Indium doping in sol-gel synthesis of In-Sm co-doped xIn-0.05%Sm-TiO₂ composite photocatalyst

2.1 Synthesis of xIn-0.05%Sm-TiO₂ photocatalyst

In-Sm co-doped TiO₂ and pure TiO₂ were synthesized through sol-gel route. A solution was composed of 0.9 ml distilled water and 4 ml anhydrous ethanol. Another solution was composed of 0.0025 g Sm(NO₃)₃, In(NO₃)₃, 2 ml tetrabutyl titanate, 8 ml anhydrous ethanol, and 0.1 ml concentrated hydrochloric acid. The first solution was slowly dropped into the second solution to form a transparent sol. Subsequently, the sol turned into a sticky gel after stirring for about 1 h. The gel stayed at ambient temperature for 12 h and was dried in the furnace at 80°C for another 12 h. The solid was ground into fine powders and transferred into an oven. The temperature of the oven increased from room temperature to 400°C at a heating rate of 5°C/min. The powders were calcinated for 3 h and then cooled to room temperature. The samples were ground again and marked as xIn-0.05%Sm-TiO₂ [x=n(In)/n(Ti)×100%]. Samarium doping content was fixed to 0.05%. Pure TiO₂ sample was also synthesized without any dopants.

2.2 Characterization of photocatalysts

The crystalline structures of the materials were determined by a D8 X-ray diffractometer, using monochromatized Cu Kα at λ=1.5416 Å. Surface morphology of the materials were taken by a QUANTA 250 SEM. Infrared and far infrared absorption spectra were recorded by a Frontier Fourier transform infrared and far infrared (FT-IR/FIR) spectrometer in the wavenumber between 50 cm⁻¹ and 4000 cm⁻¹. A LAMBDA 35 UV-Vis spectrometer was used to record UV-Vis diffuse reflectance spectra of the materials. Specific surface area and pore characters of the materials were measured by an F-sorb 3400 analyzer.

2.3 Decoloration of methyl orange

Decoloration of methyl orange (MO) was measured to study the activity of the materials. A 100 ml quartz beaker and a 20 W UV-light lamp irradiating at 253.7 nm were set up as the lap-scale reactor. Photocatalyst (15 mg) and 50 ml of 10 mg/l MO solution were used in each experiment. Prior to turning on the light, the suspension was ensured adsorption-desorption equilibrium after stirring in the dark for 60 min. Adsorption percent of MO on the photocatalyst was measured at this moment. The photocatalyst powders were removed from the solution through a millipore filter. The average irradiation intensity was 1300 μW/cm², as measured on the surface of the suspension using an actinometer. The irradiation time in the subsequent experiments was set to 30 min except for the prolonged time reaction. Absorbance of the solution was measured by a 721E spectrophotometer at the maximum absorption wavelength of MO, i.e., 466 nm. MO concentration was calculated according to Lambert-Beer theory.

3.1 Characterization results

Figure 1 shows XRD patterns of xIn-0.05%Sm-TiO₂ and pure TiO₂. The typical and pure anatase phase can be seen in the pattern of TiO₂. The diffraction angles slightly move to low angle after the doping of Sm and In, accompanied with broaden peak and shrinking intensity. Scherrer formula was used to calculate the crystallite size of anatase TiO₂ based on the (101) plane. The crystallite sizes are 26.9 and 11.5 nm in pure TiO₂ and 0.05%Sm-TiO₂. With further doping of samarium ions in 0.05%Sm-TiO₂, the crystallite sizes are 10.1, 9.2, 8.9, and 8.7 nm for the samples containing 1%, 3%, 5%, and 10% samarium, respectively. The addition of Sm and In leads to continuous shrinking crystallite size. Anatase TiO₂ crystal formation and growing are disturbed by the dopants.

Figure 1:

XRD patterns of xIn-0.05%Sm-TiO₂ and pure TiO₂.

Substances containing indium and samarium cannot be distinguished in the XRD patterns, although the indium doping content is as high as 10% in the 10%In-0.05%Sm-TiO₂. Similar result is reported in our previous work using aluminum and indium as dopants [22]. As can be seen from the figure, the diffraction angles of anatase TiO₂ slightly move to low angle, which can be explained by the replacement of Ti⁴⁺ in the lattice skeleton with Sm³⁺ and In³⁺. The radii of Sm³⁺ (108 pm) and In³⁺ (81 pm) are larger than that of Ti⁴⁺ (68 pm). The substitution of smaller Ti⁴⁺ ions by the larger Sm³⁺ and In³⁺ ions causes the expansion of the TiO₂ cell. Table 1 lists the lattice parameters of anatase TiO₂ in xIn-0.05%Sm-TiO₂ and TiO₂. The cell volume of anatase TiO₂ expands with the addition of both indium and samarium.

The surface morphologies of the sol-gel synthesized xIn-0.05%Sm-TiO₂ samples are shown in Figure 2. The samples are composed of irregular shaped particles in the size as large as several micrometers. Particles aggregation during synthesizing process is responsible for the large particles. Grinding is necessary before using the materials, while the particle size is not even after grinding. There is no noticeable difference in the surface morphologies of the samples containing different content of indium.

Figure 2:

SEM images of xIn-0.05%Sm-TiO₂ samples with different indium contents. (A) 0.05%Sm-TiO₂, (B) 1%In-0.05%Sm-TiO₂, (C) 3%In-0.05%Sm-TiO₂, and (D) 5%In-0.05%Sm-TiO₂.

The FT-IR and FT/far IR spectra of xIn-0.05%Sm-TiO₂ and TiO₂ are shown in Figure 3. The adsorbed hydroxyl groups are proven by the bending and stretching vibrations of O-H bond at 1640 cm⁻¹ and 3450 cm⁻¹ [23], [24]. Surface adsorbed hydroxyl is important for photocatalytic generation of holes after electron exciting. Although the materials are dried before IR examination to remove the adsorbed water molecules, there is still detectable hydroxyl group combining with TiO₂. The absorption of stretching vibration of C-O bond is also found at 1387 cm⁻¹ [25].

Figure 3:

(A) FT-IR and (B) FT/far IR spectra of xIn-0.05%Sm-TiO₂ samples and pure TiO₂.

It seems that the absorption intensity of O-H becomes a little stronger after doping more indium ions. Although it is hard to ascertain the reason for this phenomenon, the increase in O-H group is beneficial to produce hydroxyl radicals after irradiation. FT-far IR spectra are used to distinguish the bonding information of metal oxide. The broad adsorptions centered at 470 cm⁻¹ and 347 cm⁻¹ are attributed to bending vibration of Ti-O-Ti [26]. The doping of indium and samarium ions has no noticeable influence on the bonding characters of TiO₂. The major skeleton of anatase TiO₂ lattice is maintained in the doped samples.

UV-Vis diffuse reflectance spectra of TiO₂ and xIn-0.05%Sm-TiO₂ are presented in Figure 4. The spectra can be used to calculate the band gap of semiconductors through Tauc plot method [27]. The pure TiO₂ synthesized in this work has a band gap of 3.08 eV. The value is a little smaller than the common 3.2 eV band gap of anatase TiO₂. After doping of 0.05% Sm in TiO₂, the band gap of 0.05%Sm-TiO₂ slightly changes to 3.05 eV. The indium doped samples have comparatively strong absorption in the visible region. The band gaps of xIn-0.05%Sm-TiO₂ samples are 3.03, 2.93, 2.91, and 2.85 eV when indium doping contents are 1%, 3%, 5%, and 10%, respectively. A continuous red shift of the absorption edge occurs at increasing indium doping content. Lower conduction band level is usually formed in the doped TiO₂. Bandgap energy is reduced after doping. The doped ions can also act as low potential electron trap in the original semiconductor, which is beneficial to promote quantum efficiency and the response to visible illumination.

Figure 4:

UV-Vis diffuse reflectance spectra of TiO₂ and xIn-0.05%Sm-TiO₂.

Enhanced absorption of the incoming photons means the possibility of photon-electron conversion and the subsequent degradation of organic substances. The UV lamp used in this work can illuminate UV light at 253.7 nm, which has enough power to excite all the xIn-0.05%Sm-TiO₂ samples. Meanwhile, the lamp can emit in longer wavelength including visible light. That means the co-doped xIn-0.05%Sm-TiO₂ materials can absorb more incoming photons to generate electrons and holes. Materials with small bandgap energy may have strong activity on photocatalytic degradation.

Porous characters are examined for the xIn-0.05%Sm-TiO₂ samples, as shown in Figure 5 and Table 2. Figure 5 gives N₂ desorption isotherms of xIn-0.05%Sm-TiO₂. All of the desorption curves can be identified as mesoporous materials in the International Union of Pure and Applied Chemistry classification. An almost constant adsorbed N₂ volume at low relative pressure indicates the saturated adsorption of N₂ molecules on the surface of xIn-0.05%Sm-TiO₂. Capillary condensation of N₂ molecules in the pores of the materials leads to obvious increase of the adsorbed N₂ volume. The materials also contain certain number of macropores that can be proved by the abruptly increased N₂ volume at relative pressure over 0.95.

Figure 5:

N₂ desorption isotherms of xIn-0.05%Sm-TiO₂.

Brunauer-Emmett-Teller (BET) surface area, average pore size, and total pore volume of TiO₂ and xIn-0.05%Sm-TiO₂ are listed in Table 2. The undoped TiO₂ has the minimum specific surface area and the smallest pore volume. Porous characters of photocatalyst are very important in discussing the possible activities on both adsorption and photocatalytic degradation. The doping of samarium into bulk TiO₂ can apparently enhance BET surface area, average pore size, and total pore volume. Further doping of indium into 0.05%Sm-TiO₂ has complex influences. The sample doped with 0.05% Sm and 5% In has the maximum specific surface area. A slight shrinking in surface area is found in the 10%In-0.05%Sm-TiO₂ sample containing as much as 10% indium.

The average pore size is enlarged from 11.2 nm to 19.8 nm after doping 0.05% samarium in TiO₂. That means smaller pores in TiO₂ merge into larger ones after the addition of samarium. The additional doping of indium into 0.05%Sm-TiO₂ causes continuous pore size declining in xIn-0.05%Sm-TiO₂, accompanied with shrinking pore volume with increasing indium content. The doping of Sm and In leads to continuous shrinking in crystallite size of anatase TiO₂. It is responsible for the enhancement in BET surface area. The average pore size and total pore volume are reduced after doping more indium ions due to the refinement of crystallite size. Obviously, the number of pores in the materials decreases after doping excessive amount of indium. Although the pore size varies with the doping of samarium and indium, the pores are large enough for most organic pollutants. The difference in pore size may not have much effect on adsorption of methyl orange molecules. On the other hand, both of adsorption and photocatalytic degradation of methyl orange take place on the surface of the xIn-0.05%Sm-TiO₂ materials. In this case, the enlarged surface area as the result of doping may have positive effect on removal of the dye.

3.2 Photocatalytic activity

Methyl orange is used as the model pollutant to examine the activity of xIn-0.05%Sm-TiO₂. Either adsorption or degradation of the dye leads to decoloration of the solution. Figure 6 shows adsorption and photocatalytic activities of xIn-0.05%Sm-TiO₂ samples with the variation of indium doping contents. The adsorption of methyl orange on the surface of xIn-0.05%Sm-TiO₂ increases with rising indium content. The maximum adsorption percent is a little higher than 10% when indium content is as high as 10%. Normally, the adsorption of methyl orange depends on the surface properties of the materials. Adsorption amount is usually proportional to specific surface area. Meanwhile, surface polarity and bonding characters are also important to the adsorption capacity.

Figure 6:

Adsorption and photocatalytic activities of xIn-0.05%Sm-TiO₂ samples with respect to indium contents.

The influence of indium content on photocatalytic degradation of methyl orange on xIn-0.05%Sm-TiO₂ is more complex. Degradation efficiency increases from 41% on 0.05%Sm-TiO₂ to the maximum value of 69.8% on 5%In-0.05%Sm-TiO₂. The degradation efficiency sharply decreases after the optimal indium doping content. In a common sense, nano-sized material usually has strong activity. The resulted smaller crystallite size after doping of indium can be beneficial to photocatalytic activity of the materials. However, this might not be the main reason causing the variation of the activity.

The adsorption capacity and photocatalytic activity of the photocatalyst have close correlation. Adsorption capacity is related to specific surface area. Meanwhile, surface area of photocatalyst is also important for absorption of incoming irradiating photons. The adsorption of pollutant on photocatalyst surface is also beneficial to the subsequent transition of photogenerated electrons and holes to the target substances. As stated before, either the formation of lower conduction band level or the introduction of electron trap in TiO₂ is capable of promoting quantum efficiency and absorption of illumination at longer wavelength. The optimal doping content is usually reported by most of the researchers regarding doping effect on photocatalysts. After exceeding the optimal doping content, degradation activity can be eventually reduced. The dopants can become recombination centers for electrons and holes at excessive concentration if they cannot enter into the lattice skeleton of TiO₂.

Figure 7 compares the removal of methyl orange with extending irradiation time on 0.05%Sm-TiO₂ and 5%In-0.05%Sm-TiO₂. The photocatalysts and methyl orange solution were stirred in the dark to reach adsorption-desorption equilibrium. Subsequently, the lamp was turned on to trigger photocatalytic process. As can be seen from the figure, 97.7% of the initial methyl orange is decolorized on 5%In-0.05%Sm-TiO₂ after 45 min of irradiation. As a comparison, only 76.4% of the dye is removed from the solution after 60 min of irradiation with the existence of 0.05%Sm-TiO₂. Photocatalytic degradation plays the main role in decoloration of the dye. Meanwhile, it is believed that adsorption of pollutant on the surface of photocatalyst is important to the photocatalytic degradation process. Enhanced adsorption of methyl orange on 5%In-0.05%Sm-TiO₂ is also beneficial to photocatalytic activity as compared to 0.05%Sm-TiO₂. Photogenerated electrons and holes can readily move from photocatalyst to the adsorbed methyl orange molecules, leading to the subsequent degradation process. These charge carriers can hardly move from the surface of the photocatalyst to the methyl orange molecules in the solution due to fast recombination rate of electrons and holes.

Figure 7:

Removal of methyl orange during irradiation with the existence of 0.05%Sm-TiO₂ and 5%In-0.05%Sm-TiO₂ photocatalysts.

UV-Vis absorption spectra of methyl orange aqueous solution during photocatalytic degradation are shown in Figure 8. Some groups in methyl orange molecule have absorptions in both visible and ultraviolet regions. The spectrum of the initial methyl orange solution has a broad and strong absorption peak in the visible region. It is attributed to the conjugated chromophores in methyl orange molecule. The weak absorption peaks in the ultraviolet region are due to the absorptions of benzene ring in methyl orange molecule. Methyl orange is a typical azo dye and is characterized by its orange juice-like color after entering into the aquatic system. Although the initial concentration of methyl orange solution used in this work is only 10 mg/l, the solution has an apparent color so that treatment is needed before discharging.

Figure 8:

UV-Vis absorption spectra of methyl orange aqueous solution catalyzed by (A) 0.05%Sm-TiO₂ and (B) 5%In-0.05%Sm-TiO₂.

The solution containing 5%In-0.05%Sm-TiO₂ can be decolorized much faster than the solution using 0.05%Sm-TiO₂. The absorption in the visible region almost disappears after 60 min of photocatalytic degradation using 5%In-0.05%Sm-TiO₂ as photocatalyst. Methyl orange molecules are decomposed into small fragments as a result. The conjugated chromophores in methyl orange molecule can be decomposed thoroughly at this time, leading to total decoloration of the solution. However, weak absorptions in ultraviolet region can still be found in the spectra. Some organic intermediates still remain in the solution. Total mineralization of the organic substances needs much longer time.