3.1 Characterization of Zn doped Fe3O4 octahedrons
The XRD patterns of the pure Fe3O4 and octahedral Zn doped Fe3O4 samples are exhibited in Fig. 1. As shown in Fig. 1a, both the diffraction peaks of Zn doped Fe3O4 and pure Fe3O4 are almost the same. Both the patterns show that the reflection peaks ascribed to the crystal planes, (220), (311), (400), (422), (511), and (440), respectively, which matches well with the standard patterns of cubic spinel structure (JCPDS 77-1545). Observing the patterns, no impurity diffraction peaks of ZnO appeared, suggesting that crystalline ZnO did not form during the synthesis process. However, it can be seen that the (311) crystal plane of Zn-doped Fe3O4 slightly shifted to a smaller angle and the corresponded lattice constant changed from 0.8386 nm to 0.8402 nm (as depicted in Fig. 1b). This change could be ascribed to the substitution of a small amount of Fe2+ (ion radius = 0.61 nm) and Fe3+ (ion radius = 0.49 nm) in the magnetite. The phenomena could relate to Zn2+ with a larger ion radius of 0.74 nm. The results indicate that Zn has been successfully doped into the crystal structure of magnetite. Herein Fe3+ ion and a small amount Fe2+ ion occupied in tetrahedral and octahedral sites could be substituted by Zn2+ ion (Yang et al., 2013; Zélis et al., 2013).
Raman spectra analysis was conducted to further investigate the phase purity and degree of crystallization for the as-prepared Zn doped Fe3O4. Raman spectroscopy was conducted at excited 633nm with 20% and 40% power as shown in Fig. 2. The Raman spectra of the samples excited at 633 nm with 20% power depicted at Fig. 2a. The band appeared at 670 cm-1 related to the characteristic peak of magnetite, which agrees with a symmetric Fe − O breathing A1g mode (D.L.A. de Faria, 1997; Jan Kaczmarczyk). It is clearly seen that no characteristic signal of ZnO is observed, which corresponds with the results of XRD. Figure 2b exposes the Raman spectra of the as-synthesized product excited at 633 nm with 40% power. It is worth mentioning that the significant phase changes appeared at the surface of the samples when the samples were excited at 633 nm with 40% power. A few new bands observed at 220, 286, 397, 486, 604, 656 and 722 cm-1 could relate to α–Fe2O3. The peaks at 220 cm-1 and 486 cm-1 are ascribed to the A1g modes. The five peaks observed at 286, 397, 604, and 656 cm-1 are in accordance with the Eg modes. The signal at 722 cm-1 could be attributed to the feature of γ-Fe2O3 and the signal at 673 cm-1 corresponds to the feature of Fe3O4 (D.L.A. de Faria, 1997; Jan Kaczmarczyk; K. S. K. Varadwaj, 2004). The signal at 892 cm-1 agrees with vibrations of oxo-bridge di-iron O − Fe − O bonds at intermediate phase (Amine Mezni, 2013). There is no band ZnO in the Raman spectra that usually appear at about 436 cm− 1(Prakash Chand, 2015; V. Pazhanivelu, 2015). These results demonstrate that the magnetite phase was reduced under laser irradiation of 633 nm with 40% power (Guo et al., 2011). It is worth noting that the symmetric stretch of Zn-O in Raman mode of ZnFe2O4 is at about 650 cm-1 (Thota et al., 2016), which was not observed in both the spectra excited at 633 nm with 20% and 40% power, indicating the Zn doped into Fe3O4 successfully.
3.2 Morphology and crystalline structure of as-obtained mesocrystal
The morphology of the as-prepared Zn doped Fe3O4 was observed by SEM, TEM, SAED technique as shown in Fig. 3. From Fig. 3(a), it is clearly seen that the product is consists of the uniform octahedrons with their sizes in the range 300–550 nm. The TEM image of several individual Zn-doped Fe3O4 (Fig. 3b) confirms its octahedral structure. The small pic in Fig. 3(b) shows a normal structural model of an octahedral Zn doped Fe3O4 composed by eight [111] facets. Figure 3c depict the related high-resolution TEM (HRTEM) image and clearly reveal their high crystalline because of the well-resolved lattice fringes. The distances of the two different sets of crystal planes are 0,435 nm, which match well with the crystallographic planes (111) of Zn doped Fe3O4 (Sun et al., 2013). The corresponding selected-area electron diffraction pattern (SAED) of the single Zn doped Fe3O4 octahedron shown in Fig. 3d further reveal the single crystal nature of the octahedral sample (Chen et al., 2013). These above results confirms that the as-obtained Zn doped Fe3O4 are in fact single crystal octahedrons.
The FTIR is the normal method to ascertain the chemical composition and structure of spinel Zn ferrites. The FTIR spectrum of pure Fe3O4 and Zn doped Fe3O4 was depicted in the Fig. 4. The spectra data exhibited the distribution of cations in the crystal structure through their vibration modes. It is said that the metal cations usually expose at two different sub-lattices as the tetrahedral sites and octahedral sites in magnetite. The peak around 600 cm− 1 is ascribed to the stretching vibration mode of the tetrahedral site (A-site) and peak around 400 cm− 1 to that of the octahedral site (B-site) (Köseoğlu et al., 2008). The band at 565 and 424 cm− 1 in Fe3O4 spectrum confirm the formation phase which corresponds to the Fe–O absorption bonds at the A-site and B-site positions in the crystalline lattice of pure Fe3O4 (Cen and Nan, 2018). Observing Fig. 4, the FTIR spectra of pure Fe3O4 is slightly different from Zn doped Fe3O4 samples. The presence of the blue shift to 565 and 424 cm− 1 in the Zn doped Fe3O4 spectrum is attributed to the absorption bands of the Zn–O bond vibrations. The transition of the absorption bands to a higher level, suggesting Zn pieces occupied in the Fe3O4 crystal lattice (Mohapatra et al., 2013).
To investigate the valence states and chemical formation of Zn doped Fe3O4 octahedrons, the as-prepared product was checked using X-ray photoelectron spectroscopy (XPS) measurement. As shown in Fig. 5a, the survey spectrum of Zn doped Fe3O4 sample could be fitted with characteristic signals of elements as Zn, Fe, O, and adventitious C. The appear of peak at 284.8 eV in the spectrum can be attributed to carbon contamination, and CO2 penetrating in the surface of the sample after synthesis. The peak observed at 530.1 eV is O 1s, which can be ascribed to oxygen in metal oxides. In a spinel ferrite, the core level binding energies of Fe 2p electrons will different in the two Fe cations at octahedral sites and tetrahedral sites. In the as-prepared sample, the valence states of the Fe cations are mixed (Fe2+ and Fe3+), thus different from a normal spinel ferrite. It can cause the emitted photoelectrons from the 2p states with different energies. As shown in Fig. 5b, the Fe 2p3/2 broad band of the octahedral sample can be consists of four sub-signal named with “1–4”, where “1” belong to Fe2+ (B-site) sub-signal at 708.9 eV, “2” and “3” are Fe3+ sub-signal at the A and B sites with binding energies at 710.1 and 712.2 eV, respectively and along with associated satellite “4” at 718.3 eV. In the XPS spectrum of Zn 2p as shown in Fig. 5b, the presence of peaks at 1022 eV and 1045 eV are characteristic of to Zn2+ 2p3/2 and Zn2+ 2p1/2.. The results agree with the reported data [33]. In addition, no observed shoulder in the photoelectron spectrum suggests that zinc is selectively substituted into tetrahedral sites.The results futher confirm the formation of Zn doped Fe3O4.
3.2 The possible formation mechanism of hollowsphere Zn ferrite
To reveal the formation mechanism of octahedral Zn doped Fe3O4, time-dependent experiments were conducted at 4,8,12,16, 20, 24h. The time-dependent morphological evolution processes have been checked by SEM. Figure 6a–d depicted the SEM images of products obtained at the as-fixed reaction time. It can be seen that when the reaction time was 4h, bulk particles could be obtained (Fig. 4a). When the time reaction is longer than 8h, the micro quasi-spherical morphology products can be observed as Fig. 6b, which is ascribed to the Ostwald
ripening processes. As well known, Ostwald ripening processes is the growth of larger particles at the expense of smaller particles through a recrystallization process due to the energy difference among them. These processes have been widely used to explain successfully fabrication formation of many sphere nanomaterial systems (Costi et al., 2010; Guo et al., 2013). Keep it in mind and carefully observe SEM images in Fig. 7c showed that the micro quasi-spheres changed into the like-polyhedrons at 12 h reaction time. When the reaction time rises up to 16 h, the polyhedrons can be clearly achieved as shown at Fig. 6d. That alteration could be ascribed to the crystallographic surfaces that enclose the particles. In the hydrothermal synthesis, ethylene glycol molecules play an important role in gradually reducing the difference among the surface energies, leading to the selective adsorption on the (111) facets and increase the growth of (100) facets. As a result, the Zn doped Fe3O4 octahedrons are formed because the growth rate along the [100] direction is higher than that along the [111] direction (B. Y. Geng, 2008; WeiLei, 2017). The morphology of the obtained products changed significantly with the presence of the octahedron rolls, when the reaction time is up to 20 h. Finally, the uniform Zn doped Fe3O4 octahedrons were prepared at the reaction time of 24h. The arrangement of the energy surface process and the crystallographic growth process while long heating time at high temperature might is a possible reason for this significant morphology evolution.
It is noted that some amount of N2H4.H2O is also critical to the formation of Zn doped Fe3O4 octahedrons. To uncover the role of hydrazine hydrate in the synthesis process, a series of parallel experiments were conducted by altering the amount of hydrazine hydrate. In our experiments, without N2H4.H2O added into the reaction system, only particles obtained. When a small quantity (2 mmol of N2H4.H2O) is accessed in system only the like-sphere products prepared. The octahedrons products in 300–500 nm size are formed when 5 mmol of N2H4.H2O. When increasing the amount of N2H4.H2O (10 mmol), the size of the as-prepared octahedrons are 400–700 nm. Figure 5a-d presents the SEM images of the as-synthesized samples derived at 0, 2, 5, 10 mmol of hydrazine hydrate solution (N2H4.H2O), respectively. Herein, N2H4.H2O could serve as a very necessary factor added for electrostatic stabilization to prevent agglomeration of particles thus affecting the morphological evolution and size distribution for the Zn doped Fe3O4 octahedrons. In the reaction, N2H4.H2O is hydrolyzed at a high temperature in solution resulting in NH4OH. And NH4OH is a weak acid-base, thus divided into NH3 and H2O as gaseous bubbles as below reactions.
3N2H4 + 4H2O → 4NH4OH + N2 (1)
NH4OH→NH3 + H2O (2)
Then, the generated gaseous bubbles with high surface energy due to their small size diameter and could serves as the heterogeneous nucleation center for aggregating newly formed nanoparticles around the gas-liquid interface. In other hand, in the solvothermal systems, these product gaseous bubbles could provide as a soft template to form the quasi-sphere products. Without N2H4 introduction the decomposition reaction of NH4OH into gaseous bubbles might not occur completely. Without a heterogeneous nucleation center also as soft template, particle products are dominant rather than like-sphere structure (Fig. 7a-b). When more N2H4.H2O added into the synthesis at 5 and 10 mmol, the significant morphology change happens leading to form single-octahedral Zn doped Fe3O4 (Fig. 7c-d). The sequence of the surface energy (γ) for different crystallographic facets usually is γ[111]< γ[001] < γ[101] for iron oxide. Thus, N2H4.H2O or NH4+ adsorb selectively on [111] facets, which result in a higher reaction rate along the other directions. Then, [111] facets with the lowest growth rate will be dominated leading to an octahedron shape products (Chen et al., 2014; Mitra et al., 2014). As a result, the simultaneously occupation of the processes as the Ostwald ripening, the crystallographic surfaces and the growth rate of the basal surfaces could play an important for evolution morphology of the as-prepared products. The results suggest that the different sizes of the Zn doped Fe3O4 octahedrons could be synthesized controllably through modifying the amount of hydrazine hydrate reactant.
3.3. Photo-Fenton activities using Zn doped Fe3O4 octahedrons as catalysts
Many reports have demonstrated that the photocatalytic activity of photocatalysts is enhanced due to their exposed crystal facets, crystallinity, morphology, size particle (Fernando Santos Domingues, 2019; Rakkesh, 2015). Thus, the as-synthesized Zn doped Fe3O4 octahedrons are expected to show the higher photo-Fenton performance compared to the Zn ferrite prepared by solid-state reaction method.
The photocatalytic reaction of the samples was evaluated through the degradation of rhodamine B (RhB) aqueous solution in the presence of H2O2 under visible irradiation. To survey the photocatalytic activity of the as-prepared Zn doped Fe3O4 octahedrons, the results are compared with the efficiency of Zn doped Fe3O4 nanoparticles. The results of photocatalytic activities of the samples prepared at different conditions are shown in Fig. 8. When the photocatalyst is added but absence of H2O2 only slight photodegradation can be observed. The above experiments demonstrate that in this case, the degradation of RhB needed both catalysts and H2O2. The photocatalytic activity of the Zn doped Fe3O4 octahedrons under visible light irradiation are further checked by comparison with that of Zn doped Fe3O4 nanoparticles. The octahedrons are much more photocatalytically efficient than Zn doped Fe3O4 nanoparticles. As shown in Fig. 8a, about 97.5% of RhB is photo decomposed at the irradiation time of 60 minutes with Zn doped FeO photocatalysts while only 26% and 20% of RhB removed when ZnFe2O4 nanoparticles or Fe3O4 nanoparticles added at the same condition, respectively. The excellent photocatalytic performance of Zn doped Fe3O4 octahedrons can be associated with their crystalline structure and octahedral morphology. The adsorption of RhB on the as-prepared octahedrons in the dark was also examined. The RhB adsorption ability was negligible, indicating that the degradation of RhB was due to photodegradation but not adsorption. Figure 8b displays the absorption spectra of the RhB solution show a characteristic peak at 554 nm. From Fig. 6b, it is seen that the intensity of the absorption peak at 554 nm decreased gradually following time reaction. At a reaction time of 60 minutes the absorption peak intensity is minimal, demonstrating that the RhB was removed. The mineralization of RhB was also investigated as depicted in Fig. 8c. The results reveal that the TOC removal of RhB using Zn doped Fe3O4 octahedrons as catalysts gained about 72%. The results suggest that Zn doped Fe3O4 as-synthesized exposes high capacity for the mineralization of contaminants.
Effect of catalyst amount and pH in the range of 2 to 8 on RhB degradation efficiency was also studied (as shown in Fig. 9). The results show the degradation rate of RhB enhanced with an increase in catalyst amount as depicted in Fig. 9a. However in higher catalyst dosage, the dye removal percentage slightly decreased. Based on the experiment, 1.5 mg/L of Zn doped Fe3O4 octahedrons is consistent for the RhB photo-Fenton removal. The experiment results on the effect of pH reveal that the optimum pH was 5.5 (seen Fig. 9b). With pH below 5.5, in high H+ concentration, the formation of stable oxonium ion [H3O2]+ makes hydrogen peroxide more stable and then decrease its activity with ferrous ions. Moreover, the formation of Fe(II) complexes and ferric oxyhydroxides precipitation at a pH above 5.5 are probably reasons for efficiency decreases in the photo-Fenton RhB removal processes (Domingues et al., 2017).
As known, the Fenton reaction is one of the most effective advanced oxidation processes for wastewater treatment in which the active hydroxyl radicals generated by reaction between Fe2+ and H2O2. It is reported that the presence of Fe2+ in the oxide plays an important role for the activation of H2O2 (Cheng et al., 2014; Feng et al., 2013). With the present of both visible light and H2O2, active hydroxyl radicals will be generated by main reactions following:
H2O2 + Fe (II) → Fe (Ш) + •OH + OH− (3)
Fe (Ш) + e- → Fe (II) (4)
Then more •OH can be produced resulting in a reaction between regenerated Fe(II) with H2O2 (Eq. (1)). Therefore, the kinetics of the reaction between •OH and RhB is enhanced remarkably via visible light irradiation (Feng et al., 2013). Moreover, the enhancement of the photo-Fenton reaction of Zn-doped Fe3O4 could be ascribed to the Zn2+ substitutes in tetrahedral sites and octahedral sites leading to the electron transfer process accelerated. As a result, the interface between Fe3+ and H2O2 improved, which result in more ·OH radical from high rate of decomposition H2O2 (Alimard, 2019). Furthermore, the crystalline structure and octahedral morphology are important factors that could enhance the photocatalytic perform of the as-prepared Zn doped Fe3O4.
In addition, the recycle tests were conducted to survey the stability of the as-obtained products in the oxidation process under Vis light irradiation. The results show that the catalyst was easily separated by an internal magnet and the RhB degradation effectively has no significant change during the four successive cycles, indicative of high stability of the catalyst (Fig. 8d). XRD spectra of Zn doped Fe3O4 before and after 4 runs show no significant changes as depicted in Fig. 9e. These properties play a very important role in application for water treatment at industry scale. The high photocatalytic activity, the stability and the easily separation suggest that the Zn doped Fe3O4 octahedrons can be promising candidates for the photo-Fenton degradation application.