INTRODUCTION
Photocatalysis is an efficient alternative for common treatment techniques, which are unable to remove complex and persistent contaminants from the aqueous media [1, 2]. Annually, millions of tons of organic compound, including surfactants, dyes, pesticides, and oil based chemicals, are discharged into water resources as wastes, posing a significant threat to marine ecosystem, and in turn to public health [3]. Recent increase in the water contamination has raised serious concerns, as we humans rely on oceans and rivers to supply food and drinking water. Of the pollutants discharged into the environment, effluent from textile dyeing industries is the big challenge, because of its adverse impacts on the human health. It is well known that organic dyes can cause cancerous diseases or neurological disorders [4, 5].
Although photocatalytic process is the promising approach for the environmental remediation, its large-scale applications restrict by some limitations [6]. Common photocatalysts such as TiO2 and ZnO have the wide range band gap corresponding to the energy of light in the UV region [7]. This point specifies that the high energy is needed to activate the photocatalysts [8]. Furthermore, rapid recombination rate of charge carriers is strongly decreased the photocatalytic performance [8]. Tailoring the photocatalyst composition has been proposed to surmount these limitations and render the photocatalysts suitable for degradation of wide range of the organic pollutants. In this regard, mixed metal oxides have received considerable attention due to their superior properties in batteries [9], energy production [10], solar cell [10], and photocatalysis [11]. Electronic transition between different energy levels of coupled metal oxide materials effectively minimize electron–hole pair recombination and improve photocatalytic activity in the visible light range [8, 12]. For instance, Abed et. al synthesized CuEr2TiO6 nanoparticles by sol-gel Pechini and studied their photocatalytic activity for degradation of acid red 88 under visible light irradiation [13]. Also, photocatalytic activity of CoFe2O4/Bi2MoO6 was studied for degradation of rhodamine blue under visible light irradiation [14].
Besides, reusability of photocatalysts is crucial to for their effective use in the removal of pollutants [15]. Hindering the loss of photocatalysts during the successive uses not only reduces the cost of the photocatalytic utilization, but also minimizes the possible side effects caused by the accumulation of photocatalysts in the environment [16]. To meet this demand, magnetic photocatalysts are the great solution, providing the ease of recovery and reuse of photocatalysts [17, 18].
Herein, efficient and magnetically separable visible-light photocatalyst nanoparticles, YbFeTi2O7, were prepared using straightforward sol-gel reaction. Characterization of the nanoparticles was performed using XRD, FTIR, FESEM, TEM, VSM, and DRS analyses. To evaluate to the photocatalytic activity of the synthesized nanoparticles, the degradation of an aqueous solution of methylene blue (MB) was studied under the visible light irradiation. Effect of different determinants including pH of MB solution and dosage of the prepared photocatalyst were assessed on the photocatalytic efficiency of the MB degradation.

MATERIALS AND METHODS
Synthesis of photocatalyst
The sol-gel synthetic route was employed to prepare the YbFeTi2O7 (YFTO) nanoparticles. Initially, 0.02 mmol of citric acid was added to 50 mL of absolute ethanol as a complexing agent, then 2.0 mmol of tetra butyl orthotitanate (TBOT), 1.0 mmol of Fe(NO3)3.9H2O, and 1.0 mmol of Yb(NO3)3.5H2O were respectively added under vigorous stirring. After 2 hours of stirring, the obtained sol solution was heated overnight in an oven at 80 °C to form the dried gel. Finally, the dried gel was heated at 600 °C for 6 hours to obtain light brown nanoparticles. 

Photocatalytic experiments
The methylene blue (MB) degradation was studied using the synthesized YFTO photocatalyst under the visible light irradiation. For that, 50 mL of MB solution (50 ppm) was illuminated from the distance of 30 cm by an LED lamp (daylight white, 50 W) in the presence of different dosages of the YFTO nanoparticles (0.01, 0.03, 0.05, and 0.07 g). Also, the photocatalytic degradation was investigates using different pH (3, 5, 7, 9, and 11) of MB solution.
The dye solution was first allowed to interact with the photocatalyst particles until adsorption–desorption equilibrium was achieved before illumination. So that the MB solution containing the YFTO photocatalyst was stirred in darkness for 30 min. After that, the degradation level was monitored in constant interval time (30 min) using UV/Vis spectrophotometer at maximum wavelength of MB (668 nm).

Characterization
Crystalline phase of the synthesized YFTO nanoparticles was studied using X-ray diffraction (XRD) pattern by Philips diffractometer (X’pert Pro MPD, Cu Kα = 1.54 Å). Functional groups on the surface of the synthesized YFTO nanoparticles were detected using Fourier transform infrared (FTIR) spectroscopy by Bruker Tensor 27 FTIR spectrophotometer. Morphology of the YFTO nanoparticles was studied using field emission electron microscope (FESEM) (TESCAN mira 3) and transmission electron microscope (TEM) (Zeiss, EM 900). Vibrating sample magnetometer (VSM) (VSM MDKB) was used to study the magnetic properties of the YFTO nanoparticles. The diffuse reflectance spectroscopy (DRS) (JASCO UV/Vis V-670) was used to assess the visible-light absorption ability by the nanoparticles.

RESULTS AND DISCUSSION
Structural, morphology, magnetic, and optical investigations
Fig. 1a exhibits the XRD pattern for the synthesized YFTO nanoparticles, confirming all the diffraction planes correspond to the orthorhombic phase of Iron Ytterbium Titanium Oxide (JCPDS file no. 047-0314, space group: Pcnb). The average crystallite size of the synthesized nanoparticles was calculated to be 18.31 nm using Scherrer equation, base on the peak at 2ϴ = 30.5°. 
The FTIR spectrum (Fig. 1b) depicts the functional groups of the YFTO nanoparticles. The broad absorption at around 3400 cm-1 is related to the stretching vibration of O-H bonds. The characteristic absorption of metal-oxygen bonds occur at 686 and 577 cm-1 which are attributed to the Ti-O and Fe-O respectively [19]. Additionally, the weak peak centered 480 cm-1 can be assigned to Yb-O bonds [20]. Two bands at 1490 and 1381 cm-1 are assigned to the carbonated group adsorbed on the surface of nanoparticles due to exposing to the ambient condition [21, 22].   
The FESEM image in Fig. 2a exhibits the morphology of the YFTO nanoparticles, showing semi-spherical nanoparticles in size range between 40-100 nm. Due to the large surface area of the synthesized nanoparticles, the agglomeration is inevitable, so that larger size nanoparticles are observed in the FESEM image. Moreover, the morphology of the YFTO nanoparticles was studied using TEM image (Fig. 2b), which reveals the semi-spherical nanoparticles with average size below 100 nm. Fig. 2c shows the EDX spectrum of the synthesized YFTO nanoparticles, confirming the existence of constituents of the nanoparticles and success of the synthetic route. Also, inset to the Fig. 2c represents the microanalysis of the YFTO with relative amounts of each components. The M-H curve, shown in Fig. 3, reveals the superparamagnetic behavior for the synthesized YFTO nanoparticles with the saturation magnetization (Ms) of 40.95 emu/g. This character assists to recover the photocatalyst nanoparticles from the aqueous media [23]. Using a bar magnet, the YFTO nanoparticles were easily separated from the MB solution. Additionally, due to the zero remnant magnetization in the absence of external magnetic field, the nanoparticles tend to disperse in the MB solution without occurring considerable agglomeration, resulting in the enhanced photocatalytic activity [24]. 
The optical responses of the synthesized YFTO nanoparticles was investigated by the DRS spectroscopy. The highly promoted visible light absorption is observable in the DRS spectrum (Fig. 4), confirming the potential of the YFTO nanoparticles to serve as the highly activated visible light photocatalyst. This result is truly attributed to the electron transitions between the components, that is, Yb2O3, Fe2O3, and TiO2, which substantially increases the visible light sensitivity of the nanoparticles and also decreases the recombination rate of the charge carriers [25-27]. The inset to Fig. 4 shows the determined the band gap of the nanoparticles was calculated by the Tauc method. The optical band gap of the YFTO nanoparticles was determined to be 2.44 eV which corresponds to the energy of photons in the visible light region.

Photocatalytic investigations
Photocatalyst amount 
The photocatalytic oxidation of MB solution was studied using different amount of the YFTO nanoparticles (0.01, 0.03, 0.05, and 0.07 g), which is shown in Fig. 5.  As can be seen in Fig. 5, the YFTO nanoparticles is able to efficiently degrade the MB molecules under 120 min illumination by the visible light irradiation. The highest degradation performance was 95.32% using 0.03 g of the synthesized YFTO. The exceeded amount of the YFTO nanoparticles led to the significant reduction of the photodegradation efficiency. Because photocatalyst particles require photons with sufficient energy for activation, excessive dispersion of nanoparticles in the turbid dye solution limits light penetration and consequently decreases photocatalytic efficiency [28]. 
Furthermore, Fig. 5 shows without addition of any the YFTO nanoparticles, the MB solution hardly decomposes on its own when exposed to visible light. Also, under no light condition, the reduction in the MB absorption is insignificant, confirming that the reduced absorption of the MB solution during illumination is attributed to the degradation of the MB molecules nor adsorption on the surface of the nanoparticles.

Effect of pH 
The pH of dye solution impacts the surface charges of photocatalyst, thereby affecting the adsorption of dye molecules. Since the photocatalytic reaction primarily occurs on the surface of the photocatalyst, it is clear that the pH of solution has the significant effect on the photocatalytic efficiency [28]. Fig. 6 shows the photocatalytic efficiency of the YFTO in different pH conditions of the MB solution (3, 5, 7, 9, and 11). As can be seen, the degradation level of the MB solution in alkaline conditions (97.83% at pH = 9) is higher than that of in the acidic conditions (84.7% at pH = 5). These observations are attributed to the increased adsorption of MB molecules on the surface of YFTO photocatalyst in alkaline media. Due to the cationic characteristic of MB [29], there is the increased affinity to adsorb on the negatively surface charged of YFTO and in the higher pH values. However, more alkaline condition (pH = 11) led to the significant decrease in the photocatalytic degradation.

Reusability experiments
Reusability tests were carried out over 5 consecutive reaction cycles to evaluate the stability of the synthesized YFTO photocatalyst. After reaction cycle, the photocatalyst was separated by a bar magnet, washed several times with ethanol/deionized water solution.  Then, the rinsed photocatalyst was dried at 90 °C for 1 h. Fig. 7 exhibits the performance of the MB photo-oxidation using 0.03 g of the YFTO photocatalyst over 5 reaction cycles. As seen, the decline in the photocatalytic efficiency is insignificant during 600 min of the consecutive reaction cycles (5 reactions), confirming superior stability of the YFTO photocatalyst. The loss of the photoactivity was only 2.85% after the fifth photodegradation reaction.     
Additionally, Fig. 8 represents the XRD patterns for the YFTO photocatalyst before and after performing 5 successive experiments. There are no prominent changes in the crystalline structure, which validate the structural stability of the YFTO. 

Photodegradation mechanism
In order to investigate the mechanism of the MB degradation using the YFTO photocatalyst, the radical scavenging experiments were carried out using varied radical scavengers. Eethylenediaminetetraacetic acid (EDTA) [30], ascorbic acid [31], and isopropanol were used to suppress the photogenerated holes, superoxide radicals, and hydroxyl radicals, respectively [32].  Fig. 9 exhibits the effect of different radical scavengers on the visible-light degradation of MB solution using the YFTO photocatalyst. As can be seen form the Fig. 9, the degradation level is significantly decreased in the presence of isopropanol and ascorbic acid, confirming the role of superoxide (•O2‾) and hydroxyl (•OH) radicals to contribute in the degradation of MB molecules.
Fig. 10 depicts the mechanism of MB degradation using the synthesized YFTO nanoparticles. The light brown color YFTO nanoparticles dispersed in MB solution are activated by visible light beams emitted from the LED lamp. The photo-generated electrons and holes contribute in the redox reactions to form the highly activated •O2‾ and •OH radicals. The radicals are capable degrading the MB molecules during 120 min illumination.

CONCLUSION
In summary, we have presented the sol-gel synthesis of YbFeTi2O7, YFTO, nanoparticles as the magnetically separable photocatalyst. The DRS analysis confirmed that the synthesized YFTO nanoparticles have capability serving as the highly efficient visible light photocatalyst. The band gap of the nanoparticles was calculated to be 2.44 eV. The YFTO nanoparticles exhibit the superparamagnetic behavior, making them viable magnetic photocatalyst, which is crucial for efficient recovery of the photocatalyst particles from the reaction media using a bar magnet. The visible light photocatalytic performance of the synthesized nanoparticles was investigated for degradation of aqueous solution of MB. Using 0.03 g of the YFTO photocatalyst, the photocatalytic efficiency was approximately 95% under 120 min of visible light illumination. In addition, pH of MB solution influenced the photoactivity of YFTO nanoparticles, so that alkaline condition was in the favor of achieving the higher photodegradation level of MB. The reusability experiments showed the conspicuous photocatalytic performance over 5 successive reaction cycles, revealing the great structural stability as confirmed by XRD patterns. The radical scavenging investigations were also performed to study the degradation mechanism. It was found that the MB photodegradation reaction proceeded through oxidation using •O2‾ and •OH radicals.  

CONFLICT OF INTEREST
The authors declare that there is no conflict of interests regarding the publication of this manuscript.