Solar-light-driven LaFexNi1−xO3 perovskite oxides for photocatalytic Fenton-like reaction to degrade organic pollutants

Material characterization of various photocatalysts

X-ray powder diffraction (XRD) was used to reveal the structure of the materials. In the synthesis step of LaNiO3, the calcination temperature was set to 500, 600, 700, and 800 °C, respectively, and the samples were named LaNiO3-500, LaNiO3-600, LaNiO3-700, LaNiO3-800 in sequence. The uncalcined sample was noted as LaNiO3-NC. As shown in Figure 1, there was no crystalline LaNiO3 signal for the sample of LaNiO3-NC, but there were other signals for other samples indicating the presence of La2NiO4 (42.8°, JCPDS Card #011-0557), NiO (37.3°, 43.3°, JCPDS Card #04-0835), and La2O2CO3 (13.1°, 22.8°, 29.6°, 31.3°, JCPDS Card #23-0322) signals, respectively [40]. The sample calcined at 500 °C exhibited the more apparent signals belonging to La2O2CO3 and NiO. The signal of LaNiO3 did not appear for LaNiO3-500, indicating that the temperature of 500 °C was not enough to form LaNiO3. With the increase of temperature up to 600 °C, some obvious 2θ signals of LaNiO3 crystalline appeared, including 23.3°, 32.8°, 41.2°, 47.3°, 58.6°, and 68.8°. These peaks indicated the crystal planes (012), (110), (202), (024), (211), and (220) of LaNiO3, respectively, conforming to JCPDS Card #033-0711 [41]. LaNiO3-700 and LaNiO3-800 performed similar peak positions with higher intensity of signals. Accordingly, the crystal diameter of LaNiO3 was calculated by Scherrer's equation [42]. The crystal diameters of LaNiO3-600, LaNiO3-700, and LaNiO3-800 were 8.5 nm, 11.9 nm, and 14.6 nm, respectively. It could be seen that the higher the calcination temperature, the higher the crystal size will form.

[2190-4286-13-79-1]

Figure 1: The characteristic XRD patterns of LaNiO3 at different calcination temperatures.

Similarly, LaFeO3 samples with different calcination temperatures of uncalcined, 500, 600, 700, and 800 °C were noted as LaFeO3-NC, LaFeO3-500, LaFeO3-600, LaFeO3-700, and LaFeO3-800, respectively. In Figure 2, there was no signal for the uncalcined sample. While the calcination temperature was set at 500 °C, the signals of the LaFeO3 crystalline phase appeared. The diffraction peak of LaFeO3 became stronger as the calcination temperature increased. As shown in Figure 2, the 2θ peaks of 22.6°, 32.2°, 39.6, 46.3°, 57.4°, and 67.4° indicated the crystal planes of (101), (121), (220), (202), (240), and (242), according to JCPDS Card #037-1493 [43]. With the increase in calcination temperature, the higher crystal diameters of LaFeO3 calculated by Scherrer's equation were obtained. The crystal diameters was 18.5 nm, 25.4 nm, 29.1 nm, 35.0 nm for LaFeO3-500, LaFeO3-600, LaFeO3-700, LaFeO3-800, respectively, suggesting higher calcination temperature caused higher crystallinity for LaFeO3. Moreover, all LaFeO3 samples revealed higher crystal diameters than LaNiO3, indicating LaFeO3 tended to grow crystal than LaNiO3 at a certain calcination temperature. However, high temperature might cause particle aggregation, leading to the lower surface area. Therefore, a moderate temperature of 700 °C was selected to obtain the perovskite materials. On the other hand, the XRD of LaFeO3-800 indicated the appearance of Fe2O3 at 2θ peaks of 32.9°, 38.3°, 47.3° (JCPDS Card #39-0238) and La2O3 at 2θ peaks of 25.3°, 52.0°, 54.1° (JCPDS Card #40-1281). Especially, the peak of 32.9° for LaFeO3-800 was much clearer than that of LaFeO3-700, suggesting that the calcination temperature was too high and caused the formation of Fe2O3.

[2190-4286-13-79-2]

Figure 2: The characteristic XRD patterns of LaFeO3 at different calcination temperatures.

In addition, samples with various doping proportions of Fe/Ni, including 0.1/0.9, 0.3/0.7, 0.5/0.5, 0.7/0.3, and 0.9/0.1 were synthesized. The pH value was adjusted at 0 during synthesis, and the calcination temperature was fixed at 700 °C. The samples were named LaFeO3, LaFe0.9Ni0.1O3, LaFe0.7Ni0.3O3, LaFe0.5Ni0.5O3, LaFe0.3Ni0.7O3, LaFe0.1Ni0.9O3, LaNiO3 to represent the different Fe/Ni doping ratios. In Figure 3, the signals of LaFeO3 and LaNiO3 were identified by JCPDS Card #037-1493 [43] and JCPDS Card #033-0711 [41], respectively. Among these peaks, the main characteristic peak around 32° was slightly shifted for different Fe/Ni doping ratios, indicating that Fe and Ni were successfully doped into the structure of perovskite oxides. It is interesting to note that LaFe0.7Ni0.3O3 had a stronger signal than other doped samples. The crystal diameters of samples with various Fe/Ni doping ratios were also calculated by Scherrer's equation, as shown in Figure 4, which ranged from 9.5 to 31.3 nm. When the Fe/Ni ratio was manipulated at 7/3, the sample had the largest crystal diameter of 31.3 nm. The better crystallinity caused less recombination of electron-and-hole pairs, and subsequent reactions might occur more effectively [44]. On the other hand, The unit cell parameters and cell volume were also estimated from the XRD patterns and summarized in Supporting Information File 1, Table S2. Since the peaks of LaFe0.5Ni0.5O3, LaFe0.7Ni0.3O3, and LaFe0.9Ni0.1O3 were closed to LaFeO3, the unit cell parameters and interplanar spacing were calculated based on the model of orthorhombic LaFeO3 (JCPDS card: 037-1493) using Bragg’s law [45]. Meanwhile, LaFe0.5Ni0.5O3, LaFe0.3Ni0.7O3, and LaFe0.1Ni0.9O3 were also calculated based on the model of hexagonal LaNiO3. As a result, the lattice constant and cell volume were slightly expanded. It was interesting to note that LaFe0.7Ni0.3O3 revealed a relatively larger expansion, which had a higher cell volume, suggesting a better separation of photo-induced electron and hole pairs.

[2190-4286-13-79-3]

Figure 3: The characteristic XRD patterns of LaFexNi1−xO3.

[2190-4286-13-79-4]

Figure 4: The crystal diameters of samples with various Fe/Ni doping ratios.

The UV–vis absorption capability with diffuse reflectance spectroscopy (DRS) and photographs of various LaFexNi1−xO3 perovskite oxides with different proportions were shown in Figure 5a and 5b. Except for LaFeO3, which was brown, the rest of the perovskite oxides doped with Ni became black. Since pristine LaNiO3 was black, it exhibited the total absorption in the ultraviolet–visible light spectrum, consistent with the literature comparison [46]. For comparison, LaFeO3 revealed an apparent absorption shoulder between 500 and 600 nm in Figure 5a, similar to the previous study [47]. To enhance the light absorption of LaFeO3, it was an effective method to dope Ni into the perovskite oxides. Accordingly, the samples were doped with Ni to form LaFexNi1−xO3 perovskite oxides that could absorb the most visible and ultraviolet light spectrum. Thus, the prepared Ni-doped LaFeO3 perovskite oxides were presented as black, as shown in Figure 5b. They successfully increased the absorption efficiency of visible light and utilized more visible light effectively.

[2190-4286-13-79-5]

Figure 5: (a) The DRS spectrum and (b) the pictures of samples with various Fe/Ni doping ratios.

In order to determine the specific surface area, pore size, and pore volume of the prepared perovskite oxides, the analysis of nitrogen adsorption and desorption was performed. As shown in Supporting Information File 1, Figure S1, all LaFexNi1−xO3 perovskite oxides were in line with Type IV isotherm according to their hysteresis phenomenon. The Brunauer–Emmett–Teller (BET) result of the pristine LaFeO3 and LaNiO3 corresponded to the result in the literature [48,49]. The pore size distribution of the samples was shown in Figure S2. The distribution between 2 and 50 nm indicated that the prepared perovskite oxides were mesoporous. The summary of the specific surface area, pore size, and pore volume for all the samples with different Fe/Ni ratios was presented in Table S1. In Figure 6, it could be found that the LaFe0.7Ni0.3O3 with the Fe/Ni ratio of 7/3 had the highest specific surface area, pore volume, and pore size, suggesting there was more possibility for LaFe0.7Ni0.3O3 to adsorb and react with the molecules on the surface. On the other hand, the prepared samples might also be considered non-porous materials with inter-particle pore voids, since their low surface area might come from the external surface, indicating that LaFe0.7Ni0.3O3 had the highest external surface area for reaction.

[2190-4286-13-79-6]

Figure 6: Specific surface area, pore size, and pore volume of the samples with different Fe/Ni ratios.

Figure 7 revealed the field emission scanning electron microscopy (FESEM) images at the magnification of 100,000×, the surface of the samples with different Fe/Ni ratios were irregular and slightly different from each other. The grain surfaces of the pure LaNiO3 and LaFeO3 were chestnut-like. The observed appearance was also similar to the literature situation [50]. With the increased Fe content, some small lumps formed on the surface. When Fe/Ni ratio reached 7/3, relatively abundant small particles were generated than other samples. Small particles of LaFe0.7Ni0.3O3 would increase the surface area, which was consistent with the trend of the results detected by BET. On the other hand, it is interesting to note that the pH value during synthesis could affect the appearance of LaFexNi1−xO3. The samples prepared at pH 0 showed more uniform than that at pH 7 (shown in Supporting Information File 1, Figure S3). It indicated that the protons in the sol–gel solution could help the separation of LaFexNi1−xO3 crystals, leading to less particle aggregation. Moreover, the elemental analysis of the samples was also carried out using energy dispersive spectroscopic (EDS). The Fe contents of the samples with different Fe/Ni atomic ratios were identified. EDS detection showed that the synthesized samples exhibited the accurate Fe/Ni atomic ratios as designed. The detailed EDS data was provided in Table S3. The lanthanum, nickel, iron, and oxygen were analyzed from the samples, and the carbon was detected from the carbon tape.

[2190-4286-13-79-7]

Figure 7: The FESEM images of LaFexNi1−xO3 prepared at pH 0 (at the magnification of 100,000×).

MB removal test using various photocatalysts prepared at pH 0 or pH 7

The standard concentration of MB aqueous solution was prepared at 20 ppm. Various LaFexNi1−xO3 perovskite oxides prepared at pH 0 were examined for dark adsorption and photocatalytic degradation. First, dark adsorption was not significant for all perovskite oxides since their specific surface areas were low. After 30 min, to confirm the achievement of dark adsorption-desorption equilibrium, the photocatalytic reaction occurred under the irradiation of simulated AM 1.5G solar light with adding H2O2. By monitoring the C/C0 of MB, the performance of all perovskite oxides prepared at pH 0 or pH 7 were depicted in Figure 8a and 8b, respectively. LaFeO3 had the highest content of Fe3+ ions so that it could generate more hydroxyl radicals with H2O2 in the solution, resulting in the best performance of MB degradation. LaFeO3 could completely degrade MB after 45 min of simulated solar light irradiation due to the Fenton effect of Fe3+[51]. On the contrary, LaNiO3 did not conduct the Fenton-like effect; therefore, it exhibited poor photocatalytic ability. Fe's phenomenon revealed better catalytic activity than Ni, similar to the previous study [52]. It might result from Ni's apparent activation energy being higher than Fe's for producing oxidizing species [53]. Although Fe was attempted to be doped into LaNiO3, LaFe0.1Ni0.9O3 and LaFe0.3Ni0.7O3 still exhibited low photocatalytic capability. Until Fe doped amount was up to 50% for replacing Ni, the photocatalytic performance of LaFe0.5Ni0.5O3 was enhanced much more apparent than that of LaFe0.3Ni0.7O3. It indicated that Fe3+ plays a vital role in involving MB degradation. As the Fe doped content increased, LaFe0.7Ni0.3O3 reached the highest photocatalytic capability, originating from its largest surface area and crystal diameters. A larger surface area would enhance the surface reaction with the aqueous solution, and larger crystal diameters could decrease the possibility of recombining electron-and-hole pairs. However, while the Fe doped content was set at 90%, replacing Ni, the photocatalytic capability was reduced due to the lower surface area and smaller pore size. The above inference was consistent with the cases of the samples prepared at pH 0 and pH 7. Interestingly, the performance of the perovskite oxide prepared at pH 0 was much better than that prepared at pH 7. It was derived from the fact that more uniform structural features of the perovskite oxide prepared at pH 0 were achieved than the condition at pH 7. Moreover, the photocatalytic performance of physically mixed 70% LaFeO3 and 30% LaNiO3 could be estimated based on the result of MB degradation using 70% LaFeO3 (since LaNiO3 showed no MB degradation). The degradation of 70% LaFeO3 in 30 min was approximately 60.0%. Considering the light shading effect by LaNiO3, the MB photodegradation of physically mixed 70% LaFeO3 and 30% LaNiO3 might be lower. On the contrary, the LaFe0.7Ni0.3O3 showed 78.5% of MB degradation in 30 min. Therefore, we believed there would be a benefit of doping Ni to improve the photodegradation.

[2190-4286-13-79-8]

Figure 8: MB degradation experiments using various LaFexNi1−xO3 with different Fe/Ni ratios prepared at (a) pH 0; (b) pH 7.

In Figure 9, by taking the negative natural logarithmic value of C/C0, it was observed that all the degradation trends were in line with first-order kinetics:

[Graphic 1]

where k was the rate constants of MB degradation reaction; t was the reaction duration in min. As shown in Table 1, for photocatalytic Fenton-like reaction to decompose MB dye, LaFeO3 had the fastest degradation rate (k = 0.1072 prepared at pH 0; k = 0.0086 prepared at pH 7). It might be due to the higher content of Fe ion for LaFeO3 than LaFe0.7Ni0.3O3. However, we would like to focus on the doping effect on photocatalytic reaction so that Fe0.7Ni0.3O3 was the target material to be further analyzed and characterized. Comparing the samples co-doped with Fe and Ni, LaFe0.7Ni0.3O3 exhibited a higher k value of 1st order reaction than other co-doped samples. LaFe0.7Ni0.3O3 had a larger crystal diameter and higher specific surface, which improved the separation of photogenerated charge carriers and the efficiency of the surface reaction. For comparison, the second-order kinetics analysis was also conducted for the samples prepared at pH 0 in Table 1. However, the R2 values were too low to represent their kinetic model. Therefore, 1st order reaction kinetics was more suitable for describing the kinetic model of LaFexNi1−xO3.

[2190-4286-13-79-9]

Figure 9: Kinetic analysis of MB degradation experiments using various LaFexNi1−xO3 with different Fe/Ni ratios at (a) pH 0 and (b) pH 7.

Table 1: The kinetic analysis of LaFexNi1−xO3 perovskite oxides prepared at pH 0 and pH 7.

pH value Reaction kinetics analysis LaFeO3 LaFe0.9Ni0.1O3 LaFe0.7Ni0.3O3 LaFe0.5Ni0.5O3 pH 0 1st order k 0.1072 0.0339 0.0506 0.0254 R2 0.9616 0.9808 0.9584 0.9876 2nd order k 0.8043 0.0313 0.4098 0.0125 R2 0.6892 0.4846 0.5404 0.6184 pH 7 1st order k 0.0086 0.003 0.0039 0.002 R2 0.9903 0.9691 0.9838 0.9517
MB removal test under different conditions using LaFe0.7Ni0.3O3 prepared at pH 0

The pH value of the solution was a strong effect on photocatalytic degradation [54]. Accordingly, different pH values of solution using LaFe0.7Ni0.3O3 perovskite oxides prepared at pH 0 were examined for photocatalytic degradation. Thus, the MB aqueous solution was adjusted to pH 1.5, 3.5, and 5.5. The performance of the photocatalytic Fenton degradation was measured at different pH values in Figure 10. After 120 min of simulated solar light irradiation, the degradation performance at 1.5, 3.5, and 5.5 were 97.9%, 100%, and 25.2%, respectively. The pH value at 3.5 revealed the highest photocatalytic performance, in which MB pollutants were completely degraded within 105 min. Based on the 1st order kinetic analysis, the reaction rate constants (k) at pH 1.5, 3.5, and 5.5 were 0.0254, 0.0506, and 0.002, respectively. While the pH value was too high, the hydrogen peroxide in the solution was easily decomposed into oxygen and water [55]. On the other hand, when the pH value was too low, the solution tended to generate too many hydrogen ions, which would react with hydrogen peroxide to form water, leading to a decrease in the photocatalytic degradation performance [56]. Therefore, the operating condition of the pH value was set at 3.5.

[2190-4286-13-79-10]

Figure 10: (a) The C/C0 and (b) 1st order kinetic analysis of the MB photodegradation using LaFe0.7Ni0.3O3 operating at different pH values.

After manipulating the operating pH value, the pH of the solution was fixed at 3.5. The dosages of LaFe0.7Ni0.3O3 perovskite oxides were further examined. In this study, the amount of the MB solution was 150 mL, and various LaFe0.7Ni0.3O3 perovskite oxides, including 30 mg (200 ppm), 50 mg (333 ppm), 80 mg (533 ppm), and 100 mg (666 ppm) of catalyst were put into the reactor for photocatalytic degradation reaction. As shown in Figure 11a, the degradation performances after 120 min photocatalytic reaction were 55.7%, 100%, 98.8%, and 97.3%, respectively. Comparing the conditions of 30 mg (200 ppm) and 50 mg (333 ppm), it could be found that if more LaFe0.7Ni0.3O3 perovskite oxides were added to the solution, a better photocatalytic degradation performance would be obtained. However, while too many LaFe0.7Ni0.3O3 perovskite oxides were added, the excess catalyst would inhibit light penetration into the suspension. Thus, the availability of light energy was reduced, resulting in a reduction of the photocatalytic degradation capability [30]. Accordingly, the degradation performance was calculated by the 1st order kinetics, as shown in Figure 11b. It could be known that when the catalyst dosage was 30 mg (200 ppm), 50 mg (333 ppm), 80 mg (533 ppm), and 100 mg (666 ppm), their reaction rate constants k were 0.007, 0.0242, 0.0314, and 0.0506, respectively. Therefore, the pH of the solution was 3.5, and the catalyst dosage was 50 mg for the following examination.

[2190-4286-13-79-11]

Figure 11: (a) The C/C0 and (b) 1st order kinetic analysis of the MB photodegradation using LaFe0.7Ni0.3O3 operating at different catalyst dosages.

Moreover, the amount of hydrogen peroxide added was manipulated in advance. Similarly, the amount of contaminant solution was 150 mL. The pH value of the solution was fixed at 3.5, the catalyst dosage was 50 mg, and hydrogen peroxide was added in various amounts, such as 5 µL (33 ppm), 10 µL (66 ppm), 20 µL (133 ppm) and 60 µL (400 ppm), respectively. The photocatalytic degradation experiment results were shown in Figure 12a. With the increase of hydrogen peroxide addition, the degradation capability after 120 min simulated solar light irradiation was 69.4%, 97.2%, 100%, and 99.9%, respectively. Then, their 1st order kinetics analysis was also depicted in Figure 12b. The reaction rate constants k of various conditions by adding different amounts of H2O2 were obtained as 0.0086, 0.0251, 0.0326, and 0.0506, respectively. It could be concluded that adding more H2O2 could increase the hydroxyl radicals in the aqueous solution and improve the performance of photocatalytic degrading pollutants. However, when too many hydroxyl radicals were generated, the hydroxyl radicals in the aqueous solution would combine with excess H2O2 to form superoxide hydrogen radicals and water [56], reducing the degradation capability.

[2190-4286-13-79-12]

Figure 12: (a) The C/C0 and (b) 1st order kinetic analysis of the photodegradation using LaFe0.7Ni0.3O3 operating at different H2O2 addition.

Accordingly, the optimal operating conditions were summarized together: (1) the pH value of the aqueous solution was 3.5; (2) the amount of catalyst added was 50 mg in 150 mL solution (333 ppm); (3) the amount of hydrogen peroxide added was 20 µL. The MB removals contributed to the photocatalytic reactions and the Fenton reactions simutaneously. In order to further understand the effect of the photocatalysis and the Fenton reaction in the degradation reaction, some degradation tests were also carried out as controlling experiments for comparison, including: (1) without adding photocatalyst (No catalyst), (2) without light (No light), and (3) without adding H2O2 (No H2O2), as shown in Supporting Information File 1, Figure S4a. It could be observed that in the case of no catalyst added, the degradation was only 2.2% after 120 min, suggesting that H2O2 might not be easy to form ∙OH radicals under visible light irradiation to carry out the degradation reaction. Then, the degradation of the case without adding H2O2 was approximately 4.9% after 120 min. It indicated that the heterogeneous Fenton reaction might play a more critical role than the photocatalytic reaction. Another possibility, even though LaFe0.7Ni0.3O3 was a visible-light-driven photocatalyst, its electron–hole recombination was still too severe, resulting in relatively poor performance of MB removal. Next, the result of the case without light indicated the Fenton reaction was carried out without light. After 120 min, it was found that the degradation reached 69.2%. LaFe0.7Ni0.3O3 catalyst contained iron ions, which could certainly convert H2O2 into ∙OH and achieve a certain degree of degradation. Finally, compared with the above-mentioned controlling cases, the degradation under original conditions (All) reached 100% at 120 min, attributed mainly to the Fenton reaction and the photocatalysis, which could multiply the function of hydroxyl radicals. Their performances of MB degradation were calculated by 1st order kinetics, as shown in Figure S4b. The reaction rate constants k of different conditions: (1) No cat., (2) No H2O2, (3) No light, and (4) All, were 0.0002, 0.0005, 0.0096, and 0.0506, respectively. Moreover, the comparison of the prepared LaFe0.7Ni0.3O3 with other materials from the literature was also listed in Table S4. It showed that Fe0.7Ni0.3O3 samples revealed comparable photodegradation performance to other composite materials.

Moreover, the degradation of TC using LaFe0.7Ni0.3O3 prepared at pH 0 was also carried out under the optimal conditions of pH 3.5, catalyst = 333 ppm, and H2O2 = 20 µL as shown in Supporting Information File 1, Figure S5a. The TC concentration could drop by nearly 93.3% in 30 min of light irradiation compared with the initial concentration (20 ppm). The TC was degraded entirely in 60 min. Since there were multiple polar groups (hydroxyl group) within the chemical structure of TC [57], it was easy to attract the hydroxyl radicals in the TC solution, resulting in a better photocatalytic performance than MB. The kinetic rate constant of TC degradation was also calculated, as shown in Figure S5b. Accordingly, the reaction rate constants k of photodegrading TC and MB were 0.10991 and 0.0506, respectively, indicating LaFe0.7Ni0.3O3 revealed a great potential to decompose other organic pollutants.

留言 (0)

沒有登入
gif