Synthesis of BiFeO₃/ZnMgAl-LDH nanocomposite for instantaneous decolorization of methylene blue by ultrasonic induced adsorption system: characterization and equilibrium studies

The following chemicals were applied without extra purification: For the preparation of BFO, Fe(NO₃)₃.9H₂O (99%), Bi(NO₃)₂.5H₂O (99%), ethylene glycol (C₂H₆O₂), and nitric acid (HNO₃, 68 wt%) were purchased from Sigma Aldrich. For the fabrication of LDH, Zn(NO₃)₂.6H₂O (98%), Al(NO₃)₃.9H₂O (98%), Mg(NO₃)₂.6H₂O (98%), potassium carbonate, and KOH (98%) were supplied by Loba Chemi Co. The pollutant solution of MB was prepared by dissolving a proper weight in deionized water.

BFO was prepared according to Safizade et al [41]. The details of the preparation method are as follows: firstly, solution A was formed by liquefying 16 mmol of ferric nitrate in ethylene glycol. Solution B was created by liquefying 16 mmol of bismuth nitrate in nitric acid. Afterward, solution B is gradually added to A, evolving brown bubbles of nitrogen oxide, indicating a redox reaction between NO₃ and O–H of ethylene glycol. The precipitate was drying at 130 °C and annealed in air at 600 °C for 60 min. The sample was washed with acetic acid and deionized water and then dried at 70 °C to remove the impurities.

ZnMgAl LDH was synthesized by the co-precipitation method. The metal salts (0.326 gm zinc nitrate hexahydrate, 1.95 gm magnesium nitrate hexahydrate, and 1.08 gm aluminum nitrate nonahydrate) were liquified in 30 ml D.I.W. to form solution A (based on cations). Solution B (1.8 g potassium carbonate, 1.7 g potassium hydroxide) in 40 ml deionized water Solution A was dropped by solution B under continuous stirring while maintaining a pH of 10. The resulting slurry was stirred vigorously for 18 h at 80 °C. The precipitated LDH was harvested with a centrifuge and washed several times with a mixture of distilled water and ethanol to remove excess ions. Then dried at 70 °C for a day.

BFO/LDH nanocomposite was fabricated via the in situ co-precipitation method using similar initial procedures as LDH. About 0.5 g of BFO was weighed into 50 ml of D.I.W. and sonicated for 30 min. After adjusting the pH of the LDH suspension, BFO solution was slowly added to it and kept for 18 h with vigorous stirring at 80 °C. The obtained brick-colored product was filtrated. Subsequently, the product was cleaned numerous times with D.I.W. and dried at 70 °C. The exact percent of BFO loading is 12%.

The structure and composition of the prepared nanoparticles were studied by x-ray diffraction (XRD) using a Philips diffractometer in the range 10°–80° 2θ by Cu Ka radiation. The surface area (BET) and related results were determined by N₂ sorption using a NOVA 3200 S unit. Fourier transform infrared (FT-IR) spectra were achieved using Nicolet IS-10 in the range of 4000–500 cm⁻¹. The Raman analysis was measured by a Bruker-Senterra spectrometer. The TEM of the prepared composites was captured using a JEOL-JEM-2100 working at 200 kV. XPS analysis was verified by XPS SPECS GmbH, Germany.

The ultrasound effect of factors such as ultrasonic time (3–40 min), adsorbent dose (0.3–2 g l⁻¹), pH⁻¹ (4–12), and initial dye concentration (20–200 mg l⁻¹) on MB dye removal was investigated. The optimal experimental conditions were determined. Typically, adsorption experiments are carried out using an ultrasonic bath (Model UC-30A, Frequency: 40 KHz, Capacity: 6 l, Power: 400 W, Input: AC 250 V, 50 Hz). The adsorption process was carried out in a double-jacket beaker to keep the temperature constant during the adsorption process. After the adsorption process, magnetically, the catalyst separated by centrifugation at adjusted speed of 5000 rpm for 10 min and the remaining dye was examined using a UV–vis spectrophotometer. The kinetics and isotherm adsorption were evaluated by measuring the dye's adsorptive removal at varied initial dye concentrations and time intervals under optimal parameters. The MB adsorption removal and the adsorption capacity (A.C.) under various experimental settings were measured by equations (1) and (2).

$$\begin{eqnarray} &&\% \,{\rm{MB}}\,{\rm{adsorption}}\,{\rm{removal}}=\displaystyle \frac{{c}_{i}-{c}_{f}}{{c}_{i}}\times 100\end{eqnarray} \tag{ 1 }$$

$$\begin{eqnarray}&&{{\rm{q}}}_{{\rm{t}}}=\displaystyle \frac{{\left({c}_{i}-{c}_{f}\right)}^{V}}{W}\end{eqnarray} \tag{ 2 }$$

Where C_i and C_f denote the initial and final concentrations of the MB dye at time t, respectively; q_t represents the amount of dye adsorbed at time t; V is the dye solution volume used in the test, and w means the adsorbent amount.

The optimization of the removal processes of MB dye by BFO/LDH was assessed based on statistical design, considering the impact of time interval, pH, and BFO/LDH dosage as the essential inputs or variables using Design-Expert Software (Version 6.0.5). The design was built considering the properties of the central composite statistical design (CCD), which is associated with Response Surface Methodology (RSM). The upper and lower levels of the affecting variables were added based on the previously addressed experimental results and listed in table 1 in their coded and actual values. The impact of the evaluated input, as well as the interaction between them, was illustrated based on the second-order polynomial equation considering the removal percentages of MB as the target response (equation (3)). The inserted Y, Xi, and Xj, β0, βi, βii, βij, and K symbols donate the target response (removal of MB., %), input variables, constant, linear coefficient term, quadratic coefficient term, cross-product coefficient term, and the number of variables, respectively.

$$\begin{eqnarray}&&Y={\beta }_{o}+\displaystyle \sum _{i=1}^{k}{\beta }_{i}{X}_{i}+\displaystyle \sum _{i=1}^{k}{\beta }_{{ii}}{X}_{i}^{2}+\displaystyle \sum _{i=1}^{k-1}\displaystyle \sum _{j=i+1}^{k}{\beta }_{{ij}}{X}_{i}{X}_{j}\,\end{eqnarray} \tag{ 3 }$$

Figure 1 illustrates the XRD pattern of MgZnAl-LDH, BFO, and BFO/ZnMgAl-LDH nanocomposite. The XRD of LDH demoinstarte its characteristic peaks at 2θ values of 11.47° (003), 22.99° (006), 34.71° (012), 39.2° (015), 46.9° (018), and 62.05° (113). The strong diffraction peaks at 11.47 demonstrate the high purity of prepared LDH [42, 43]. XRD results revealed that the prepared LDH presented good crystallinity. The sharp and powerful peaks of ZnMgAl-LDH are attributable to LDH produced under a high pH environment, enhancing crystallinity. This result is in agreement with the previous study [44].

Zoom In Zoom Out Reset image size

The XRD of BFO at 600 °C sintering temperature shows typical diffraction peaks at 2θ values (22.3° (102), 31.7° (104), 31.9^o (110), 38.3^o (006), 39.5^o (202), 45.5^o (024), 51.4^o (116), 51.9^o (122), 56.3^o (018), 56.9^o (214)) that are ascribed to the phases of BiFeO₃ [45], and one peak attributable to Bi₂Fe₄O₉ (2θ = 27.8^o). The secondary phase (Bi₂₅FeO₄₀) was eliminated by nitric acid (HNO₃) [46]. It is reported that most BiFeO₃ is accomplished by secondary phases such as Bi₂Fe₄O₉ and Bi₂₅FeO₂₉. The impurity phases were attributed to the volatilization of Bi³⁺ ions, resulting from the excess addition of bismuth. The high pressure of the bismuth atom tends to facilitate the easy evaporation of the Bi⁺³ ion during synthesis conditions such as drying and calcination. A similar trend was observed in the previous study [47–49].

When comparing pure BFO to BFO/LDH composite, the overall peaks move to the right, the strength of twin peaks of planes (104) and (110) was reduced, and the dual peaks were somewhat split and shifted to a higher angle (2θ = 32.7^o and 33.6^o) compared to pure BFO. The peak intensity of plane (003) at 2θ = 11.4^o was reduced and marginally shifted to the right, revealing that composite formation reduced the crystallinity of LDH.

The FT-IR spectra of synthesized materials are presented in figure 2. At the LDH spectra, a broad peak at 3000–3600 cm⁻¹ is assigned to the stretching vibration of the hydroxyl group (O–H), while a peak at 1620–1640 cm⁻¹ is attributed to the bending vibration of the interlayer water molecule. The band at 1362–1375 cm⁻¹ may be ascribed to the stretching vibration of the interlayer carbonate anion [44]. The peak at 1010–1015 cm⁻¹ is related to the deformation mode of Al–OH. Moreover, the adsorption band at 732–532 cm⁻¹ is associated with M–OH vibration. Bands at 432–558 cm⁻¹ can be related to M–O and M–O–M stretching vibrations [50, 51]. The broad band of BFO located at 3000–3500 cm⁻¹ is attributed to stretching O–H, but at 1650 cm⁻¹ corresponds to O–H bending vibration. The bands at 400–650 cm⁻¹ were related to Fe–O stretching and O–Fe–O bending vibrations of the FeO₆ species in BFO nanoparticles [52]. The FT-IR spectrum of the composite matches well with the peaks of BFO and LDH. It indicates the coexistence of BiFeO₃ and ZnMgAl LDH phases in the BFO/LDH composite.

Zoom In Zoom Out Reset image size

Raman analysis is a valuable technique that gives information about the material structure. Figure 3 shows the Raman analysis for LDH, BFO, and BFO/LDH composites. BFO should have active Raman modes (A1 and E) representing lattice distortions of prepared BFO. Where A1 is the longitudinal optical mode and E is the transverse optical mode (TO). The E modes are located at the higher cm⁻¹ (212, 316, 380, 450, and 537 cm⁻¹) and can be ascribed to vibration modes of octahedral Fe O₆. But the modes related to Bi-O are located below 200, at 97 and 120. The position of Raman peaks is in good agreement with the previous study [53–55], and the results are consistent with XRD results.

Zoom In Zoom Out Reset image size

Raman spectra of LDH show a sharp band at approximately 1056, which was ascribed to the symmetric vibration of carbonate CO₃ ⁻² that bonded to surface OH, but the lower intensity band at 1079 cm⁻¹ is related to the symmetric vibration of CO₃ ⁻² bonded to interlayer water [56, 57]. The composite's Raman spectra also demonstrated a band at 1324 cm⁻¹ that corresponded to the carbonate anion on the external surface of LDH [58], which is similar to the prepared LDH, as well as vibration modes at 534, 446, 316, 218, 177, and 135 cm⁻¹ that corresponded to BFO, confirming that the BFO/LDH composite was successfully synthesized.

The BET isotherm was carried out to characterize the surface area, pore diameter, and total pore volume. According to the IUPAC classification, the LDH isotherm belonged to type IV isotherms with an H3 hysteresis loop, as described in figure 4 [59]. Pore size distribution of LDH suggests that LDH is made up of mesopores with a BET of 36 m² g^–1, an average pore diameter of 24 nm, and a pore volume of 0.22 cc/g. The BFO isotherm follows the type IV isotherm with a specific surface area of 11.9 m² g^–1. These results agree with those described in previous literature [60]. When BFO nanoparticles were incorporated into LDH, the surface area, pore volume, and average pore diameter of the composite decreased compared to those of LDH. This might be due to blockage of LDH pores in the nanocomposite, as seen in table 2.

Zoom In Zoom Out Reset image size

The structure of the prepared materials was confirmed by transmission electron microscopy (TEM) and scanning electron microscopy (SEM), as shown in figure 5. A TEM image of LDH demonstrated that LDH has hexagonal plate-like particles (the typical morphology of co-precipitated LDH), confirming the successful synthesis of LDH nanosheets with lateral size. The TEM image of BFO reveals irregular spherical and rectangular-shaped nanoparticles with particle sizes ranging from 38 nm to 105 nm. Also, there is an accumulation of BFO nanoparticles due to the high temperature of calcination. For comparison, the TEM image of the composite exhibits morphological change where the presence of BFO is visible on the layered structure of LDH and represented by dark spots randomly dispersed over the light gray plate-like LDH. The SEM of the BFO/LDH composite is shown in figure 5(d) and exhibits the aggregation of uneven sheet-like nanoparticles. The homogenous distribution of Mg, Al, Bi, Fe, Zn, and O that constitute BFO/LDH was validated by nanocomposite mapping and EDX. The SEM, mapping, and EDX confirmed the formation of the prepared composite.

Zoom In Zoom Out Reset image size

XPS was used to explore the chemical state of component elements and the BFO/LDH composite catalyst composition. The XPS profile of prepared materials is shown in figures 6(a)–(f). The BFO/LDH comprises O, C, Al, Mg, Zn, and Bi components. O 1s of BFO has one peak at 532.01 eV, corresponding to binding energy in O-M. C1s have three peaks at 284.9 eV for the C–C bond, 286.2 eV for the C–O bond, and 288.9 eV for the C-OO bond. The profile of Al 2p has a peak at 74.6 eV related to Al-O. The Mg 1s have two peaks at 1305.04 and 1303.8 eV, corresponding to Mg-O and Mg metal, respectively. Zn 2p has two binding energies at 1045.27 and 1022.15 eV, attributed to Zn 2p_1/2 and Zn 2p_3/2, respectively. The profile of Bi4F exhibits two peaks at 159.1 and 164.4 eV that are attributed to Bi 4F_7/2 and Bi 4F_5/2, respectively, and this confirms that Bi is in the Bi⁺³ oxidation state [61]. This result indicates that BFO is successfully supported on LDH.

Zoom In Zoom Out Reset image size

Figure 7(a) shows the ultrasonic time effect on the percentage of MB adsorption. For both LDH and BFO/LDH, the percent MB adsorption removal increases dramatically within the first 10 min, with a slight change in dye removal. Shortening the time to reach adsorption equilibrium is due to the effect of ultrasound waves. The first reason is that the high pressure produced by ultrasound wave radiation in water creates bubble cavitation. Then bubbles collapse violently on the catalyst surface. Consequently, the boundary layer thickness on the catalyst solid surface has been reduced, and mass transfer has improved [62, 63]. The second reason is that ultrasound wave radiation enhances MB molecules diffusion to the adsorption sites available at this point [67, 68]. As a result, a higher number of dyes binding to the surface occurs. Within 20 min, 57.3 and 77.5% of removal were obtained for LDH and BFO/LDH, respectively. It can be shown that the modification of LDH with BFO increases the percentage of MB removal due to the many functional groups on the composite surface compared to LDH. In the current study, 15 min is considered the optimum contact time since the difference in percent removal of methylene blue using a longer time (20 min) is just 2%.

Zoom In Zoom Out Reset image size

The adsorbent amount is a significant parameter that disturbs the adsorption process. A series of experiments are tested with doses in the range of 0.3–2 g l⁻¹ of MB (50 mg. l⁻¹). As shown in figure 7(b), the removal of MB increases for LDH and BFO/LDH as the adsorbent dosage increases from 0.3 g l⁻¹ to 1.0 g l⁻¹ . However, after 1.0 g l^–1 adsorbent weight, the MB dye molecules in the medium were insufficient to fully connect with all adsorption active sites on the composite surface, resulting in an equilibrium state.

The influence of pH on percent MB removal and adsorption capacity A.C. was investigated throughout a pH range of 4 to 12. As shown in figures 7(c) and (d), increasing the pH value improved both the percent of MB removal rate and A.C. The significant adsorption rise with pH increase can be correlated to the surface charge and the accessibility of binding sites on the surface of the LDH composites [32]. The above results indicate that the highest MB adsorption by LDH or BFO/LDH was achieved at pH 12.

The validation and significance properties of the design and its descriptive model (the second-order quadratic polynomial model) were assessed based on the dep inspection of the variance function (ANOVA). The statistical design suggests 16 suggested random tests considering the number of studied variables (three variables (A. time, B. pH, and C. dosage) (table 6).

The fitting degree of the linear regression relationship between the theoretically predicted removal percentages of MB and the experimentally detected values demonstrates significant agreement between them, with a determination coefficient (R²) higher than 0.98 (figure 12(A)). Therefore, the suggested experiments, according to the statistical assumption of the quadratic polynomial model, exhibit significant properties and can be addressed to describe the effect of the experimental factors on the removal of MB dye as well as the expected interaction effects between these variables. Moreover, the recognized prediction deviation curve of the 16 experiments reveals the remarkable regular distribution of the plotted points above and below the reference line within the range from −1.34 up to 1.5 (figure 12(B)). Additionally, the presented linear regression relation of the studentized residuals reflects the high accuracy and normality properties of the selected model during the prediction processes for the effects of the studied variables (figure S2).

Zoom In Zoom Out Reset image size

The analysis of ANOVA functions considering the values of model-F, model-Prob > F, lack of fit, and the sum of squares gives significant indications about the validation of the studied models and the assessed variables as inputs (table 7). The observed values of model F are 56.5, demonstrating the high significance of the applied polynomial model during the prediction processes with less than 0.01% chance of the presence of noise effects (table 7). The recognized model-Prob-F values of the three assessed variables are lower than 0.05%, which validates their significant properties as experimental factors. Moreover, this reveals the nonlinear relations between the studied variables as inputs and the determined removal percentages of MB as a response (table 7). Additionally, the considerable agreement between the estimated values of Pred R-Squared (0.84), Adj R-Squared (0.96), and Adeq Precision (30.49) reflects the adequate signal of the removal processes of MB by BFO/LDH and the significance of the applied polynomial model to navigate the space of the statistical design (table 7).

By Considering the results of the variances analysis of the approaches and the validation of the studied model, the obtained relations between the studied variables as inputs and the determined removal percentages of MB as response can be described according to the polynomial regression equation (equation (9).

$$\begin{eqnarray*}&&Y=+85.05+5.22X\,A+5.58X\,B+5.98X\,C\,\end{eqnarray*}$$

$$\begin{eqnarray*}&&-3.63X\,A2+1.47X\,B2-0.23X\,C2\end{eqnarray*}$$

$$\begin{eqnarray}&&+0.66X\,A\,X\,B+0.14X\,A\,X\,C-0.11X\,B\,X\,C\,\end{eqnarray} \tag{ 9 }$$

7.2.1. The time interval and pH

The effect of the adsorption time interval of MB (50 mg l⁻¹) as a function of the adjusted pH of the solutions was studied considering the applied dosage of BFO/LDH at the intermediate value (0.75 g l⁻¹). The interaction relationship between the solution pH and the adsorption interval as well as their impacts on the removal percentages of MB were assessed based on the obtained 3D regression curve (figure 13(A)). The detectable trends demonstrate a significant enhancement in the effect of the adsorption intervals on the removal percentages of MB with a regular increase in the solution pH from pH 8 up to pH 12. At pH 8, the increase in the time intervals by 5 min, 10 min, and 15 min resulted in an increment in the MB removal percentages by 72.7%, 80.3%, and 81.8%, respectively. By adjusting the solution pH to pH 10, the removal percentages enhanced to 77.4% (5 min), 84.2% (10 min), and 86.3% (15 min). At the upper level of the studied pH (pH 12), the MB removal percentages enhanced significantly up to 82.5% (5 min), 93.6% (10 min), and 95.3% (15 min). This behavior was previously assigned to the impact of the high alkaline conditions on inducing the electrostatic attraction of the dissolved MB molecules on the deprotonated surface of BFO/LDH.

Zoom In Zoom Out Reset image size

7.2.2. The time interval and BFO/LDH dosage

7.2.3. The solution pH and BFO/LDH dosage

Based on the optimization function of the used software (Design-Expert software) as well as the prediction option according to quadratic polynomial programming, the suggested or predicted solutions to optimize or enhance the removal percentages of MB by BFO/LDH are presented in table 8. The suggested experimental conditions and their related MB removal percentages were estimated considering the upper and lower levels of the three test variables. The suggested solution as well as the determined removal percentages are similar to the experimentally detected optimum conditions (15 min contact time, 1 g l⁻¹ dosage, and pH 12). The predicted condition suggested a neglected reduction in the time interval, used dosage, and solution pH. Additionally, the suggested enhancement in the removal percentage is of neglected value and cannot be considered a remarkable enhancement. This can reflect the successful detection of the best conditions experimentally.