Facile Synthesis of Tin Dioxide Nanoparticles for Photocatalytic Degradation of Congo Red Dye in Aqueous Solution

: This research work reports an approach used to prepare a SnO 2 photocatalyst by precipitation and calcination pathways and describes an investigation of the e ﬀ ects of preparation parameters on SnO 2 yield. The SnO 2 photocatalyst was further used for the photocatalytic degradation of Congo red (CR) dye, and the removal e ﬃ ciency was optimized using response surface methodology. The results indicate that the SnO 2 photocatalyst yield was the highest in 0.05 M of the precursor, stannous chloride and 28 wt % ammonia as the precipitant, pH 10, at 30 ◦ C. The transmission electron microscopy results of the SnO 2 photocatalyst illustrate that the average particle size was mainly around 30–50 nm and had a solid spherical shape. The X-ray di ﬀ raction results reveal that the prepared sample had a highly crystalline SnO 2 rutile crystal structure. The prediction and experimental results of the Response surface methodology (RSM) indicate that, when the reaction time was 97 min, the operating temperature was 47 ◦ C, the photocatalyst dosage was 751 mg / L, and the optimal degradation rate of the CR dye was 100%. After ﬁve consecutive photodegradation reactions, the degradation rate remained at 100%. The results demonstrated that the SnO 2 photocatalyst prepared in this study possesses excellent reusability.


Introduction
As the demand for dyed products has increased, the discharge of dye wastewater has increased proportionally. Globally, dye wastewater released into the environment has become one of the primary sources of serious pollution [1]. Its release is not only a problem due to the color of the wastewater but also because several of its intermediate products are toxic. Congo red (CR) mainly exists in wastewater discharged from the textile, papermaking, printing, and leather industries [2]. CR is a type of azo dye derived from benzidine. Due to its complex chemical structure and high solubility in aqueous solutions, once discharged into the natural environment, CR is persistent and has low biodegradability. The metabolic intermediate of CR is benzidine, which is a known carcinogen and may also cause some allergic reactions [3].
Environmental problems caused by wastewater affect the environment and human health. Numerous wastewater treatment technologies exist (e.g., coagulation, adsorption, ion exchange, and activated sludge [4][5][6][7][8][9]); however, they have several drawbacks and limitations-they are expensive, time-consuming, and release a considerable amount of secondary pollutants during treatment. was conducted to evaluate the range of experimental parameters. Subsequently, the conditions of the optimal photocatalytic degradation efficiency were identified through the RSM experimental design, and factors affecting degradation efficiency were discussed. Therefore, this study could provide useful data for the development of SnO 2 photocatalyst in the future.

Operating Conditions for the Preparation of SnO 2
In the SnO 2 preparation processes, the precursor concentration, precipitant agent, and reaction temperature are considered the most important parameters that can affect the SnO 2 photocatalyst prepared using the precipitation method. In this work, stannous chloride was dissolved in water and then alkali (precipitant) was added to facilitate the following reaction [42,43]: The calcination process is a dehydration reaction and the oxidation produces SnO 2 nanoparticles [44], where the SnO to SnO 2 reaction occurs at temperatures above approximately 400 • C [42,45], as follows: 2SnO + O 2 → 2SnO 2 (4) Figure 1 depicts the correlation between different precursor concentrations and SnO 2 yield. The SnO 2 yield of the experiments using sodium hydroxide as the precipitant reached 92.95%, and the highest yield of the experiments using ammonia as the precipitant reached 97.58%. The SnO 2 yield increased as the precursor concentration increased, reaching equilibrium at 0.05 M. In addition, when the precursor concentration was found to be overly high, precipitation occurred without the addition of precipitant. This finding is consistent with those of previous studies [29,38]. The precipitate was white insoluble stannous-chloride-precipitated Sn(OH)Cl in water. Because the experimental condition of high-concentration precursors might have affected the stability of the precipitation process, a precursor concentration of 0.05 M was used as the experimental follow-up condition for the preparation of photocatalyst materials.
Catalysts 2020, 10, x FOR PEER REVIEW 3 of 18 optimal photocatalytic degradation efficiency were identified through the RSM experimental design, and factors affecting degradation efficiency were discussed. Therefore, this study could provide useful data for the development of SnO2 photocatalyst in the future.

Operating Conditions for the Preparation of SnO2
In the SnO2 preparation processes, the precursor concentration, precipitant agent, and reaction temperature are considered the most important parameters that can affect the SnO2 photocatalyst prepared using the precipitation method. In this work, stannous chloride was dissolved in water and then alkali (precipitant) was added to facilitate the following reaction [42,43]: The calcination process is a dehydration reaction and the oxidation produces SnO2 nanoparticles [44], where the SnO to SnO2 reaction occurs at temperatures above approximately 400 °C [42,45], as follows: 2SnO + O2 → 2SnO2 (4) Figure 1 depicts the correlation between different precursor concentrations and SnO2 yield. The SnO2 yield of the experiments using sodium hydroxide as the precipitant reached 92.95%, and the highest yield of the experiments using ammonia as the precipitant reached 97.58%. The SnO2 yield increased as the precursor concentration increased, reaching equilibrium at 0.05 M. In addition, when the precursor concentration was found to be overly high, precipitation occurred without the addition of precipitant. This finding is consistent with those of previous studies [29,38]. The precipitate was white insoluble stannous-chloride-precipitated Sn(OH)Cl in water. Because the experimental condition of high-concentration precursors might have affected the stability of the precipitation process, a precursor concentration of 0.05 M was used as the experimental follow-up condition for the preparation of photocatalyst materials.  Figure 2 depicts the correlation between the concentrations of different precipitants (sodium hydroxide and ammonia hydroxide) and yield. The trends indicated that the use of different precipitant concentrations has slight effects on the yield. The yield of strong alkali sodium hydroxide at 1-4 M was in the range of 94% ± 3%, and the yield of weak alkali ammonia at 7-28 wt % was in the  concentrations has slight effects on the yield. The yield of strong alkali sodium hydroxide at 1-4 M was in the range of 94% ± 3%, and the yield of weak alkali ammonia at 7-28 wt % was in the range of 98% ± 1%, indicating that the yield of ammonia was higher than that of sodium hydroxide. As the precipitant concentration of the ammonia increases, the light transmission in the samples are reduced, as shown in Figure 3a. The UV-vis full-wavelength spectrum indicated that the Sn 6 O 4 (OH) 4 prepared using ammonia had a good stability of suspension. After the samples were left overnight, the ammonia sample was still suspended, and the sodium hydroxide sample was mostly precipitated. Therefore, 28 wt % ammonia was used as the follow-up experimental condition.
Catalysts 2020, 10, x FOR PEER REVIEW 4 of 18 range of 98% ± 1%, indicating that the yield of ammonia was higher than that of sodium hydroxide. As the precipitant concentration of the ammonia increases, the light transmission in the samples are reduced, as shown in Figure 3a. The UV-vis full-wavelength spectrum indicated that the Sn6O4(OH)4 prepared using ammonia had a good stability of suspension. After the samples were left overnight, the ammonia sample was still suspended, and the sodium hydroxide sample was mostly precipitated. Therefore, 28 wt % ammonia was used as the follow-up experimental condition. Finally, the precursor of 0.05 M stannous chloride and precipitant (28 wt % ammonia water) were used to indicate the effect of reaction temperature for the yield of products, as shown in Figure  3b. Our calculations indicated that the yields were all above 98%, and the range of yield change was not considerable. The reason for the slight decrease in yield at 70 °C could be that, as the temperature rose, the solubility of the precipitate increased, causing the yield to decrease slightly. Overall, the preparation process uses a single reactor and low-toxicity chemicals, and is inexpensive. It provides an economically viable and environmentally friendly method for synthesizing a SnO2 photocatalyst.

Characterization of Photocatalyst
The morphology and elements of the photocatalysts were determined from SEM images and EDX analysis, respectively. Figure 4 depicts the SnO2 samples synthesized using different precipitants; both were calcined at 600 °C and neither had an additional dispersant. This indicated range of 98% ± 1%, indicating that the yield of ammonia was higher than that of sodium hydroxide. As the precipitant concentration of the ammonia increases, the light transmission in the samples are reduced, as shown in Figure 3a. The UV-vis full-wavelength spectrum indicated that the Sn6O4(OH)4 prepared using ammonia had a good stability of suspension. After the samples were left overnight, the ammonia sample was still suspended, and the sodium hydroxide sample was mostly precipitated. Therefore, 28 wt % ammonia was used as the follow-up experimental condition.
Finally, the precursor of 0.05 M stannous chloride and precipitant (28 wt % ammonia water) were used to indicate the effect of reaction temperature for the yield of products, as shown in Figure  3b. Our calculations indicated that the yields were all above 98%, and the range of yield change was not considerable. The reason for the slight decrease in yield at 70 °C could be that, as the temperature rose, the solubility of the precipitate increased, causing the yield to decrease slightly. Overall, the preparation process uses a single reactor and low-toxicity chemicals, and is inexpensive. It provides an economically viable and environmentally friendly method for synthesizing a SnO2 photocatalyst.

Characterization of Photocatalyst
The morphology and elements of the photocatalysts were determined from SEM images and EDX analysis, respectively. Figure 4 depicts the SnO2 samples synthesized using different precipitants; both were calcined at 600 °C and neither had an additional dispersant. This indicated Finally, the precursor of 0.05 M stannous chloride and precipitant (28 wt % ammonia water) were used to indicate the effect of reaction temperature for the yield of products, as shown in Figure 3b. Our calculations indicated that the yields were all above 98%, and the range of yield change was not considerable. The reason for the slight decrease in yield at 70 • C could be that, as the temperature rose, the solubility of the precipitate increased, causing the yield to decrease slightly. Overall, the preparation process uses a single reactor and low-toxicity chemicals, and is inexpensive. It provides an economically viable and environmentally friendly method for synthesizing a SnO 2 photocatalyst.

Characterization of Photocatalyst
The morphology and elements of the photocatalysts were determined from SEM images and EDX analysis, respectively. Figure 4 depicts the SnO 2 samples synthesized using different precipitants; both were calcined at 600 • C and neither had an additional dispersant. This indicated that the samples with ammonia as the precipitant had smaller particles. EDX element analysis revealed that the presence of only Sn and O, and other impurities were not found. The appearance of other peaks (e.g., 2.2 keV) are due to borosilicate glass, gold sputter coating and carbon substrate tape used to paste the sample in the holder [39]. The morphology and the average particles size of SnO 2 samples synthesized by precipitation were revealed using TEM. Figure 5a shows the morphology of un-calcined Sn 6 O 4 (OH) 4 , where un-calcined particles appear as irregularly-shaped polygons and an agglomeration phenomenon can be observed. Figure 5b shows that the SnO 2 particle profile after calcination was round and granular structures, but agglomeration still existed. SnO 2 particle size, as shown in Figure 5c, indicated a range of 20-80 nm and an average particle size of approximately 40 nm.
Catalysts 2020, 10, x FOR PEER REVIEW 5 of 18 that the samples with ammonia as the precipitant had smaller particles. EDX element analysis revealed that the presence of only Sn and O, and other impurities were not found. The appearance of other peaks (e.g., 2.2 keV) are due to borosilicate glass, gold sputter coating and carbon substrate tape used to paste the sample in the holder [39]. The morphology and the average particles size of SnO2 samples synthesized by precipitation were revealed using TEM. Figure 5a shows the morphology of un-calcined Sn6O4(OH)4, where un-calcined particles appear as irregularly-shaped polygons and an agglomeration phenomenon can be observed. Figure 5b shows that the SnO2 particle profile after calcination was round and granular structures, but agglomeration still existed. SnO2 particle size, as shown in Figure 5c, indicated a range of 20-80 nm and an average particle size of approximately 40 nm.
To understand the crystalline phase and structure, the synthesized SnO2 samples were characterized by XRD. Figure 6a compares the XRD patterns before and after calcination. The sample before calcination was crystalline Sn6O4(OH)4, the angle was 2θ, and the corresponding plane coordinates were 25   Before the photocatalysis experiment, we used UV-vis to measure the full-wavelength absorbance of SnO2 powder and to ensure that that the synthesized SnO2 had photocatalytic activity. Figure 6a shows that the absorbance wavelength of the SnO2 powder was in the ultraviolet wavelength range, and the band gap value of 3.575 eV was calculated using the Tauc plot method [46,47]. These values were similar to the band gap values found in earlier research studies [32,47]. As  To understand the crystalline phase and structure, the synthesized SnO 2 samples were characterized by XRD. Figure 6a compares the XRD patterns before and after calcination. The sample before calcination was crystalline Sn 6 O 4 (OH) 4 4 crystal phase material. Besides, according to the XRD pattern of the standard card (JCPDS No. 00-041-1445) of SnO 2 sample calcined at 600 • C, it seemed to exhibit in the tetragonal rutile structure. No other impurity diffraction peaks were observed, which indicated that the purity of the prepared SnO 2 photocatalyst sample was high.
Before the photocatalysis experiment, we used UV-vis to measure the full-wavelength absorbance of SnO2 powder and to ensure that that the synthesized SnO2 had photocatalytic activity. Figure 6a shows that the absorbance wavelength of the SnO2 powder was in the ultraviolet wavelength range, and the band gap value of 3.575 eV was calculated using the Tauc plot method [46,47]. These values were similar to the band gap values found in earlier research studies [32,47]. As the literature [38] indicated, the band gap energy of SnO2 of 3.1-3.9 eV should be activated by photons of a wavelength of 315-400 nm. That was why electrons from the valence band to the conduction band in SnO2 synthesized could be excited by ultraviolet light (365 nm) irradiation, thereby producing the photocatalytic effect. The specific surface area and pore size of the photocatalysts are also essential characteristics that can be used as indicators for the evaluation of the active sites and photocatalytic activity. The uncalcined samples and calcined samples at 600 and 700 °C were compared in terms of BET specific surface area, and the results are presented in Table 1. The BET-specific surface areas of each sample were 27.9671, 19.7177, and 14.3532 m 2 /g, respectively. The specific surface area of the un-calcined sample was the largest. As the calcination temperature rose, the specific surface area decreased, potentially because calcination at a higher temperature indicates a higher likelihood of SnO2 melting and recrystallizing [48]. This would collapse the original pore defects and decrease the pore size distribution and total BET surface area. Before the photocatalysis experiment, we used UV-vis to measure the full-wavelength absorbance of SnO 2 powder and to ensure that that the synthesized SnO 2 had photocatalytic activity. Figure 6a shows that the absorbance wavelength of the SnO 2 powder was in the ultraviolet wavelength range, and the band gap value of 3.575 eV was calculated using the Tauc plot method [46,47]. These values were similar to the band gap values found in earlier research studies [32,47]. As the literature [38] indicated, the band gap energy of SnO 2 of 3.1-3.9 eV should be activated by photons of a wavelength of 315-400 nm. That was why electrons from the valence band to the conduction band in SnO 2 synthesized could be excited by ultraviolet light (365 nm) irradiation, thereby producing the photocatalytic effect.
The specific surface area and pore size of the photocatalysts are also essential characteristics that can be used as indicators for the evaluation of the active sites and photocatalytic activity. The un-calcined samples and calcined samples at 600 and 700 • C were compared in terms of BET specific surface area, and the results are presented in Table 1. The BET-specific surface areas of each sample were 27.9671, 19.7177, and 14.3532 m 2 /g, respectively. The specific surface area of the un-calcined sample was the largest. As the calcination temperature rose, the specific surface area decreased, potentially because calcination at a higher temperature indicates a higher likelihood of SnO 2 melting and recrystallizing [48]. This would collapse the original pore defects and decrease the pore size distribution and total BET surface area.

Effect of the Operation Conditions on Photocatalytic Efficiency
To evaluate the photocatalytic efficiency of the prepared SnO 2 catalyst, the photocatalyst dosage, operating temperature, and initial concentrations were main factors considered as influencing the photocatalysis process. At a fixed operating temperature of 45 • C and a photocatalyst dosage of 500 mg/L, we conducted photodegradation of CR dyes with different initial concentrations (20,30, and 40 ppm). The results indicated that the degradation rate of 20 ppm was faster at lower concentrations; for example, in a 45 min degradation reaction employed, degradation rates at 20, 30, and 40 ppm were 58.80%, 55.32%, and 47.19%, respectively, as shown in Figure 7. We can infer that the active site of the catalyst was blocked, which slowed the degradation. Alternatively, in the reaction with high dye concentration conditions, a considerable amount of light might have been absorbed by the dye molecules and decreased the penetration of light, thereby reducing the catalytic efficiency.

Optimization and Interactions of Photocatalysis Process
In this study, the BBD in the RSM was used to identify the optimal parameters for the SnO2 photocatalytic degradation of CR. The control factors were the reaction time (15-150 min), operating temperature (30-60 °C), and catalyst dosage (100-1000 mg/L). In this experiment, the initial concentration of CR was fixed at 20 ppm, the volume of the CR solution was 500 mL, and the magnet was stirred at 400 rpm for photocatalytic degradation. Table 2 presents the experimental design table and experimental results. The experimental data were applied to evaluate linear, two-factor respectively, as shown in Figure 7b. The amount of SnO 2 catalyst is a crucial factor in the photocatalytic process due to its considerable influence on the rate of photocatalytic degradation. As the amount of catalyst increases, the active sites available for photocatalytic reactions increase; thus, the rate of dye removal increases. However, as the concentration of the catalyst increases and the particles aggregate, the specific surface area is likely to decrease, which would eventually lead to reduced photocatalytic activity [32]. An excessive amount of suspended catalyst particles would block the passage of ultraviolet light or sunlight and increase light scattering, thereby reducing the photocatalytic efficiency [12]. Nevertheless, under our experimental conditions, no reduction in efficiency was found.
Subsequently, 500 mg/L SnO 2 catalyst were used for the reaction condition in a 20 ppm CR solution. The effect of different degradation temperatures on the degradation rate was further tested. When the temperature was 25, 45, and 60 • C and the degradation time was 45 min, the degradation rates of CR were 45.18%, 62.53%, and 74.39%, respectively, as shown in Figure 7c. This indicated that, as the temperature increased, the photodegradation efficiency of the CR dye also increased. To understand the photocatalytic efficiency, we computed the rate constants using pseudo-first-order reaction kinetics, as shown in Figure 7d. The rate constants (k) of photodegradation at 5, 45, and 60 • C were 0.0134, 0.0228, and 0.0352 min −1 , respectively, and the R 2 (coefficient of determination) values were 0.9954, 0.9981, and 0.9832, respectively. The photodegradation of pollutants can be divided into two steps: adsorption and desorption on the catalyst and degradation on the catalyst's surface. First, increases in temperature seemed to be more favorable to dye adsorption and desorption on the photocatalyst [49]. Therefore, during the degradation of pollutants, the operating temperature slightly affects the ability of the photocatalyst to adsorb the dye. The size of the band gap changes with the temperature-as the temperature rises, the thermal energy could make the interatomic spacing increase, thereby weakening electrons bond and reducing the band gap [50,51]. The photogenerated electrons and electron holes generate more free radicals to react with pollutants to improve the degradation rate [52].

Optimization and Interactions of Photocatalysis Process
In this study, the BBD in the RSM was used to identify the optimal parameters for the SnO 2 photocatalytic degradation of CR. The control factors were the reaction time (15-150 min), operating temperature (30-60 • C), and catalyst dosage (100-1000 mg/L). In this experiment, the initial concentration of CR was fixed at 20 ppm, the volume of the CR solution was 500 mL, and the magnet was stirred at 400 rpm for photocatalytic degradation. Table 2 presents the experimental design table and experimental results. The experimental data were applied to evaluate linear, two-factor interaction, and quadratic models. The software recommended a quadratic model based on the analysis of the photocatalytic degradation model of CR. Following the model fitting, the second-order polynomial equation could be obtained, the calculation result of which was as follows: where A, B, and C are reaction times, temperature, and catalyst dosage, respectively. The predicted CR degradation rate was computed using Equation (5). The quadratic model of the ANOVA results is summarized in Table 3. The determination coefficient (R 2 ) of the empirical model was 0.9946, indicating that the model was suitable for analyzing the experiment results. The p-value of the model was <0.0001, which indicated that the effect of the second curvature on the response value was significant, which meant that the effect of the control factor on the response value was a curved relationship [34]. In Table 3, the p-values of A, B, C, A-B interaction, A-C interaction, A 2 , and C 2 were all less than 0.05. This indicated that the contribution to the regression model was significant, whereas the contribution of the B-C interaction and B 2 to the regression model was non-significant [35]. The results of the lack-of-fit test were non-significant (p-value > 0.05), indicating that the quadratic model was suitable. The result of the analysis of the conformance of the CR degradation rate model's CV (coefficient of variation) was 1.64%. Predicted Catalysts 2020, 10, 792 9 of 17 R 2 (Pred. R 2 ) was 0.9427, suggesting that the model predicted the variance of the new experimental response value to be 94.27% [53,54]. Besides, Adjusted R 2 (Adj. R 2 ) was 0.9877 and the difference from Pred. R 2 value was less than 0.2, which expressed that the model was well fitted [36]. A higher Adequate Precision value also referred to the signal to noise ratio (SNR), which can indicate greater precision. An Adequate Precision value is more significant than four reasonable model prediction efficiencies [55]. Therefore, the results indicated that the model could effectively predict the degradation rate of CR. The RSM can be used to analyze the response surface plot in a model to elucidate whether an interaction between various factors exists. The response surface plot depicts a three-dimensional (3D) plot and a contour map (2D) of the relationship between response factors and control factors. Figure 8a illustrates the relationship between reaction time and temperature. The 3D image shows a slightly distorted surface, which is a characteristic of an interaction effect [35]. The ANOVA suggested a p-value < 0.05, as shown in Table 3, which also meant that an interaction effect existed. However, the contour lines appeared to be straight, indicating that, although an interaction existed, the trend of the interaction relationship displayed on the graph is not evident because the effect of the reaction time was overly broad. The dense distribution of lines on the x-axis indicates that the reaction time was the main factor affecting the level of degradation rate. Figure 8b also presents the relationships between reaction time and catalyst dosage. The contour line shows a micro-curved appearance and confirms the existence of an interaction. The ANOVA indicated a p-value < 0.05, and thus an interaction existed. The line distribution density on the x-axis indicates that the reaction time still had a major effect on the level of degradation rate. Figure 8c presents the relationship between operating temperature and catalyst dosage. The 3D image was flat, which suggested that the operating temperature did not interact with the catalyst dosage. The ANOVA indicated a p-value > 0.05, which meant that no interaction existed, and the density of the contour plot was evenly distributed among the two factors. Therefore, it was hard to determine the impact level between the operating temperature and catalyst dosage on the degradation rate from the response surface plot. However, the large F-value can be used to check statistical significance [33]. A higher F-value implied a more considerable influence on the CR degradation rate. Table 3 shows that the significant effects on CR degradation rate were in the following order: reaction times > catalyst dosage > temperature.
interaction between various factors exists. The response surface plot depicts a three-dimensional (3D) plot and a contour map (2D) of the relationship between response factors and control factors. Figure  8a illustrates the relationship between reaction time and temperature. The 3D image shows a slightly distorted surface, which is a characteristic of an interaction effect [35]. The ANOVA suggested a pvalue < 0.05, as shown in Table 3, which also meant that an interaction effect existed. However, the contour lines appeared to be straight, indicating that, although an interaction existed, the trend of the interaction relationship displayed on the graph is not evident because the effect of the reaction time was overly broad. The dense distribution of lines on the x-axis indicates that the reaction time was the main factor affecting the level of degradation rate. Figure 8b also presents the relationships between reaction time and catalyst dosage. The contour line shows a micro-curved appearance and confirms the existence of an interaction. The ANOVA indicated a p-value < 0.05, and thus an interaction existed. The line distribution density on the x-axis indicates that the reaction time still had a major effect on the level of degradation rate. Figure 8c presents the relationship between operating temperature and catalyst dosage. The 3D image was flat, which suggested that the operating temperature did not interact with the catalyst dosage. The ANOVA indicated a p-value > 0.05, which meant that no interaction existed, and the density of the contour plot was evenly distributed among the two factors. Therefore, it was hard to determine the impact level between the operating temperature and catalyst dosage on the degradation rate from the response surface plot. However, the large F-value can be used to check statistical significance [33]. A higher F-value implied a more considerable influence on the CR degradation rate. Table 3 shows that the significant effects on CR degradation rate were in the following order: reaction times > catalyst dosage > temperature. The RSM model can be used for residual analysis (e.g., the model's normality, equal variance, and independence) to evaluate suitability [56]. If the residuals are normally distributed, the points depicting the residuals will resemble a straight line. The focus of the residual plot observation was between 25% and 75%, indicating that both sides were not in an extreme state. Figure 9a presents a typical probability diagram of the residual of the CR degradation experiment. The distribution of each point of the experiment was closely aligned with the straight line and no outlier was observed. Accordingly, this experimental model did not violate the assumption of the normal behavior of the degradation rate of CR. We used residuals to analyze the scatter plots of the predicted values. If the graph had no shape or was it divergent or convergent, then the hypothesis that the number of residual variances is equal would be supported. The extra points in Figure 9b are randomly distributed and no flared funnel-shaped structure can be observed, and thus no divergence or convergence occurred. The even distribution of the residuals indicated that the predicted values were in a normal state. Therefore, the hypothesis that the variances of the CR degradation rate model are equal was supported. The RSM model can be used for residual analysis (e.g., the model's normality, equal variance, and independence) to evaluate suitability [56]. If the residuals are normally distributed, the points depicting the residuals will resemble a straight line. The focus of the residual plot observation was between 25% and 75%, indicating that both sides were not in an extreme state. Figure 9a presents a typical probability diagram of the residual of the CR degradation experiment. The distribution of each point of the experiment was closely aligned with the straight line and no outlier was observed. Accordingly, this experimental model did not violate the assumption of the normal behavior of the degradation rate of CR. We used residuals to analyze the scatter plots of the predicted values. If the graph had no shape or was it divergent or convergent, then the hypothesis that the number of residual variances is equal would be supported. The extra points in Figure 9b are randomly distributed and no flared funnel-shaped structure can be observed, and thus no divergence or convergence occurred. The even distribution of the residuals indicated that the predicted values were in a normal state. Therefore, the hypothesis that the variances of the CR degradation rate model are equal was supported. Residuals were plotted in chronological order of experimental data collected in this study. If a series of positive or negative residuals is observed, a strong correlation between the residuals exists, which would violate the assumption of residual independence. Figure 9c indicates that the residuals were randomly distributed and no series of positive or negative residuals existed. Therefore, the hypothesis of independence of the CR degradation rate model was supported [35].
To verify the feasibility of the RSM, we compared the predicted maximum degradation rate of CR and the experimental results under the optimum conditions. The predicted maximum for the degradation rate of CR using this model was 100% and the optimized photocatalytic conditions were: reaction time = 97 min; operating temperature = 47 • C; photocatalyst dosage = 751 mg/L. Under the optimized conditions, the experimental results suggested that the degradation rate of CR was 100%, which was consistent with the model prediction.

Photocatalyst Comparison and Reusability
As the literature revealed, dyes have been removed by photocatalysts for water treatment [31,[57][58][59][60][61][62]. The photocatalytic degradation rate of SnO 2 prepared by the precipitation method, which compared with various photocatalysts used to remove dye from aqueous solutions was very promising, as shown in Table 4. The results suggest that using the RSM to optimize dye treatments could provide a better photocatalytic degradation rate than other studies. This optimization of photocatalytic processes using the RSM could be more favorably used in high concentration and short reaction time for dye removal. For the industrial applications of photocatalysts, the recycling and reuse of the SnO 2 are essential components of the practical implementation of the water treatment process [18]. To investigate the reusability of the prepared SnO 2 , we have selected the optimum conditions for the photocatalytic degradation of CR. At a CR concentration of 20 ppm, solution volume of 500 mL, operating temperature of 47 • C, catalyst dosage of 751 mg/L, and degradation reaction time of 97 min, five consecutive photodegradation experiments were conducted on the SnO 2 photocatalyst and the results are presented in Figure 10. As the number of cycles increased, the degradation rate remained at 100%, which indicates that the SnO 2 photocatalyst prepared in this research study has excellent reusability.

Preparation of SnO 2 by Precipitation
First, stannous chloride was placed in a 500 mL glass beaker to which 400 mL of deionized (DI) water was added and the solution was stirred with a magnetic stirrer (300 rpm). Synthesis was conducted at a rate of 1 mL of alkali (precipitant) added every 15 s until the pH reached 10. Subsequently, the solution was centrifuged in a 50 mL centrifuge tube to remove the supernatant, obtaining a white precipitate. After the precipitate was washed several times with 20 mL DI water, it was placed in an oven and dried at 50 • C for 24 h. The dried sample was ground into powder, the weight of which was recorded, and it was then placed in a glass sample bottle and stored in a dry box. Finally, the total amount of powder was placed in a crucible that was placed in a furnace for calcination at 600-700 • C for 2 h to obtain the SnO 2 photocatalyst.
The SnO 2 photocatalyst (final product after calcination) yield was calculated according to the following equation and used in the ensuing preparation experiments of SnO 2 : SnO 2 photocatalyst yield = number of SnO 2 moles number of stannous chloride moles × 100% (6)

SnO 2 Photocatalyst Characterizations
The surface morphology of photocatalysts was observed by a SEM (S-3000H, Hitachi Science and Technology, Tokyo, Japan). The elements of the SnO 2 photocatalyst were identified by EDX (Emaxx-act, Horiba, Kyoto, Japan). The size and morphology of SnO 2 particles were determined by employing TEM (H-7100, Hitachi, Tokyo, Japan). The XRD (Empyrean, Malvern Panalytical, Malvern, UK) was equipped with CuKα source (λ = 0.151418 nm) to understand the crystal structure of the sample under the following conditions: 40 kV/40 mA and 2θ range of 20 • -70 • . The specific surface area and pore size of the SnO 2 photocatalyst were analyzed using a BET measurements analyzer (Gemini™ V Series 2380, Micromeritics, Norcross, GA, USA). A GENESYS 10 Series spectrophotometer from Thermo Scientific (Waltham, MA, USA) was used for the observation of the UV-vis absorption spectra and the assessment of the photocatalytic degradation of the CR dye. The experiments were carried out in quartz cuvettes with 1 mm thickness.

Photocatalytic Degradation of CR Dye
In this experiment, before photodegradation, the CR dye contaminant solution (500 mL) and the photocatalyst were thoroughly stirred before being well dispersed with ultrasound by the ultrasonic cleaner (DC150, Delta Ultrasonic, Taiwan) for 1 min to disperse the catalyst evenly in the solution. Subsequently, they were placed in a dark room to enable the photocatalyst to adsorb the dye. When the adsorption reached equilibrium, we turned on the light source for degradation, which was a 400 W high-pressure mercury lamp with a main band of 365 nm. Magnetic stirring was performed at 400 rpm to conduct a photodegradation test for pollutants. During the degradation process, we used a syringe to draw 3.5 mL of solution at each sampling point, centrifuged the solution twice in a centrifuge (6000 rpm, 5 min), and then performed UV-vis to detect the absorbance value of CR and measure the change in its concentration. The dye degradation rate was calculated according to the following equation and used in the ensuing photocatalytic experiments: degradation rate (%) = C 0 − C C 0 × 100% (7) where C 0 and C are the dye concentration (mg/L) at initial time and at time t (min), respectively.

Process Variables and Experimental Design
For our experimental design, we used the software package Design Expert (Minneapolis, MN, USA) with the RSM as an extension method. The RSM combines mathematics and statistics and is the optimal tool for determining experimental design or operating parameters. First, we discussed the effect of single factors on the degradation rate of CR to determine the trend of each factor's degradation effect. To better understand the interaction of various factors, we employed the Box-Behnken design (BBD) which helps to explore the influence of the following three factors: X 1 : operating time (range 15-150 min), X 2 : operating temperature (range 30-60 • C), and X 3 : catalyst dosage (range 100-1000 mg/L). The BBD was composed of five central points and three-factor points, and its three levels were (−1, 0, +1). In total, 17 experiments were conducted. The experimental conditions and data are displayed in Table 2.

Conclusions
In the above research study, a SnO 2 photocatalyst was successfully prepared using the precipitation and calcination pathways method. The chemicals used in the synthesis process cause low harm to the environment and humans. Compared with other preparation methods, precipitation has various advantages, including speed, low cost, high purity, and higher yield. The results of the prepared process indicated that the photocatalyst yield of SnO 2 is affected by the precursor concentrations, synthesis temperature, and precipitant concentrations. Specifically, the effects of three single variables (initial concentration of pollutants, catalyst dosage, and operating temperature) on degradation efficiency were investigated for the CR dye. The degree of effects from high to low was as follows: operating temperature > catalyst dosage > initial concentration of pollutants. In the RSM, the interaction of the CR dye with three variables (reaction time, catalyst dosage, and operating temperature) was discussed. In the model simulation, only the interaction effects of B 2 (temperature) and BC (temperature, catalyst) was not noticeable and its model regression contribution was relatively small compared with the contribution of other factors. The prediction results of the model indicated that, when the degradation time was 97 min, the operating temperature was 47 • C, and the photocatalyst dosage was 751 mg/L, and the optimal degradation rate of the CR dye was 100%. We verified that the experimental values were the same as the predicted value of the model, which confirmed that the model was reliable. Finally, catalyst reusability was tested. After five consecutive photodegradation reactions, the degradation rate remained at the same level and the degradation efficiency was also 100%. This demonstrated that the SnO 2 photocatalyst prepared in this study has excellent reusability. The proposed precipitation method provides a simple and environmentally friendly preparation photocatalyst method. In the photocatalytic degradation of CR, through the RSM, these data could identify the optimal parameters for SnO 2 photocatalytic degradation and acknowledge the primary factors and two-factor interactions for water treatment applications.