Structural, Magnetic, Optical Properties and Photocatalytic Activity of Nanocrystalline Cobalt Ferrite Prepared by Three Different Methods


 Cobalt ferrite (CoFe2O4) nanoparticles were synthesized through three different routes (chemical co-precipitation, ceramic, and sol-gel). The products were characterized by XRD, SEM, TEM, Raman spectroscopy, FTIR, photoluminescence (PL), BET, XPS, vibrating sample magnetometer (VSM). The X-ray analysis validates the spinel structure of the samples with particle sizes between 21 – 36 nm. XPS revealed the impact of the preparation route on cation distribution at the tetrahedral and octahedral sites in spinel ferrite lattice. All ferrite surfaces exhibit a mesoporous structure with a surface area between 25 – 99 m2/g. The results show that the prepared samples exhibit either ferromagnetic or super-paramagnetic behavior. The ferrite samples showed a photocatalytic performance under visible light. The present work demonstrates that each of the particle size and cation distribution is effective in controlling the structural, magnetic, and optical properties. The results obtained reveal that this kind of CoFe2O4 spinel ferrite nanoparticles has promising applications in magnetic designs and water treatment fields.


Introduction
Recently, magnetic nanoparticles have drawn an excessive arrangement of research due to their characteristic properties and their scienti c, technological uses. Ferrite spinels consider as the most important class of these materials. The spinel structure exhibits a general formula (A) [B 2 ]O 4 crystallizes in a face-centered cubic structure, with two lattice types for cation distributions of A and B sites in the tetrahedral and octahedral organization, respectively. The physical properties of ferrite spinel are linked to the cation distribution over the tetrahedral sites A (8a) and octahedral sites B (16d) of the structure. In practice, the spinel structure is more complex because molecules do not always correspond fully to a normal or inverse structure. Hence, an inversion parameter, d, de ned by 0 < d < 1 is introduced [1]. Cobalt ferrite is one of the most appropriate spinel ferrites which hold distinctive properties such as strong spin-orbit coupling, high coercivity, high Curie temperature, high magneto crystalline anisotropy, distinguished mechanical, and chemical stability [2][3][4][5][6][7].
Generally, cobalt ferrite exhibits a partially inverse spinel structure in which both sites (A and B) hold a fraction of Co 2+ and Fe 3+ cations and shows the formula of (Co 2+ 1-d 4 where d is a fraction of tetrahedral sites occupied by Fe 3+ ions, which is known as the degree of inversion. The degree of inversion is sensitive to numerous parameters such as the microstructure and synthesis process. This work is designed to prepare CoFe 2 O 4 with three different methods (co-precipitation, ceramic, and sol-gel) and investigate how the change in each of the cation distribution, particle size, and surface area affect the optical and magnetic properties, which ultimately determine the magnetic and photocatalytic applications. Experimental

Materials
All chemicals were reagent grade and utilized without further puri cation. Cobalt

Characterization Methods
XRD analysis was performed on a Philips X' Pert Pro Super diffractometer with Cu Kα radiation (λ=1.54 Å).
Electron microscope analysis was done by SEM and TEM electron microscopy model JEOL JEM-100CXII and JEOL-2010, respectively. The FT-IR spectra of the specimens were obtained by employing a Brucker-FT-IR in the range of 4000-400 cm -1 . A U-1000 laser Raman Spectrometer with 514.5 nm line of an ArC laser was used to determine Raman spectra. Optical absorption was measured by employing Cary 5G equipment, at wavelengths ranging from 200 to 1000 nm. The photoluminescence (PL) studies were made using a 225 nm excitation wavelength source. Traditional surface area and textural surface properties were studied using the BET-N 2 gas adsorption technique. The electron binding energies for the elements were determined using a PHI-5702 multifunctional spectrometer with AlKα radiation (XPS). The magnetic properties of the prepared samples were studied using the vibrating sample magnetometer model (VSM-9600M-1, USA) in a maximum applied eld of 10 kOe.
A programable ammonia desorption technique (TPD-NH 3 ) was used to determine the type and the amount of acidity in the studied samples. This is performed in a reactor lled with 0.4 g of the sample, which was previously activated at 250 ℃ for 3 hours with N 2 ow. The sample was then permitted to expose to NH 3 at room temperature for one hour. The over ow of NH 3 was swill out of the reactor with N 2 ow. The temperature was then elevated linearly at a rate of 10 ℃/min to liberate NH 3 gas from the sample at a temperature interval of 100 ℃ up to 500 ℃. The liberated NH 3 is allowed to ow through a known large amount of H 2 SO 4 solution. Finally, the amount of acid was determined by back titration with NaOH solution using M.O indicator.

Photocatalytic Study
Basic Red 18 (BR 18) dye was selected as an ideal system for a catalytic reaction because of its intense color in an aqueous medium and low biodegradability due to the existence of benzene rings. The photocatalytic degradation of Basic Red 18 (BR 18) dye over ferrite nanoparticles was performed in a photoreactor with a total capacity of 0.5 L. The illumination source was a UV-C lamp (200-980 nm, 9 W, Philips) put in the internal quartz tube of the photoreactor. The impact of ferrite dosage on Basic Red 18 (BR 18) dye oxidation was examined by utilizing a dose of 0-0.05 g/l. The photocatalytic study was investigated at a pH range of 2-9 and initial dye concentration in the range of 50-200 ppm. The solution pH was regulated using NaOH or HCl. Specimens were removed from the arrangement at speci c periods and centrifuged, and the concentration in the supernatant dye solution was then analyzed using UV-vis Spectrophotometer (a Cary 5G equipment).  Table 1.

Results And Discussion
Where "h k l" are miller indices. It is noted that the lattice constant follows the order: CoFe s > CoFe p > CoFe g .
This may be attributed to that a certain number of Co 2+ ions (0.78 Å) transfer from octahedral sites, accompanied by opposite migrate of equivalent number of Fe 3+ ions (0.645 Å) from tetrahedral to octahedral sites to relax the compressive strain. The average crystallite sizes (D XRD ) of the investigated samples were calculated (based on 3 different peaks) from the widening of re ection peaks using the Scherrer formula [7]: Where λ is the X-ray wavelength and β is the half peak width of the diffraction peak in radiant. The results are listed in Table 1 and found to lie in the range of 21 -33 nm. The average X-ray density (ρ x ) of the cobalt ferrite nanoparticles was determines using the following equation: Where M is the molecular weight of cobalt ferrite, N is Avogadro's number. The results are also listed in Table  1.
The formation of the CoFe 2 O 4 spinel structure was also supported by infrared spectra shown in Fig.1S. The spectra of all samples demonstrate two speci c bands for spinel structure, ν 1 in the range 581-563 cm −1 , relates to M tetr -O vibration at the tetrahedral site, and ν 2 in the range 428-402 cm −1 attributed to M octa -O vibration at the octahedral sites [9][10][11]. The mean feature bands observed are recorded in Table 1 Table 2, [13] proving that pure ferrite had been produced.
To investigate the cation valence states and their distribution in the CoFe 2 O 4 spinel, the high-resolution XPS spectra of Fe 2p, Co 2p, and O1s peaks of the CoFe 2 O 4 specimen were studied and given for CoFe g in Fig. 4.
The integrated intensities of the tted peaks of Co 2+ and Fe 3+ ions were used to determine their distributions in both octahedral and tetrahedral positions. The table also shows an increase in the concentration of Fe 3+ cations on octahedral sites with the increase in the particle size of the sample, which agrees well with XRD data. In conclusion, it can be said that the selection of the preparation route is effective in controlling the cation distribution within the spinel lattice. The presence of high intense satellite structure on the high binding energy side of the Co 2p 3/2 and Fe 2p 3/2 might be attributed to the band structure related to octahedral Co 2p in the oxide lattice.

Magnetic Study
In order to investigate the effect of synthetic method on the magnetic properties, the VSM test was done at room temperature in an applied eld of 10 kOe. The results are represented in Fig For the ideal inverse spinel crystal structure of CoFe 2 O 4 , with all the Co 2+ ions located at the octahedral site, the magnetization per formula unit can be theoretically evaluated using Neel's two sub lattice model by considering the difference of total magnetic moments in octahedral and tetrahedral sites [15][16][17][18] M=M octahedral -M tetrahedral . The magnetic moment of Fe 3+ and Co 2+ cations are 5.0 and 3.8 µB respectively, a theoretical magnetic moment of CoFe 2 O 4 is 3.8 µB per formula unit. Based on the cation distribution obtained from XPS, the magnetization per formula units were also calculated and listed in Table 3, which shows that the magnetic moments changes with the preparation methods. The magnetic moment values of the investigated samples can be also determined experimentally by the following equation in Bohr magneton; µ B = Mol.wt×Ms/5585 (4) However, the evaluated data from XPS and VSM are not equal, which can be associated to the nite size of nanoparticles conducting to the noncollinearity of magnetic moments on the surface of the nanoparticles. The disordered moments are developed due to the broken exchange bonds at the outer layer. On the other hand, the competition antiferromagnetic interactions precedes to a noncollinear arrangement of magnetic moments within interstitial sublattices which is induced because of the non-equilibrium cation distribution among tetrahedral and octahedral sites [15,17,18] [17]. The low M s -value of the investigate ferrite samples compared with that of the bulk one (80.9 emu/g) [3] can be also explained on the basis of the core-shell model, which clari es that the nite-size effects of the nanoparticles manage to canting or non-collinearity of spins on their surface, in that way reducing magnetization [16].
To sum up, it can be said that the change in saturation magnetization with the variation in the preparation methods is possibly due to the rearrangement of the cation distribution, i.e., the exchange of Co 2+ and Fe 3+ ions from octahedral and tetrahedral sites and vice versa. The low values of M s for the investigated samples could be credited to surface distortion which destabilizes the collinear spin arrangement and producing various canted spin structures at the surface. This effect is especially noticeable for ultra ne particles owing to their large surface to volume ratio. The reduction in coercivity with increasing particle size could be accredited from the combination of surface anisotropy and thermal energies [19].
The values of the squareness ratio (M r /M s ) of investigated samples, shown in Table 3, are below 0.5 refers to that these samples are multidomain and the particles interact by magnetostatic interaction [20].
The magnetic anisotropy (K / ) has been also calculated using the following relation [21], H c = 0.98 K / /M s (5) and the results obtained showed high values of 16711, 13571, and 6181 emu.Oeg -1 for CoFe p , CoFe g , and CoFe s , respectively. The increase in K value is going parallel with increasing the presence of Co 2+ ions in the octahedral sites, as shown in XPS results, Table 2.
The effect of the preparation method on the magnetic parameters of our CoFe 2 O 4 nanoparticles is compared with other methods present in the literature and listed in Table 4.

Surface Properties
The surface properties of the spinels investigated were studied using the BET technique. The isothermal N 2 adsorption-desorption plots of these samples Fig.2S can be classi ed as type IV for CoFe p and type V for CoFe g and CoFe s according to IUPAC, which is characteristic of the mesoporous material in which the adsorption proceeds via multilayer adsorption followed by capillary condensation. The hysteresis loops show the type H3 (aggregates of platelike particles forming slit-shaped pores) for all samples. The textural properties of the studied samples including BET-surface area, average pore diameter, pore-volume, and pore size distribution calculated from BJH method are derived from N 2 nitrogen adsorption/desorption isotherms and listed in Table 5. The data obtained refer to that the surface properties depend to a large extent on the method of preparation and CoFe g exhibits the highest surface area.

Optical Properties
The prepared CoFe 2 O 4 nanoparticles show still high magnetization, that photocatalyst appropriate for magnetically separable by a magnetic eld and separation of photocatalyst from solution. Thus, the photocatalytic activity of the investigated samples has been studied. The optical absorption property related to the electronic structure characteristic is documented as the main factor in deciding the photocatalytic activity [30]. The diffuse re ectance spectra of our ferrites were recorded and converted to the Kubelka-Munk function, K-M, Fig. 6, using the following equation: Where R is absolute re ectance. The results obtained are listed in Fig. 6A is a Planck , s constant, n is the light frequency and n is a constant relating to a mode of transition ( n = ½ for allowed direct transition and n= 2 for indirect transition). Tauc's plots, shown in Fig.6-b, for every one of the specimens, demonstrated that the band-to-band direct transitions are more inclined to happen than the indirect transitions. The optical energy gaps, E g , obtained from the intercept of the plot with the X-axis are recorded in Table 6, from which it can be seen that E g value decreases with increasing the particle size and showed the smallest value for CoFe g sample.

Raman Spectroscopy
Raman spectra were used to acquire vision on the vibrational energy states within the spinel ferrite obtained, as well as to review the structural characteristics and compositional regularity throughout the samples [34].
The spinel ferrite exhibit ve active Raman vibration modes [35,36]. The Raman spectra of our samples showed only three bands, due to peak overlapping, at 284-297 cm -1 , 460-470 cm -1 , and 640-650 cm -1 , as shown in Fig. 7, which was assigned as E g , 3T 2g and A 1g (1), respectively [37]. The A 1g (1) band is related to symmetric stretching vibration mode at the tetrahedral (A) site. While A 1g (1) band might be related to the vibration of Co-O bonds at the tetrahedral (A) site. The results obtained in Fig. 7; show that the frequency of Raman modes of the investigated spinels is slightly changed with the preparation method. This could be attributed to the variation in the cation distribution in the spinel lattice, as mentioned above.

Photoluminescence Study
Photoluminescence (PL) spectroscopy is an outstanding procedure to get valuable information concerning energy and the dynamics of charge carriers yielded during the exposure of light. The photoluminescence of the ferrite nanoparticles was studied using a 225 nm excitation wavelength source and the results obtained are presented in Fig. 8. The spectra of all samples show broad visible emission peaks at 434 -442 nm, which are attributed to the charge transport between Co 2+ at tetrahedral sites and Fe 3+ at octahedral sites that are surrounded by O 2ions [38]. The variation in the position and the intensity of luminescence can be explained on the basis that the PL-spectra are sensitive to the character of nanoparticles surface, due to the existence of gap surface disorders developing from surface non-stoichiometry and unsaturated bonds. The defects produced in the nanomaterial lattice during preparation are the base of luminescent properties [39]. The emission intensity of CoFe g is lower than that of the rest samples. This indicates that this sample acted as traps for the photo-induced charge carriers. These outcomes con rm the previously mentioned results on the in uence of the preparation methods on the surface and optical properties.

Surface Acidity:
Temperature programmed desorption of ammonia (TPD-NH 3 ) is an appropriate procedure for measuring the quantity and the spreading of the acid sites on the surface of our samples. The acid site distribution results for the studied samples are summarized in Table 6. The ammonia desorbed at 100 o C contains some physisorbed ammonia as well as overstating the proportion of weak acid sites. Whereas, the ammonia desorbed at 220 -370 o C and that at 450-600 o C are attributed to medium and strong acid sites, respectively [40]. The results obtained show that the acidity varies with preparation methods. For all samples, the strength of acidic sites follows the order: Weak acid sites > medium sites > strong acid sites. CoFe g exhibits the highest total acidity than that of other samples.

Photocatalytic Activity of CoFe 2 O 4 Samples
According to the above-mentioned optical properties, we studied the photocatalytic activity of the investigated degradation results are illustrated in Fig. 9. From which it can be seen that the degradation of the dye is very slow in the absence of the catalyst and the CoFe g sample showed the highest photocatalytic e ciency due to the high optical absorptions in vis. Light region with lower bandgap energy and a larger surface area. Therefore, this sample was selected to test the impact of catalyst dosage, dye concentration, and pH of the solution on the dye degradation rate. The results obtained given in Fig. 3S show that the rate of dye degradation has the highest rate at 3 mg/L catalyst dosage and decreases with increasing dye concentration in the range of 10 -100 ppm and has the highest rate at pH= 7 of the solution.
In view of literature reports, the kinetic of photocatalytic reaction can be calculated according to: -ln (C/Co) = k obs t Where Co and C are the concentrations of dye at zero time and time t, respectively, and k obs are the pseudorst-order rate constant. The rate constant, k obs , evaluated from the slope of the straight line of plotting −ln(C/C0) vs. reaction time, Fig. 9-b showed a value of 0.1 min -1 .

Mechanism of Photocatalysis
The major oxidative species in the photocatalytic progression are positive holes (h + /VB) and the OHhydroxyl radical formed during the irradiation process. In the present work, the trapping experiments were used to determine which one of these species is active for organic degradation. EDTA-2Na and isopropyl alcohol were used as an h + , and isopropyl alcohol as an OH· scavenger, respectively [32]. The results showed that the additive of isopropyl alcohol slightly changed the dye degradation indicating that OH . radicals were minor factors in the photocatalytic degradation process, whereas the addition of h + capture (EDTA-2Na) caused a great decrease in the degradation e ciency as shown in Fig. 10. This foundation distinctly denoted that positive holes are the major active species of the dye dissociation.

Conclusions
Nanosized cobalt ferrite (CoFe 2 O 4 ) particles are synthesized through three different methods (sol-gel, solidstate blending, and co-precipitation). The proposed preparations are inexpensive and thus appropriate for the large-scale production of such type nanoparticles. XRD approves the formation of a spinel phase in all preparation methods. The FT-IR spectra displayed two characteristic metal-oxygen, vibrational bands for the spinel structure. The results reveal that the synthetic route affected each of the particle size, morphological structure, surface textures, surface acidic properties, cation distribution, optical and magnetic properties. Both Fe 3+ and Co 2+ are distributed in octahedral and tetrahedral sites in a ratio depending on the preparation method. The variation in this ratio controlled the studied properties. The results showed that the prepared ferrite exhibits either ferromagnetic or super-paramagnetic behavior with magnetic parameters of particular importance for allowing the magnetic recovery and reuse of the catalyst and the possibility to be used in recording media. Ammonia TPD analyses showed that weak acid sites prevail medium-strength sites, whereas the number of strong acid sites is the least. The obtained ferrite is utilized as photocatalytic for