

The amount of NOx removed was 1.04 μmol (Table 1). This value was similar to that of commercial products (Fig. 5). The NOx adsorption and desorption amounts were 0.08 and 0.8 μmol, respectively. The amount of nitric oxide removed was 3.0 μmol. The amount of nitrogen dioxide generated was 1.24 μmol.

Figure 5
figure 5

Comparison of NOx decomposition characteristics with those of commercial products certified by the Photocatalysis Industry Association of Japan.

Full size image

The regeneration efficiency upon washing with water was 1,400% (no conversion). The first and second elution amounts of NOx were 12.50 and 2.42 μmol, respectively. The regeneration efficiency exceeded 100%, suggesting that NOx was generated from the sample during the test and was added to the NOx elution amount. This result indicated that nanostructured TiO2 contained nitrogen inside and/or on its surface. Ammonium hexafluorotitanate is one of the sources of nitrogen.

Annealed nanostructured TiO2 with high photocatalytic activity

Nanostructured TiO2 on FTO films were annealed at 500 °C in air for 4 h. In general, high-temperature treatment was considered to improve crystallinity but reduces internal vacancies and surface defects. Photocatalytic properties are affected by both changes.

Notably, the acetaldehyde removal rate increased greatly from 2.80 to 10.16 μmol/h upon annealing, and the removal ratio was 74.8% (Table 1). The rate exceeded that of commercial products (Fig. 4). This was one of the advantages of nanostructured TiO2. Further, acetaldehyde was converted to CO2 at the rate of 16.84 μmol/h, and its conversion ratio was 60.2%.

The NOx removal rate decreased from 1.04 to 0.42 μmol/h upon annealing (Table 1, Fig. 5). NOx removal was affected by the photocatalyst’s surface area rather than its crystallinity. The NOx adsorption and desorption amounts were 0.08 and 0.7 μmol, respectively. The amount of nitric oxide removed was 2.7 μmol. The amount of NOx generated was 1.68 μmol. The regeneration efficiency upon washing with water was 2,100% (no conversion). The first and second elution amounts of NOx were 7.34 and 1.28 μmol, respectively.

Conclusions

Nanostructured TiO2 was formed on PET films in aqueous solutions. Nanostructured TiO2 coatings showed high N2 adsorption. The BET specific surface area was ~ 503.6 m2/g and the ratio of the surface area to the substrate projected area reached around 284 times that of a bare PET film. Furthermore, the photocatalytic property of nanostructured TiO2 on FTO films was evaluated. The acetaldehyde removal rate increased greatly from 2.80 to 10.16 μmol h upon annealing. This value exceeded that of commercial products (Fig. 4). The amount of NOx removed was 1.04 μmol; this was similar to that of commercial products. This value decreased to 0.42 μmol/h upon annealing. NOx removal was affected by the photocatalyst’s surface area rather than its crystallinity. The study results revealed that nanostructured TiO2 had different nanostructure and surface conditions. These contributed to the high surface area and high photocatalytic activity. In particular, the acetaldehyde decomposition characteristics were higher than those of commercial products. Nanostructured TiO2 can be used for low-cost, large-area coating of various substrates such as polymer films. The high surface area and high photocatalytic activity can be used for various applications to building interior materials.