Introduction

Zn-containing EAFD as a solid waste produced during the steelmaking process contains different valuable metals such as iron, zinc, lead, chromium, manganese, calcium, magnesium, etc. The amount of dust emitted from this process is about 15 to 20 kg per 1 ton of molten steel1. Because of the presence of easy leachable elements like Zn, Cd, As, Pb and other heavy metals in the EAFD its land filling is not sustainable, it represents an environmental issue. The research on EAFD are concerned with recycling of zinc from EAFD by hydrometallurgical, pyro-metallurgical and electrolysis techniques1,2. On the other hand, the pickling process is applied to the surface of hot rolled steel strips to remove oxide layer, making steel strips suitable for galvanizing and drawing applications. In this process, hydrochloric acid reacts with the oxide layer on the steel surface. After pickling, hydrochloric acid becomes enriched in iron in the form of ferrous and ferric chlorides and this is called “waste pickle liquor”. Waste pickle liquor is not suitable for pickling process and its discharge represents an environmental problem3.

Ni–Zn ferrites have a wide range of applications in the various fields of science and technology such as EMI suppressors, electromagnet cores, transformer cores, antenna, video magnetic heads and magnetic heads of multiple path communication. On the other hand, magnetic ferrite nano-particles have novel applications like therapeutic applications mainly in the drug delivery systems, in addition to catalytic applications, magnetic resonance imaging, supercapacitors and gas sensors4,5,6.

A method for preparation of Ni–Zn ferrite using EAFD was introduced by Wang et al.7. EAFD was treated with 0.5 mole L−1 HCl solution to reduce its Ca content. The treated EAFD was mixed with NiCl2·6H2O with different mass ratios with subsequent grinding and calcination at different temperatures from 800 to 1100 °C. The mass ratios of EAFD and NiCl2.6H2O along with calcination temperature influenced the magnetic properties of the prepared single-phase spinel ferrite. The maximum obtained Ms was 60.5 emu/g and the minimum coercivity (Hc) was 49.8 Oe. In this work, a new preparation method for NZF was introduced using EAFD and WPL via co-precipitation technique. In this method EAFD was treated with acetic acid to reduce its Ca content, after acetic acid treatment, the EAFD was dissolved in hydrochloric acid and the leach liquor was mixed with WPL and NiCl2.6H2O followed with precipitation in NaOH solution. Co-precipitation method was selected for the proposed technique for several reasons; first is the direct usage of WPL, which is a ferric chloride liquor in the co-precipitation process. Second is making use of the dissolving step of EAFD at HCl in its purification from undesired impurities like Al and Si oxides, which have poor solubility at HCl solutions, so such impurities will not be transferred to the EAFD leach liquor that will be used for the co-precipitation process. And, finally the advantages of Co-precipitation process as a low cost method that doesn’t require high temperatures and provides crystalline nano particles with high purity in short duration.

The effect of acetic acid treatment on EAFD composition was studied and the influences of Ni ratio on the structural and magnetic properties of the prepared Ni–Zn ferrite was investigated using XRD, FTIR, FESEM and vibrating sample magnetometer (VSM) examinations. By the applied process in this paper, EAFD and WPL could be transformed from waste environmentally hazardous materials to a high added value material with both economic and environmental benefits.

Experimental details

EAFD treatment with acetic acid

Treatment of the EAFD with acetic acid was done to reduce the calcium oxide content. To determine the optimum acetic acid concentration for EAFD treatment, 150 mL of five different concentration acetic acid solutions (1, 2, 3, 4, 5 v/v%) were used for the treatment of five EAFD samples (15 g). The acetic acid treatment for each EAFD sample was carried at 80 °C for 1 h with a continuous stirring. The EAFD samples treated with 1, 2, 3, 4 and 5 v/v% acetic acid were denoted ATEAFD1, ATEAFD2, ATEAFD3, ATEAFD4, and ATEAFD5, respectively. The treated EAFD samples were collected by centrifugal separation at 2000 rpm and the five samples were washed with distilled water and dried at 250 °C for 4 h.

Dissolving of the acetic acid-treated EAFD in hydrochloric acid

A 300 g sample of 2% acetic acid-treated EAFD was dissolved in 3000 mL of 15% HCl at 80 °C with a continuous stirring for 1 h. The undissolved particles were separated with centrifugal separation at 2000 rpm, and leach liquor of the dissolved sample was obtained.

Preparation of the NZF samples by co-precipitation process

Four NZF samples with the formula Nix(Zn + impurities)1-xFe2O4 (where x = 0, 0.25, 0.5 and 0.75 and the impurities contains manganese, magnesium and calcium were prepared by co-precipitation technique. First, the mixed solutions were prepared by adding NiCl2·6H2O and WPL in stoichiometric ratios to four 200 mL solutions of the leach liquor obtained from dissolving of the acetic acid-treated EAFD in HCl. Each mixed solution was added to 0.6 M NaOH solution dropwise with stirring of speed 50 rpm at room temperature. NaOH solutions were prepared taking the ratio Fe:Na to be 1:4 for each NZF composition. pH 10 was kept during the precipitation process by continuous titration with NaOH. During the addition, dark grey precipitates were obtained. The supernatant liquors were decanted, and the remaining suspensions were placed into a dryer at 200 °C for 4 h to complete ferritization reaction. The dried precipitates were washed several times with distilled water and dried at 200 °C for 4 h. The flow chart of the preparation process is shown in Fig. 1.

Figure 1
figure 1

Flow chart of NZF preparation using EAFD and WPL.

Full size image

Characterization

Chemical compositions of the WPL and the leach liquor obtained by dissolving of the acetic acid-treated EAFD in HCl were obtained using ICP analysis (ICP-OES Prodigy, Teledyne Leeman Labs, USA).

X-ray Fluorescence (Rigaku Supermini 200, Japan) was used to get the chemical composition of the EAFD before and after acetic acid treatment. It was also used to determine the chemical composition of the prepared nickel-zinc ferrite samples.

For phase identification, estimation of crystallite size, and other structural parameters. XRD examination was performed using X ray-diffractometer (Shimadzu-XRD 6100) using Cu K∝ (λ = 1.54060 Å) radiation, with continuous scan mode and scanning speed of 12 degree/minute for scanning range of 5–80 degrees. FTIR examination for the prepared nickel-zinc ferrite samples was carried out in the range of 4000–200 cm−1 using FTIR-8400S, Shumadzu, Japan device. The dried samples were pressed with KBr matrix, and spectra measurements were according to transmittance method. Magnetic measurements were carried out at room temperature with a maximum magnetic field of 20,000 Oe using a Lakeshore vibrating sample magnetometer (model: 8600 series VSM, lakeshore, USA) and magnetic parameters including Ms, Hc and remanence (Mr) were evaluated. NZF samples were subjected to SEM examinations using JEOL scanning microscope (JSM-IT200 Series).

Results and discussion

Chemical composition of WPL

The chemical analysis of WPL is shown in Table 1. It is seen that the WPL contains mainly Fe in the form of ferrous/ferric chlorides and that Mn concentration is very low so it can be used directly for the synthesis of nickel-zinc ferrite.

Table 1 Chemical composition of the WPL.
Full size table

Effect of acetic treatment on EAFD composition

The chemical composition of the EAFD before and after leaching with acetic acid is given in Table 2. It is seen that the effect of acetic acid treatment on Fe, Zn and Ca concentrations is most obvious than its effect on the other EAFD components. Figure 2 shows the effect of acetic acid concentration on Fe, Zn and Ca concentration in the EAFD. With increase of acetic acid concentration, iron concentration is increased while calcium concentration is decreased. For low acetic acid concentrations of 1 and 2 v/v%, Zn concentration is raised while Zn concentration is declined at higher acetic acid concentrations of 3, 4, and 5 v/v%. CaO is removed upon treatment with acetic acid due to its reaction with acetic acid according to reactions (1) and (2)8,9.

$$ {\text{CaO}} + {\text{H}}_{{2}} {\text{O}} \to {\text{Ca}}\left( {{\text{OH}}} \right)_{{2}} , $$
(1)
$$ {\text{Ca}}\left( {{\text{OH}}} \right)_{{2}} + {\text{ 2CH}}_{{3}} {\text{COOH}} \to {\text{Ca}}\left( {{\text{CH3COO}}} \right)_{{2}} + {\text{ 2H}}_{{2}} {\text{O}}. $$
(2)
Table 2 Chemical composition of the EAFD.
Full size table
Figure 2
figure 2

Effect of acetic acid treatment on chemical composition of EAFD.

Full size image

In addition, acetic acid treatment depresses Pb weight percent from 1.1% ± 0.024% in the raw EAFD sample to 0.06% ± 0.024 in the 2% acetic acid treated sample due to formation of lead acetate.

The EAFD material yield after the acetic acid treatment with different concentrations was calculated from the weight difference between the EAFD samples before and after treatment. As indicated from Fig. 3, the highest yield is 88.3%, which obtained at acetic acid concentrations of 1 and 2% v/v.

Figure 3
figure 3

EAFD yield for the treated samples with different concentrations of acetic acid.

Full size image

The best acetic acid concentration accordingly is 2 v/v% which reduced the Ca content by 70.6% ± 0.40% (from 8.5 to 2.5%) with the maximum zinc enrichment (24.5% ± 0.35) in addition to the maximum EAFD material yield (88.3%).

The XRD patterns of the EAFD before and after 2% v/v acetic acid-treatment are shown in Fig. 4a,b, respectively. Untreated EAFD sample shows the diffraction peaks at 2θ around 19.2°, 29.9°, 35.1°, 36.6°, 42.7°, 53°, 56.5° and 62°. These peaks are characteristic of spinel ferrites10. On the other hand, the peaks of CaO are appeared at 2θ of 32.2, 36, 53.2°, and Ca(OH)2 peaks are displayed at around 28.5°, 34.3°, 47.8°, and 50.7°11. The XRD pattern of the acetic-acid treated sample shows that a disappearance of the Ca(OH)2 peaks and decrease of CaO peaks intensities are observed.

Figure 4
figure 4

XRD patterns of EAFD (a) and 2% v/v acetic acid-treated EAFD (b).

Full size image

Chemical composition of the dissolved acetic acid-treated EAFD leach liquor

The chemical analysis of the obtained liquor after dissolving of the acetic acid-treated EAFD sample in 15% HCl solution is illustrated in Table 3. It is noticed that Si and Al didn’t exist which indicates that their oxides have poor solubility in HCl. The weight of the undissolved particles is 29.23 g representing 9.74% of the acetic acid-treated EAFD sample weight. Incomplete dissolving of the sample could be due to the impurities of metal oxides with a lower solubility in HCl solution mainly silicon and aluminum.

Table 3 Chemical composition of the acetic acid-treated EAFD leach liquor.
Full size table

Chemical analysis of the prepared NZF precipitates

The chemical analysis of the prepared samples is shown in Table 4. Although Cu and Pb are detected in dissolved acetic acid treaded EAFD solution, they are disappeared in the prepared NZF samples due to their very low concentration in the precipitation solution.

Table 4 Chemical analysis of the prepared NZF samples.
Full size table

Structural properties of the prepared NZF samples

XRD patterns of the as prepared NZF samples with different Ni contents are shown in Figure 5.

Figure 5
figure 5

XRD patterns for the prepared NZF samples with different Ni contents.

Full size image

The results indicate that the reflections in all the cases are characteristic of the spinel structure without any impurity phases. There is a slight shifting of the 2θ of the diffraction peaks to higher values with increase of Ni ratio. This shifting is attributed to the fact that the size of Ni2+ cation is smaller than that of Zn2+ cation (69 Å and 74 Å respectively), so the lattice parameter of the spinel unit cell decreases with increasing of Ni content. This results in a larger diffraction angle because the diffraction angle θ is inversely proportional to the lattice parameter. The decrease of lattice constant with increase of Ni content was reported by many previous studies about the effect of Ni substitution ratio on the structural properties of NZF12,13,14.

The crystallites size (D) of NZF samples was calculated using Scherer formula15.

$$D\left(hkl\right)=\frac{0.90\uplambda }{{\upbeta \cos}\uptheta },$$
(3)

where β is the peaks full width at half maximum in radians, and λ is the wavelength of the used X-ray beam and 0.90 is the shape factor.

Peak of (311) reflection is used for calculations because it is the sharpest one. The crystallites size are 10.8, 10.7, 12.0 and 16.7 nm for samples N0.0ZF, N0.25ZF, N0.50ZF and N0.75ZF, respectively. The crystallite size is almost the same at Ni = 0.0 and 0.25, while further increase of Ni to 0.5 and 0.75 showed an obvious increase in the crystallite size. The observed increase of the crystallite size with increase of Ni content was observed by Chuan et al.16 and Upadhyay et al.17. It is attributed to Zn-ferrite formation is more exothermic than Ni-ferrite formation, so the surface temperature of the formed crystals becomes higher with increase of the Zn content which reduces the molecules concentration at the surface of the growing crystals.

The lattice parameters for each sample was calculated using the formula18,19.

$$a=d\sqrt{{h}^{2}+{k}^{2}+{l}^{2}},$$
(4)

where a is lattice parameter, d is the inter planer spacing and h, k, l are the miller indices of (311).

The lattice parameter shows a slight decrease in its value with increase of Ni content which is attributed to its lower cation size compared to Zn cation.

The theoretical XRD density ρx was calculated using the formula20,21.

$${\rho }_{x} =\frac{8M}{{A}_{N}V},$$
(5)

where M is NZF samples molecular weight, AN is Avogadro’s number (mol−1) and V is the lattice volume. A slight increase of ρx is observed with increase of Ni ratio due to the accompanied reduced lattice parameter.

The lattice strain ε was calculated using Williamson–Hall equation22,23.

$${{\upbeta_{\text{hkl}}} \cos \theta }= \frac{k\lambda }{D}+4 \varepsilon \sin \theta ,$$
(6)

the lattice strain can be estimated with the aid of Eq. (6) from the slope of βhkl cosθ versus \(4 sin\theta \) plot as in Fig. 6. The estimated lattice stain values declines with the increase of Ni content due to the smaller Ni2+ cation size which makes less distortion for the spinel lattice than the larger Zn2+ cations, a similar phenomenon was reported by Abu El-Fadl et al.13 in the study of NZF prepared by microwave combustion method and by Srinivas et al.24 in co-precipitated NZF and by Batoo et al. in nanoparticles of Ni–Cu–Mg ferrite prepared by sol–gel method25. The XRD calculated crystallites size, lattice parameter, lattice strain and XRD density for all samples are listed in Table 5.

Figure 6
figure 6

Plots of βhkl cosθ vs 4sinθ of NZF samples.

Full size image
Table 5 Results of the XRD test for the NZF samples.
Full size table

FTIR spectra of the prepared NZF samples are shown in Fig. 7. Two main broad absorption bands are appeared in the range of 390 to 586 cm−1. The highest wavenumber one is designated ʋ1, observed in the wavenumber range of 565 to 586 cm−1 and this band corresponds to the metal ions intrinsic stretching vibrations at the tetrahedral site. The other band appeared in the range of 392 to 409 cm−1 and it is designated as ʋ2 and corresponds to metal ions stretching vibrations in the octahedral sites. These two bands are observed in previous studies about spinel ferrites24,25,26,27,28.

Figure 7
figure 7

FTIR spectra for the prepared NZF samples.

Full size image

It is seen from Table 6 that there is a slight shifting of ʋ1 towards higher wavenumbers with increase of Ni content. This can be explained the in view of the cation distribution within NZF spinel structure. Zn ferrite has a normal spinel structure where Zn2+ cations occupy the tetrahedral sites (A sites) and Fe3+ cations occupy the octahedral sites (B sites). On the other hand, Ni ferrites have inverse spinel structure in which octahedral sites are occupied by both Ni2+ and Fe3+ cations. Accordingly, when Ni2+ cations are added in expense of Zn2+ cations, they occupy the octahedral sites pushing part of Fe3+ cations from the octahedral to the tetrahedral sites. As Fe3+ cations are smaller in size and lighter than Zn2+ cations, atomic vibration of the tetrahedral sites increases as well giving rise to shifting of the absorption band towards higher wavenumbers. The same shifting of ʋ1 towards higher wavenumbers with increase of Ni was also reported by Tahrani et al.12 and Batoo et al.29. Other bands are observed around 1630 and 3420 cm−1 and the first is attributed to deformation vibrations of adsorbed water during preparation while the second is attributed to O–H stretching indicating the remaining of some hydroxyl groups12,30. The intensity of these two bands are decreased with increase in heat treatment temperature.

Table 6 The effect Ni ratio on the wavenumber of ʋ1 and ʋ2 bands.
Full size table

Figure 8 (from a to d) display the SEM images of the prepared NZF samples. The sample with zero Ni content has non-uniform hexagonal and spherical particles. As Ni content increase, the particles become more spherical in shape and finer in size. A similar effect of Ni content on NZF morphology was reported by Khan et al.31 in the study of the effect of Zn2+dopping on Nickel ferrites.

Figure 8
figure 8

SEM micrographs of NZF samples (a) NZF, (b) N0.25ZF, (c) N0.5ZF and (d) N0.75ZF.

Full size image

Magnetic characterization

Figure 9 depicts the magnetization hysteresis loops obtained by VSM measurement for the prepared NZF samples. The obtained values of Ms, Hc and retentivity (Mr) are shown in Table 7. It is seen from Table 7 that the specific Ms values of 19.8, 28.64, 73.89 and 45 emu/g are observed for the NZF samples N0.0ZF, N0.25ZF, N0.50ZF and N0.75ZF, respectively. The saturation magnetization of magnetic ferrites depends on their chemical composition and the cations distribution within their spinel structure. In NZF spinel lattice, Ni2+ cations prefer to occupy the octahedral sites whereas, Zn2+ cations prefer the tetrahedral sites, and Fe3+ cations are distributed between both octahedral and tetrahedral sites32. The net magnetic moment of the spinel lattice is the difference between the magnetic moments of the octahedral B and the tetrahedral A sub-lattices (M = MB − MA) where MB is the magnetic moment of B sub-lattice and MA is the magnetic moment of A sub-lattice32,33,34,35. As Ni2+ content at the octahedral sites increases, more Fe3+ cations are forced to move from octahedral to tetrahedral sites. As the magnetic moments of Fe3+ cations are higher than that of Ni2+ cations (5 μB and 2 μB, respectively), the replacement of Ni2+ cations for Fe3+ cations at octahedral sites results in a decrease of MB and an increase of MA. The increase of MA with Ni addition produces a higher and super-exchange interaction between A and B sub-lattices which leads to an increase of the magnetization value. However, at a higher content of Ni up to 0.75 the reduction of net magnetic moments between A and B sub-lattices is observed and this can be elucidated based on the high super exchange interaction between the two cations sites.

Figure 9
figure 9

Magnetic hysteresis loops for the prepared NZF samples.

Full size image
Table 7 Magnetic parameters of NZF samples with different ratios of Ni.
Full size table

It is also noticed from the magnetization examinations that sample N0.0ZF with zero Ni content achieves 19.8 emu/g of Ms, which is abnormal as zinc ferrite should be paramagnetic at room temperature because all Fe3+ cations are located in the octahedral sites and all Zn2+ cations are localized in the tetrahedral sites. Therefore, the magnetic spins of the octahedral Fe3+ cations can’t be co-aligned because of the absence of super-exchange between octahedral and tetrahedral sub-lattices36.

The existence of Mn, Mg and Ca impurities in the prepared NZF samples can explain the saturation magnetization of the zero Ni sample where the presence of Ni modifies the cations distribution and the magnetic moments in the NZF spinel lattice. Mn2+ and Ca2+ cations prefer the tetrahedral sites in the spinel lattice, whereas Mg2+ cations prefer the octahedral sites37,38,39,40. As Mn2+ cations with high magnetic moment (5 μB) occupy the tetrahedral sites and Mg2+ cations with zero magnetic moment occupy octahedral sites, MA increases and MB decreases. The increase of the magnetic moments of the tetrahedral A sub-lattices increases their capability to align antiparallel to the magnetic moments of the octahedral B sub-lattices and accordingly enhances the super-exchange interaction between A and B sub-lattices and increases the net magnetic moment. The abnormal saturation magnetization of the sample N0.0ZF can also be attributed to the cations miss distribution in nanosized ferrites. In nanosized NZF some Zn2+ cations may occupy the octahedral B sites instead of their most favorable tetrahedral A sites, leading to shifting of some Fe3+ cations to the tetrahedral sites. This abnormal cation distribution illustrates the increase of saturation magnetization in the zinc ferrite sample13.

On the other hand, the minimum coercivity of 2.55 Oe is observed at Ni content of 0.25 for sample N0.25ZF whereas sample N0.75ZF exhibits the maximum coercivity of 105 Oe. The high value of the coercivity at Ni content was also obtained by Abu El-Fadl et al in NZF prepared by microwave combustion method, and by Chaudhari et al in NZF synthesized by oxalate precursor method. The phenomenon of coercivity increase with increase of Ni content is attributed to the high magnetocrystalline anisotropy energy of Ni ions in comparison to Zn ions13,40. It is also observed from Table 7 that sample N0.0ZF of zero Ni content exhibits higher coercivity compared to N0.25ZF sample (11.41 and 2.55, respectively). It was also observed by Abu El-Fadl et al in the previous mentioned study that the sample with 0.2 Ni content exhibited lower coercivity than that with zero Ni content. This was attributed to the fact that Fe3+ content on the A-site is the highest for Ni content of 0.2 in comparison with Ni content of 0.013.

Conclusion

Single phase nanosized NZF was produced using EAFD and WPL. Treatment of the EAFD with 2% v/v acetic acid reduced its Ca content by about 70% and enriched its content of Fe and Zn. Dissolving of the acetic acid-treated EAFD in hydrochloric acid resulted in removal of non-soluble SiO2 and Al2O3, which improved the purity of the prepared samples. The structural and magnetic properties of the prepared NZF samples are affected by their Ni content and the highest saturation magnetization of 73.89 emu/g was achieved at 0.5 Ni ratio and the lowest coercivity of 2.55 Oe was achieved for Ni content 0.25. The presence of Ca, Mn, and Mg impurities affected the magnetic properties of the prepared samples by readjustment of the cation’s distribution and the magnetic moments within the spinel NZF lattice. The applied process in this paper realized the transformation of the EAFD and WPL form environmentally hazardous waste by products to value-added magnetic ferrites.