1. Introduction

Chemically, biodiesel is a relatively complex mixture containing fatty acid esters, usually obtained through the transesterification reaction of triacylglycerols in bio-oils, with alcohols of short molecular chains, almost uniquely methanol or ethanol. In the industrial process, the chemical reaction is homogeneously catalyzed with a strong alkaline base, as potassium hydroxide, in the liquid medium ^[1,2,3]. This is by far the most common chemical industrial process worldwide. However, any solid heterogeneous catalyst is conceptually thought to lead to cleaner processes, relative to the liquid homogeneous catalyst, on such transesterification reactions, as the solid material requires much less water to clean the produced biodiesel from impurities relatively to that from the homogeneous route and the catalyst can be removed and reused in subsequent reaction cycles. Some recently reported reviews ^[4,5,6,7] hardly mention the chemical role of iron oxide-based solid catalysts towards the transesterification reaction to produce fatty acid methyl or ethyl esters. Regarding the envisaged industrial process, it would be still more interesting if the solid is magnetic as it can be more easily removed from the reaction medium with a magnetic field from a permanent magnet or generated with an electric coil ^[8]. Magnetic materials ^[9] are thus conceptually strong candidates to be valuable components of solid catalysts, taking into account their potential chemical catalytic activity and the feasibility of being easily separated ^[8], with an applied magnetic field, from the liquid mass in the industrial line, whenever it is required.

From the economic and environmental points of view, mining rejects containing magnetic minerals may be largely available depending on the geology and on the ore matrix being processed. In most cases, the rejects represent a real threat to the surrounding natural environment, particularly in areas around the mine. Such a specific circumstance may be found in tailings from some phosphate ores containing a given proportion of magnetite.

This report describes a work devoted to the use of the magnetic fraction of the mining reject of the Tapira Mining Complex, in the state of Minas Gerais, Brazil. The magnetic fraction from the rejected material from extracting the mainly exploited ore is ordinarily separated during the mining industrial process. A sample of that magnetic fraction was further magnetically enriched with a hand magnet and tested in the laboratory as a component of a heterogeneous catalyst, either straightly or mixed with calcium oxide, on the transesterification reaction of triacylglycerols in the soybean oil with methanol, to produce biodiesel. The chemical performance of the heterogeneous catalyst based on this magnetic fraction mixed with CaO was evaluated by comparing with that of the pure magnetic fraction and with that of the sole CaO.

---

2. Materials and Methods

The magnetic fraction was separated from tailings of an ore mine for phosphate in the municipality of Tapira, state of Minas Gerais, Brazil.

The industrial magnetic fraction was still further magnetically enriched in the laboratory by separating magnetic particles with a hand-held permanent magnet and this fraction was mixed with calcium oxide PA (Aldrich) at a mass ratio 1:1, in a porcelain crucible. This solid mixture was heated in a muffle at 200 ℃ for 4 h. The so-prepared catalyst was stored in a desiccator until being used in the transesterification reaction. The used raw source of triacylglycerols was a commercial soybean oil destined to domestic use, purchased in a supermarket.

A primary mineralogical characterization of the mineral catalyst sample was performed by powder X-ray diffraction (XRD) in a Rigaku diffractometer model Ultima Ⅳ set to a current of 30 mA and a voltage of 40 kV, with the CuKα (λ = 1.541838 Å) radiation, at a scan rate of 1° 2θ min^–1, from 4 to 80° 2θ. Silicon was used as an external standard.

The chemical states of iron were assessed by Mössbauer spectroscopy, making it also possible to infer about the proportions of the iron minerals (more specifically magnetite, hematite or ilmenite) in the sample. The Mössbauer spectrum was collected at room temperature (~298 K) in the constant acceleration transmission mode with a ~20 m Ci ⁵⁷Co/Rh gamma-ray source. Data were stored in a 512-channel MCS memory unit, with Doppler velocities ranging between approximately ±12.1 mm s^–1. Mössbauer isomer shift values are quoted relative to an α-Fe foil at room temperature. The experimental data were least square-fitted to Lorentzian-shape resonance lines, with the WinNormos™ for Igor™ Pro computer program.

The magnetization measurement of the sample was performed with a vibrating sample magnetometer (VSM LakeShore 7404) with a noise base of 5 × 10^–5 emu, a time constant of 100 ms, at room temperature, and a maximum magnetic field of 2 T.

The X-ray fluorescence analysis of the sample was made with a Shimadzu EDX-720 energy dispersive X-ray fluorescence spectrometer, with a rhodium tube and silicon-lithium detector. Data were collected without vacuum, with a collimator of 10 mm.

The triacylglycerol conversion to fatty acid methyl esters was obtained by using a molar ratio methanol:oil 30:1. The catalyst corresponded to 8 mass% of the oil mass. The reaction was monitored with thin layer chromatography (TLC) by using hexane:ethyl acetate at the ratio 9:1 as an eluent and iodine as developer. The reacting liquid was rotaevaporated and the methanol was recovered. The fraction containing glycerin and biodiesel was placed in a separation funnel. The glycerol was then removed; the volume of the biodiesel was measured and subsequently washed with distilled water.

Samples from the produced biodiesel were analyzed according to the European Standard, by gas chromatography. Analyses were performed on a gas chromatograph model HP7820A equipped with flame ionization detector. The column was the BP20 (SGE) 12 m × 0.25 mm × 0.2 µm in a temperature gradient of 10 ℃ min^–1, from 120 ℃ to 220 ℃; gun (split of 1/50) to 250 ℃ and detector to 260 ℃. The gas carrier was hydrogen (3 mL min^–1). Data acquisition software: EZChrom Elite Compact (Agilent).

---

3. Results and Discussion

---

3.1. Powder X-ray diffraction

The powder X-ray diffraction pattern (Figure 1) revealed the occurrence of magnetite (Fe₃O₄; according to the JCPDS card #19-629 ^[10]), hematite (αFe₂O₃; JCPDS card #33-664), ilmenite (FeTiO₃; JCPDS card #29-733), fluorapatite (Ca₅(PO₄)₃F; JCPDS card #15-876), anatase (TiO₂; JCPDS card #21-1272), calcite (CaCO₃; JCPDS card #5-586), zeolite (K_69.8Al_69.8Si_122.2O₃₈₄; JCPDS card #26-893) and quartz (SiO₂; JCPDS card #46-1045). The cubic unit cell edge dimension for this magnetite was found to be a = 8.394 (3) Å, as estimated with the UnitCell computer program ^[11], based on the Kα₁ reflection peaks corresponding to the planes (220), (311), (400) and (511).

---

3.2. X-ray fluorescence analysis

The X-ray fluorescence data reveal that the magnetic material is relatively rich in iron, titanium, silicon, calcium and phosphorus (Table 1).

---

3.3. Mössbauer spectroscopy

Numerically fitting the 298 K-Mössbauer spectrum provided information on some chemical characteristics of the material: (ⅰ) Fe³⁺; (ⅱ) mixed valence Fe^3+/2+ and (ⅲ) a Fe²⁺ doublet.

From the hyperfine parameters (Figure 2; the corresponding spectral parameters are presented in Table 2), the sample contains a magnetite (relative subspectral area, RA = 24.6 + 40.3 = 64.9%), with magnetic hyperfine field values for the tetrahedral [B_hf] = 49.42 (3) tesla and octahedral {B_hf} = 46.17 (2) tesla sites, along with a hematite sextet (RA = 25.4%), B_hf = 51.98 (3) tesla, and an ilmenite doublet (RA = 9.7%) with quadrupole splitting Δ = 0.88 (2) mm s^–1 and isomer shift δ/aFe = 1.04 (1) mm s^–1.

---

3.4. Magnetic measurement

From the room temperature magnetization curve (Figure 3), the saturation magnetization for this sample is M_s ~47 emu g^–1. Iron represents 81.1 mass% of all determined chemical elements (normalized values) in the sample (Table 1). The relative Mössbauer spectral area due to magnetite is 64.9 % (Table 2). If no difference of the recoilless fractions corresponding to the various ⁵⁷Fe sites is taken into account, a rough estimation of the mass proportion of magnetite in the sample is ~52.6 mass%. If so, the saturation magnetization for this magnetite is M_s ~89 emu g^–1. This value is reasonably close to the reported value for a pure bulk magnetite, or 92–100 emu g^{–1 [12]}.

---

3.5. The transesterification reaction

According to these results, this magnetic material combined with CaO showed a significant chemical catalytic activity on the transesterification, yielding 99 ± 1 mass% esters, once the reaction was complete, in a 135 ± 30 min-process (Table 3). To compare, the sole magnetic material from the mining tailings has not produced any ester on transesterification, even after 24 h reaction. On the other hand, the reaction catalyzed with sole CaO yielded 84 ± 10 mass% esters, after reaction completion, in 128 ± 16 min (Table 3).

The heterogeneous catalyst composed by the magnetic material mixed with CaO leads to high chemical yields in esters through the transesterification, and is readily recoverable to be reused, keeping its catalytic activity virtually unchanged for at least two subsequent reaction cycles, with chemical yields of 99.5 and 99.9 mass% esters for corresponding reaction times of 90 min each.

The chemical mechanisms involved the synergic action of the catalyst by this magnetite-rich mining reject mixed with calcium oxide are not fully understood, so far, and become an essential topic for further studies. From a comparable magnetic heterogeneous catalyst composed by SiO₂, γFe₂O₃ and KI promoted the transesterification of triacylglycerols in soybean oil to yield 94.4 mass% esters ^[8].

---

4. Conclusion

The magnetically-enriched fraction (saturation magnetization, M_s ~47 emu g^–1) from tailings of the phosphate mining in Tapira, Minas Gerais, Brazil, was found to contain magnetite (Fe₃O₄, corresponding to M_s ~89 emu g^–1), hematite (αFe₂O₃), ilmenite (FeTiO₃), fluorapatite (Ca₅(PO₄)₃F), anatase (TiO₂), calcite (CaCO₃), zeolite (K_69.8Al_69.8Si_122.2O₃₈₄) and quartz (SiO₂). A sample of this fraction mixed with CaO at a mass ratio 1:1 and heated at 200 ℃ for 4 h was used as heterogeneous catalyst in the transesterification reaction of triacylglycerols in the soybean oil with methanol, to produce methyl esters of fatty acids (biodiesel). Results point to an effective synergic chemical catalytic effect of this mining tailing with CaO to produce biodiesel. The relatively short time of 135 ± 30 min to complete the reaction and the expressive chemical yield of 99 ± 1 mass% esters make this heterogeneous catalyst a strong candidate to be scaled up for industrial chemical processes on the production of biodiesel. The chemical mechanisms for the so catalyzed transesterification reaction involve a synergic action of the catalyst by the magnetite-rich mining reject mixed with calcium oxide. This would represent, on one side, an enormous environmental meaning, as it tends to clean the tailings material from the mining area, and, on the other, an economic gain by a rational use of the disposable material towards a technologically advanced catalyst. Further developments on these magnetic materials are strongly envisaged and are currently in continuous progress by our research Group.

---

Acknowledgments

Work supported by FAPEMIG (Brazil; grant #CEX-PPM-00412-15). JDF is indebted to CAPES (Brazil) for granting his Visiting Professorship at UFVJM under the PVNS program and to CNPq for the grant #305755-2013-7. The authors are also indebted to Mr Abraão José Silva Viana (UFVJM) for their technical assistance on the EDXRF analysis and to Mr João Batista Santos Barbosa (CDTN) for their technical assistance on obtaining the powder DRX.

---

Conflict of Interest

All authors declare no conflicts of interest in this paper.

---