Fe2O3/TiO2 nanocomposite photocatalyst prepared by supercritical fluid combination technique and its application in degradation of acrylic acid

Fe2O3/TiO2 nanocomposite photocatalysts were synthesized by supercritical fluid combination technique, consisting of sol-gel method and supercritical fluid drying. The photocatalytic activity of the samples was evaluated by the degradation of acrylic acid. The results indicated that the Fe2O3/TiO2 nanocomposite catalysts prepared by this novel technique showed significant improvement in catalytic activity compared with pure TiO2 or Fe2O3/TiO2 catalysts prepared by traditional drying. Both infrared and ultraviolet spectrum of Fe2O3/TiO2 nanocomposite photocatalysts shift a little to lower wavelength indicating that the absorption threshold of Fe doped nanocomposite photocatalysts shift into the visible light region. This phenomenon was also attested by the photocatalytic degradation test under visible light.


Introduction
While acrylic acid and acrylic esters [1] are playing important roles in chemical industry, much attention is being given to their production because of a huge amount of wastewater released in the production.
Semiconductor photocatalysis appears as an emerging destructive technology leading to the total mineralization of most of organic pollutants [2]. It offers various advantages such as low cost, no secondary pollution, photoreaction occurring at room temperature and atmospheric pressure [3]. Although significant progress has been made on the waste water treatment, photocatalytic degradation of industrial wastewater with acrylic acid remains an untouched area.
Although TiO 2 is widely used in photocatalyst due to its good optical and electronic properties, chemical stability and non-toxicity and low cost [4], several factors limit its practical application. The first factor is that TiO 2 has a wide band gap of 3.2 eV which can only be excited by high energy UV irradiation with wavelength no longer than 387.5 nm. The second factor is that TiO 2 has a high recombination of photogenerated electron/hole pairs [5,6]. One of these materials that has aroused much research interest is the so-called composite semiconductor photocatalyst. It could extend light absorption to the visible light range [7]. A possibility to achieve this goal is to dope the titanium dioxide with transition metal ions. It has been shown that Fe (III) doping can increase the photocatalytic activity of TiO 2 [8].
In this paper, the nanocomposite photocatalyst Fe 2 O 3 /TiO 2 was prepared by supercritical fluid combination technique (SCFCT). Its photocatalytic activity for the degradation of acrylic acid under ultraviolet light and visible light was also assessed. The success of the present study should shed a light on the feasible application of acrylic acid in the waste water treatment.

Preparation of Catalyst
Surfactant (0.3v%) was added to 0.3 mol·L -1 TiCl 4 , stirring vigorously at room temperature. The pH of the solution was adjusted to 8-9. After aging for 16 h, the sol was filtered and washed until no presence of Cl -1 . The water in the sol was replaced by anhydrous ethanol. The alcogel obtained was transferred into an autoclave and dried under the supercritical condition of ethanol (262 °C, 8.5 MPa) to produce aerogel. After calcination at different temperature for 1 h, pure nano-TiO 2 was obtained. 0.1 mol·L -1 Fe(NO 3 ) 3 and surfactant (0.3v%) was added to 0.3 mol·L -1 TiCl 4 in sequence, stirring vigorously at room temperature. The following processes were carried out as mentioned in above. Finally, different molar percentage of Fe 2 O 3 /TiO 2 nanocomposite catalysts was obtained.
The composite alcogel was also dried at room temperature and then calcined at different temperature for 1 h in air atmosphere. Then (0.03mol%) Fe 2 O 3 /TiO 2 composite catalysts were obtained.

Characterizations
The UV irradiation of the high pressure mercury lamp was measured with the UV radiometer (Beijing Normal University photo electric company). The content of Fe was determined by atomic absorption spectrophotometer (Z-8000). The surface morphologies and particle size were determined from transmission electron micrographs made with Hitach-800 TEM. X-ray diffraction measurements were performed using a Shimadzu HR6000X (Cu target X tube, voltage 40.0 KV, current 30.0 mA). The IR spectra were recorded with a Shimadzu IRPrestige-21 spectrophotometer. Diffuse UV-Vis spectra were obtained with a UV-2501PC Ultraviolet spectrophotometer. The UV irradiation of the high pressure mercury lamp was measured with the UV radiometer (Beijing Normal University photo electric company).

Evaluation of Photocatalytic Activity
The photocatalytic activities of the synthesized catalysts were examined using a self-designed photocatalytic reaction apparatus. It consists of double-layered cylindrical quartz flask with an ultraviolet lamp (9 W, wavelength: 254 nm, intensity: 5210 µw·cm -2 ) or a solar lamp (18 W) positioned in the middle. Acrylic acid aqueous solution (300 mL, 300 mg·L -1 ) with the chemical oxygen demand (COD) value of 400~500 mg·L -1 was used as degradation solution. The concentration of catalyst was 1g·L -1 . The air was introduced from the bottom of the reaction solution at the rate of 30 mL·min -1 . The temperature was maintained at 25±1 ºC. Stirring was kept to make sure that catalyst was dispersed evenly in the reaction solution. A small volume of the reaction solution was obtained periodically. After centrifugation, the absorbance of these solutions was measured at 210 nm by B-UV 752 spectrophotometer. The COD value of the reaction solution was determined by potassium dichromate method. The photocatalytic activities were estimated from the degradation rate of acrylic acid.

Results of AAS and XRD measurements
The results of AAS showed that the content of Fe in the composite photocatalysts was 0.03%, 0.1% and 0.15% (molar percentage), respectively. Figure1 shows the XRD patterns of selected samples (0.03mol% Fe 2 O 3 /TiO 2 ) prepared by SCFCT with calcination at 500 ºC, 600 ºC, 700 ºC and without calcination. The observed patterns can be attributed to the TiO 2 anatase as unique phase, which has diffraction peaks at d 1 =0.35256 nm, d 2 =0.18945 nm and d 3 =0.23768 nm. No transformation from anatase type to rutile type was observed after heat treatment. It confirms the conclusion that TiO 2 prepared by SCFCT exists only as anatase type crystal, that is the drying and crystallizing could be accomplished in a single step in supercritical 2 1st International Conference on New Material and Chemical Industry (NMCI2016) IOP Publishing IOP Conf. Series: Materials Science and Engineering 167 (2017) 012037 doi:10.1088/1757-899X/167/1/012037 fluid drying process [9]. It also can be found that the crystallinity was enhanced by the calcination temperature. There is no peak corresponding to Fe 2 O 3 present in the XRD patterns. This result could be attributed to that there is no independent Fe 2 O 3 crystalline phase accumulated on the surface of TiO 2 , or Fe 2 O 3 disperses uniformly in the bulk of TiO 2 particle as very small clusters.
The size of domain crystallite estimated from Scherrer equation was 8.9 nm (without calcination), 10.9 nm (500 ºC), 15.8 nm (600 ºC) and 27.3 nm (700 ºC). These results indicate that the size of the particles increases with the increase of the calcination temperature.   The UV-Vis spectra for TiO 2 and 0.03mol% Fe 2 O 3 /TiO 2 are shown on the figure 4. From the curves we can see that UV-Vis absorbance of 0.03mol% Fe 2 O 3 /TiO 2 is shifted to longer wavelength (the red shift) compared with TiO 2 . The band gap energy (Eg) of anatase TiO 2 extrapolates the absorption edge onto the energy 3.2 ev (387.5 nm) [10]. The electron-hole pairs can only be created by the light whose wavelength is shorter than 387.5 nm. In this experiment, the value of band gap energy of pure TiO 2 shown in the figure 4 is fairly consistent with the above. The band gap energy of 0.03mol% Fe doped on TiO 2 is reduced to 2.8 ev (430.0 nm), showing that red shift occurred. It is thought that additional band gap energy was created by Fe doping so that electron-hole pairs can be excited in the visible light wavelength region and red shift occults is possible.

Effect of Fe 2 O 3 dosage on catalytic activity
The catalytic activities of Fe 2 O 3 /TiO 2 nanocomposite catalysts prepared under the same conditions (containing 0.00mol%, 0.03mol%, 0.10mol% and 0.15mol% Fe 2 O 3 respectively) are shown on figure  5. It can be seen that 0.03mol% Fe 2 O 3 doped TiO 2 is more photoactive than any other samples. 0.1mol% Fe 2 O 3 doped TiO 2 has high photoactivity in comparison with TiO 2 , but 0.15mol% Fe 2 O 3 doped TiO 2 has lower Fe 2 O 3 doped TiO 2 than pure TiO 2 .
The beneficial effect of Fe 3+ may be explained as follows: 1) The ion radius of Fe 3+ (0.063 nm) [11] is much closer to that of Ti 4+ (0.068 nm) [12], so the Ti 4+ sites could be easily substituted by Fe 3+ ions; 2) The Fe 2+ /Fe 3+ energy level lies close to Ti 3+ /Ti 4+ level (figure 6), similar to anatase, thus photogenerated electrons are transferred from TiO 2 to Fe 2+ /Fe 3+ energy level (Fe 3+ + e -→Fe 2+ ). Fe 3+ can be an effective electron trap. Meanwhile, due to the energy level of Fe 3+ /Fe 4+ being above the valence band edge of anatase TiO 2 [13], Fe 3+ can also be served as vacancy trap: Fe 3+ + h + →Fe 4+ . However, high doping concentrations are not desirable, because of the high possibility that both charge carriers (e -/h + pair) will be trapped, and on the other hand, more doping agents reduce the number of active centers on the surface of TiO 2 catalysts.  It can be seen from table 1 that 0.03mol% Fe 2 O 3 /TiO 2 nanocomposite photocatalyst has higher photoactivity than TiO 2 and Fe 2 O 3 , both under the ultraviolet light and visible light. This result is constant with the result of Uv-Vis. It is also suggested that light condition plays a vital importance in the degradation of acrylic acid. The efficient application of solar spectra still need to be further investigated.  It can also be found that the degradation rate increases with the increase of calcination temperature. Sample calcined at 600 °C has the highest activity. When the temperature raises up to 700 ºC the activity decreases to 46.75%. Low catalytic activity at low calcination temperature is attributed to the covered active centers by some organic compounds such as alcohols and surfactants. With the increase of calcination temperature, decomposition of the impurities will expose more active center on the surface, thus leading to improvement in the activity of the catalysts. However, when the calcination temperature is increased to 700 ºC, a rapid particle growth from 15.8 nm to 27.3 nm is occurred. The growth of the powder particles leads to the decrease of the surface area and photoatalytic activity.

Effect of preparation methods on the catalytic activities
It is obvious that the catalyst prepared by SCFCT shows higher photocatalytic activity than that of prepared by TD as shown in figure 7. It is due to the interfacial force of the solvent, which leads to collapse of the cavity structure and aggregation of particles. And therefore, the particle size increases and cavity volume decreases. Correspondingly, the surface area and the photocatalytic activity are reduced. This phenomenon could be avoided in the preparation of aerogels obtained by SCFCT. At supercritical condition of alcohol, the solvent in the pores expands rapidly without interfacial force. The composite particles with good dispersity and small size lead to quantum size effect and improve the carriers transfer efficiency. Thus, the catalysts prepared through SCFCT show higher catalytic activity.    9), which hinders the electron-vacancy recombination rate and increase the rate of the photocatalytic process [14]. At high concentrations of hydrogen peroxide, the reactions shown in figure 10 will occur.

Results for COD Cr
As can be seen from table 3, after 3 h treatment with 0.03mol% Fe 2 O 3 /TiO 2 and 0.167v% H 2 O 2 , the COD Cr (η COD ) value of acrylic acid suspensions is reduced from 448.4 mg·L -1 to 35.57 mg·L -1 (or by 92.07%). The degradation rate determined by absorbency method is 100%. Both results indicate that almost the whole amount of acrylic acid is decomposed into H 2 O and CO 2 .