Novel TiO2 Photocatalyst Using Nonaqueous Solvent- Controlled Sol-Gel Route

2.1. General overview

The sol-gel route [41, 42] involves the mixing of metal precursor into either water or organic solvent followed by formation of 3-dimensional network resulting in viscous gel, which in general results in amorphous powder after drying process. The process of mixing is known as hydrolysis, while formation of 3-dimensional network is called as condensation. These two processes are further controlled by many parameters including nature of metal precursor, ratio of precursor to solvent, nature of solvent, capping agents (surfactants), pH and temperature. Figure 1 summarizes the various types of sol-gel routes.

Aqueous sol-gel route [41], as its name suggests, uses water as solvent to dissolve metal precursor and to complete hydrolysis process. The hydrolysis process is extremely fast due to high reactivity of water with precursors and therefore, generally there is little control over morphology and reproducibility. Nonaqueous sol-gel routes offer a good alternative to get rid of these difficulties [43, 44]. An organic solvent (alcohols, ketones, aldehydes or ethers) is used to complete the hydrolysis process instead of water. Besides, the oxygen required for metal-oxide formation is supplied by organic solvent in nonaqueous sol-gel route, whereas water plays the role of oxygen donor in aqueous sol-gel synthesis. However, inclusion or exclusion of some surfactant (consisting of hydrophilic and hydrophobic groups) in reaction solution further classifies the nonaqueous sol-gel route into surfactant assisted and solvent controlled (surfactant free) routes respectively. The main advantage of surfactant assisted nonaqueous sol-gel route is that the surfactant acts as capping agent and results in highly mono-dispersed nanoparticles. In addition a good control over particle size, morphology with outstanding reproducibility is direct consequence of surfactant-assisted nonaqueous sol-gel route. Moreover, the surface properties of nanoparticles can be easily tailored by exchanging surfactants with other functional groups. However this method is also prone to certain limitations like impurities in nanoparticles and toxic effects due to surfactants. These limitations impose restrictions on the surface sensitive applications (photocatalytic, biomedical and sensing) of nanoparticles.

2.2. Solvent controlled nonaqueous sol-gel route

A good alternative to surfactant assisted nonaqueous sol-gel route [45, 46] is solvent controlled nonaqueous sol-gel route. The solvent in itself plays role of oxygen donor necessary for oxide formation and stabilizing agent to control shape, size and morphology of nanoparticles. This modified sol-gel route also facilitates highly pure nanoparticles completely free from toxic surfactants but suffers from tendency to agglomerate. However toxic effect of halides cannot be neglected when metal oxide nanoparticles are prepared by reacting metal halides with organic solvents. In order to avoid halide impurities from metal oxide nanoparticles it is wise to use metal alkoxides, acetates or acetylacetonates as metal precursor. In the following subsections, synthesis of pristine and metal doped TiO₂ nanoparticles using halide free, nonaqueous, solvent controlled sol-gel route [2, 10] is described.

3.1. Phase and structural characterization

X-ray diffraction (XRD) is an important tool used to determine the phase purity of sample, crystal structure, lattice parameter, average crystallite size and quantitative phase analysis. Generally, as prepared TiO₂ by sol-gel route is amorphous and therefore requires different heat treatment to form crystalline phases. Low (300–500°C), moderate (500–700°C) and high (more than 700°C) calcination temperature results in pure anatase phase, mixture of anatase and rutile, and pure rutile phase respectively [66]. The anatase to rutile (A-R) phase transformation is largely affected by type and amount of metal doping. Choi et al. [67] studied effect of single metal ion doping on A-R phase transformation temperature in TiO₂ prepared by sol-gel route. They observed that many metals such as Pt, Cr, V, Fe, La doping in TiO₂ lowers the A-R transformation temperature, whereas Ru metal shows opposite behavior. Xie et al. [65] and Singh et al. [2] also reported increased A-R transformation temperature in Na doped TiO₂ nanopowder. The presence of rutile structure in anatase phase affects the photocatalytic activity of metal doped TiO₂.

Figure 2(A) and Figure 2(B) shows the XRD pattern of Zr doped TiO₂ and Na doped TiO₂ with pristine TiO₂ respectively. Very fine powders of TiO₂ based photocatalysts prepared by nonaqueous solvent controlled sol-gel route [2, 10] are used to perform XRD using Cu Kα radiation (0.154 nm). Clearly, pure anatase phase of TiO₂ is formed for pristine as well as in Zr and Na doped TiO₂ nanopowder and matches with JCPDS card number 841286. In general, there are two types of doping (i) substitutional, and (ii) interstitial. Which one of these two types of doping is favored, depends on size of guest ion as compared to host ion and volume size of interstitial position in the host lattice. Substitutional doping is preferred in case where size of guest ion is comparable or slightly larger than lattice ion, whereas if size of guest ion is much smaller than lattice ion then it occupies interstitial position of host lattice.

XRD peaks of crystal planes of Zr doped TiO₂ (Figure 2(A)) shows red shift in 2θ values, which confirms the substitutional doping of Zr⁴⁺ at Ti⁴⁺ site [10]. The substitutional doping of Zr in TiO₂ results in an increase in cell parameter and cell volume as reported by Yu et al. [68] and Wang et al. [69] Thus, incorporation of large sized Zr results in lattice strain and hence leads to formation of oxygen vacancies and suppresses the grain growth. Similar behavior is observed for Na doped TiO₂ nanopowder (Figure 2(B)) [2]. However formation of oxygen vacancies in Na doped TiO₂ can occur not only due to strain induced by Na doping but also due to lower valence state of Na⁺¹ ion. It should be noted that there are contradictory literature reports regarding doping of Na in TiO₂. Xie et al. [65] who used aqueous sol-gel synthesis, reported that large size Na cannot substitute Ti and therefore migrates to TiO₂ surface forming Na-O bonds as there is no peak shift is observed in XRD patterns. Thus it appears that substitutional doping of Na at Ti in TiO₂, as indicated by the XRD peak shifts in Figure 2(B), is facilitated by nonaqueous solvent controlled sol-gel route, which was used by Singh et al. [10].

The crystallite size of pristine and doped TiO₂ nanopowder are calculated by well-known Debye Scherrer formula and tabulated in Table 1. In both, Zr doped TiO₂ and Na doped TiO₂, the crystallite size reduces for certain dopant concentration and hence increases the surface area of nanoparticles. Additionally, A-R phase transformation temperature is increased due to doping Zr⁴⁺ and Na⁺¹ in TiO₂ matrix [10, 65, 68, 69].

3.2. Microstructural characterization

XRD gives structural information like lattice constant and crystalline phases averaged over bulk of the material. TEM on the other hand is able to give microstructural information on a nano meter length scale. In addition, grain boundaries, dislocations and structural defects could be easily identified by TEM. In this technique, a well-focused electron beam impinges on an ultrathin specimen in a high vacuum column, with the help of electromagnetic lenses. The impinged electron beam interacts with specimen and gets transmitted. A controlled and sophisticated system of electromagnetic lenses is used to focus the transmitted electron beam on a fluorescent screen. Selected area electron diffraction, diffraction contrast imaging, high resolution imaging and energy dispersive X-ray spectroscopy are some of the most commonly used TEM techniques for the characterization of materials.

Figure 3 shows the bright field TEM images, selective area electron diffraction (SAED) patterns and HRTEM images of pristine and metal doped TiO₂ nanopowder prepared by nonaqueous solvent controlled sol-gel route. Clearly, the particle size reduces with metal doping in comparison to pristine TiO₂, which is in agreement with XRD results. The SAED ring pattern and HRTEM images confirm the crystalline nature of pristine as well as metal doped TiO₂ nanoparticles. The SAED diffraction rings could be indexed as (101), (004) and (200) lattice planes of pristine anatase TiO₂. The increase in lattice spacing with Zr and Na doping indicates large metal ion substitution at Ti site in agreement with XRD results.

3.3. Photoluminescence study

Photoluminescence (PL) spectroscopy is a versatile and powerful optical method to investigate the energy levels in materials. The material is irradiated with light photons of energy greater than or equal to band gap energy of material. This results in the excitation of electrons from valence band to the excited states (Figure 4(A)) of material. These electrons relax to conduction band by losing excess energy via nonradiative process. When these electrons return to their valence band, the energy is released in the form of photons and the process is known as photoluminescence. The energy of emitted photons is determined by the difference in atomic energy levels, while the intensity of emitted light gives us information regarding recombination of electrons and holes.

The energy and intensity of emitted light in PL spectra of particular material is highly affected by doping. Figure 4(B) shows the PL spectra of pristine and Na doped TiO₂ nanopowder at excitation wavelength 390 nm. The shape of PL signal remains same while intensity reduces with increase in doping concentration of Na in TiO₂ matrix. This is attributed to the fact that Na doping in TiO₂ matrix results in formation of oxygen vacancies resulting in energy states lying between valence band and conduction band that trap the electrons form conduction band and thereby reduce electron-hole recombination [68]. However, the PL signal intensity increases for Ti_0.9Na_0.1O₂ sample. This is attributed to fact that excessive formation of oxygen vacancies might act as recombination centers [70].

4.1. Degradation experiment

The photocatalytic activity of pristine and metal doped TiO₂ is evaluated by degrading aqueous solution of MB dye under ultraviolet (UV) irradiation. In order to compare photocatalytic activity of photo catalyst prepared by nonaqueous solvent controlled sol-gel route with commercial photo catalyst, the photocatalytic activity of well-known commercially available Degussa P25 TiO₂ is also evaluated. The initial concentration of MB aqueous solution is 5 mg/L. About 100 ml of MB dye aqueous solution (pH 5) is taken in 100 ml borosil glass beaker and 60 mg of photo catalyst is added in the solution. The beaker is kept on magnetic stirrer for uniform suspension of photo catalyst in the solution throughout the experiment. The reaction mixture is irradiated by UV light of peak wavelength at 365 nm and the intensity of UV light at the surface of reaction mixture is 10 mW/cm². The distance between reaction mixture and UV light source is 20 cm resulting in light intensity of 8 × 10⁴ lux over reaction mixture. In order to complete adsorption-desorption equilibrium between photo catalyst and dye, the reaction mixture is stirred for 20 minutes in completely dark chamber. After achieving adsorption-desorption equilibrium, the reaction mixture is irradiated with UV light and a small amount of solution is withdrawn at regular time intervals. The withdrawn sample is centrifuged to separate out nanoparticles from the solution and absorbance of supernatant measured using UV-Visible spectrometer. The photocatalytic degradation percentage of dye for different time intervals is plotted for several photo catalysts and without catalyst (WC).

The degradation percentage of dye is calculated using Eq. 1. The rate constant (k) for the photocatalytic degradation of dye is determined from pseudo first order law using Eq. 2.

where C₀ is dye concentration before UV irradiation and C_t is dye concentration after t time of UV irradiation.

In order to further confirm the degradation results, TOC of irradiated and nonirradiated dye aqueous solution is measured and TOC removal rate percentage is calculated using Eq. 3.

where TOC₀ and TOC_t are the TOC values of dye solution before and after time ‘t’ of UV irradiation respectively.

4.2. Degradation results and mechanism

The photocatalytic activity of prepared Na doped TiO₂ photo catalyst can be compared from Figure 5. Clearly, the degradation percentage of MB dye is highest for Ti_0.92Na_0.08O₂, even superior to commercially available Degussa P25 catalyst. The k values for degradation of MB dye by WC, P25, pristine TiO₂ and Ti_0.92Na_0.08O₂ are 0.86 × 10⁻³, 30.52 × 10⁻³, 3.02 × 10⁻³ and 43.24 × 10⁻³ min⁻¹, respectively. TOC results are in accordance as well. This enhancement in photocatalytic activity of TiO₂ with Na doping is attributed to its smaller crystallite size and reduced rate of electron-hole recombination. Similarly, Ti_0.9Zr_0.1O₂ shows enhanced photocatalytic activity [10] for which, in addition to reduction in crystallite size and electron-hole recombination, band gap was also reduced due to Zr doping.

The photocatalytic degradation of dye at the surface of TiO₂ is well explained in the existing literature [56, 72]. The reactions occurring at the surface of semiconductor TiO₂ under UV irradiation are depicted in the Eqs. 4–8. The electrons are excited to conduction band while holes are formed in conduction band after absorption of photons having energy greater than or equal to energy gap of semiconductor. In case of pristine TiO₂, most of these electrons recombine with holes; few of them react with adsorbed oxygen forming reactive oxygen active specie (ROS) O₂^¯. Similarly holes in the valence band reacts with water molecules and form ROS OH* radicals. These ROS actually reacts with dye molecules and degrade them into simple hydrocarbons H₂O and CO₂.

In pristine TiO₂ the higher recombination rate of electron-hole results in fewer number of ROS radicals and hence less photocatalytic activity. In addition, large energy gap and small surface area also limit the degradation efficiency. Metal doping in TiO₂ results in large surface area, band gap tuning toward visible range and reduced electron-hole recombination rate. This most general proposed mechanism for degradation of dye at surface of metal doped TiO₂ is depicted in Figure 6.