Crystal phase evolution of high temperature annealed Fe3O4‐CaO catalysts for biodiesel production

Biodiesel (BD) is an alternative transportation fuel that has been commercialized. Most CaO‐based heterogeneous catalysts that are applied for BD production become strongly deactivated after not more than five cycles of reuse. Predominantly CaO and Ca(OH)2 active phases in eggshell were mixed with Fe3O4 by ball milling for 3 h followed by annealing at different temperatures (200–1000 °C) for 2 h to produce a magnetic catalyst. The catalyst was deployed to synthesize BD using a transesterification reaction with methanol. The catalyst characterization results from X‐ray diffraction and Raman analysis revealed the phase evolution of the catalysts with increasing annealing temperature, which facilitated the oxidation of the Fe3O4 to the maghemite (Fe2O3) phase. Increasing temperature led to a decrease in the catalyst surface area due to increased rigidity and loss of pores; however, the basicity increased due to the conversion of calcium carbonates mainly to CaO. Under optimal conditions of 65 °C, methanol/oil molar ratio of 12:1, and 4 wt% catalyst loading, BD yield of 95.33% could be achieved under 120 min reaction time. The catalyst could be reused seven times with minimal loss in catalytic activity. The synthesized BD showed a calorific value of 39.99 MJkg‐1, a density of 0.9 gcm‐3, a viscosity of 4.92 mm2s‐1, which satisfactorily complied with the international standards of ASTM‐D‐6751 and EN‐14214.


Introduction
T he depletion of oil reserves and the release of greenhouse gases have necessitated the drive toward decarbonization. 1 To achieve decarbonization, a change in attitude toward energy consumption and alternative energy sources is necessary. Among the main alternative energy sources, solar, wind, geothermal, and biomass sources have taken priority. In recent years, investigations have focused more on the production of green hydrogen to contribute to the planet's decarbonization process. 2 The valorization of biomass to produce biofuels such as biogas, bioethanol, and biodiesel (BD) has attracted much attention because of its ease of synthesis, application in heavy-duty transport systems, and application in current diesel engines without modification of the system. The biomass could be obtained from various waste sources including agro-residues.
Biodiesel (fatty acid methyl ester (FAME)) has various advantages and can be produced using waste grease, waste cooking oil, plant oil, animal fats, and microorganisms as oil feedstock. 3 Biodiesel is characterized by higher combustion efficiency, high cetane number, low emission of aromatic and sulfurous compounds, and high flash point (160 °C), in comparison with conventional petroleum diesel. 4 The transesterification reaction of triglyceride with alcohol (ethanol or methanol) to produce alkyl esters and glycerol is popular because it has good thermodynamic properties, and the synthesis can be carried out with or without catalysts. 5 A faster reaction rate, high conversion of the triglyceride to FAME, and milder conditions could be achieved using catalysts. 6 Several catalytic materials have been developed over the years to propagate BD synthesis. These catalysts mainly include homogeneous catalysts of acidic (H 2 SO 4 ) and alkaline (NaOH, KOH) origin, 7 enzymes (lipases), 8 and heterogeneous catalysts. Heterogeneous catalysts are popular because of their easy separation, reusability, and lower corrosiveness in comparison with homogeneous catalysts. Several heterogeneous catalysts have been applied and they can be grouped into acids (zeolites, mixed oxide metals, hetropolyacids, and polyoxometalates) and alkalis (basic zeolites, alkaline earth metal oxides, alkali metals oxides, layered hydroxides). 9 One of the bottlenecks that exist in BD production using heterogenous catalysts is the energy-intensive recovery and stability of the catalysts in the reaction media. For the recovery of the catalyst after each cycle, various methods such as centrifugation, filtration (with paper), and separation by gravity (sedimentation) have been employed. These methods involve high energy consumption, catalyst attrition, long periods, and a loss of the catalyst in each recovery cycle, thereby increasing the operational cost. 10,11 The stability of these catalysts also decreases rapidly as the number of cycles increases due to the severe leaching of active sites. Hence, various researchers have based their research on the development of new materials that can be recovered more easily, without the need to use sophisticated, expensive, and energy-intensive systems for the recovery of the material. Magnetic-based catalysts have been explored and used extensively in various fields.
There are three types of iron-based nanoparticles of interest: α-Fe 2 O 3 , γ-Fe 2 O 3 , and Fe 3 O 4 . 12 Interest is mainly focused on the Fe 3 O 4 phase, also known as magnetite, which is mostly characterized by being intensely black and with a cubic inverse spinel structure. It is also characterized by its easy availability, versatility, biocompatibility, biodegradability, and magnetic properties. 13 Magnetic iron oxide is characterized by superparamagnetism, chemical stability, and low cost. An important characteristic of this material is that at less than 578 °C it becomes ferrimagnetic. Otherwise, the maghemite (γ-Fe 2 O 3 ) presents the same crystalline phase as the Fe 3 O 4 . The main difference between these two oxides is the presence of vacancies in the cationic sublattice with a symmetry reduction. Meanwhile, the α-Fe 2 O 3 phase, also known as hematite, is a compound with a crystalline structure of the type of corundum that has a rhombohedral hexagonal structure. It has strong ferromagnetic properties below 955 K. Hematite becomes Fe 3 O 4 at a temperature of 650 °C with a loss of energy. 14 The magnetic properties of iron-based catalysts allow for easy separation. A simple magnet can be used to remove the catalyst from the reaction medium in a few minutes, which avoids the energy-intensive steps of catalyst separation and recovery. Some reports on improvements to the Fe 3 O 4 /CaO catalyst can be found in the literature. For example, it has been observed that high calcination temperatures are required during the development of the CaO/Fe 3 O 4 catalyst to improve its mechanical resistance against leaching into the reaction mixture during its application in the catalytic reaction. 15 Ali et al. reported that the Fe 3 O 4 catalyst amount and reaction time played an important role in the transesterification reaction to achieve high yields of BD. 16 Mardani et al. also reported that the Fe 3 O 4 acts better as a CaO support due to a greater amount of coating. It also changes the structural properties such as the crystal size, the type of pore shape, and the amount of coating. 17 However, iron-based heterogeneous catalysts still have some limitations.
Iron-based catalysts also have a low resistance to mass transfer due to the small particle size. Using magnetic catalysts, 70-100% conversion has been achieved; however, stability is still problematic and reusability is still low. [18][19][20] The low reusability is usually attributed to particle agglomeration, and solubility of the ferric sulfate in methanol and water media produced, especially during the esterification reaction. 20 Oxides of iron alone cannot significantly promote a high yield of BD in transesterification reactions because of their poor basicity. Hence, to promote higher conversion, alkaline metal oxides are required as the active catalytic phase. For example, impregnation of these iron oxides with CaO could compensate for the low basicity of iron, provide synergistic effect to increase stability, and allow easier separation of the catalyst from the reaction mixture. An estimated 7.2 million tons of eggshell waste is generated every year. 21 These chicken eggshells are mainly made up of calcium carbonate (CaCO 3 ) representing 95% of the eggshell. The other 5% are organic materials, such as sulfated polysaccharides, collagen, and other proteins. During calcination at high temperatures >800 °C, CaO is formed as the main crystalline phase of the eggshell. The calcination process modifies the structural properties and improves the basicity and surface area of the catalyst. 22 Eggshell is also biodegradable, recyclable, and biocompatible, the conversion of this waste into a valuable product represents an additional benefit. 23 This work makes original contributions in the following ways: (1) the production of a magnetic CaO heterogeneous catalyst that is highly reusable for more than five cycles while maintaining a BD yield of above 70%; (2) the evaluation of the crystal phase evolution of magnetic catalyst with hightemperature annealing and establishing the temperatureactivity relationship; (3) avoiding the high energy-intensive separation steps to recover the catalyst by using a simple neodymium magnet; (4) waste minimization and valorization aspects by using waste eggshell, which adds sustainability to the entire BD process, and promotes the United Nations Sustainable Development Goal 2030 of environmental ecofriendliness, and (5) the production of alternative renewable energy.
Hence, in this study, a highly reusable magnetic heterogeneous catalyst was developed using iron oxide nanoparticles (Fe 3 O 4 ) and eggshells (CaO) for the production of BD from soybean oil. The catalyst was characterized to establish the structure-performance relationship. The effect of various reaction conditions was elucidated, and the stability of the produced catalyst was tested in successive reuse cycles. The BD obtained from the transesterification process was subjected to a rigorous characterization in accordance with the requirements of the International Standard of Methods and Materials (ASTM-D6571) and the European Committee for Standardization (EN-14214).

Preparation of CaO
The collected eggshells were washed with hot water and dried at 110 °C for 24 h. Then the material was ground into powder and sieved (25 mm mesh) to obtain a homogenized grain size. Then it was annealed at 900 °C for 5 h in the air to transform the phase from predominantly CaCO 3 to CaO and Ca(OH) 2 .

Preparation of Fe 3 O 4 particles
FeSO 4 and Fe 2 (SO 4 ) 3 were dissolved in deionized water in a 1:1 mol L -1 ratio. After its complete dissolution, concentrated NH 4 OH was added dropwise to increase the pH to 9.5. Then the solution was heated at 80 °C until the complete precipitation of black particles, which were removed from the solution using a magnet. The black particles were washed with ethanol and water and then dried at 70 °C for 6 h.

Preparation of composite CaO/Fe 3 O 4
A solid-state reaction was adopted for the synthesis of the composite catalyst. About 20 g of CaO/Fe 3 O 4 was prepared using 5 wt% of Fe 3 O 4 . The materials were mixed in a ball mill for 3 h, and then 30 mL of ethanol was added to the ballmilled powder. The slurry was subjected to ultrasonication with constant stirring for 1 h. The material was dried at 100 °C overnight to remove ethanol and the composite was annealed in a temperature range of 200 to 1000 °C for 2 h at a heating rate of 10 °C min -1 .
70°. The crystallite size (D) was estimated using the Scherrer formula in Equation (1): where B is the full-width at half-maximum (FWHM) of the peak (in radians), θ is the Bragg angle, k is the shape factor (0.9) and λ is the X-ray wavelength (1.5405 Å). 24 Raman spectroscopy characterization with a confocal Raman system (NTEGRA Spectra NT-MDT, 1000 BV Amsterdam, Netherlands) was performed to determine the phases present in the samples. The excitation source was a laser at 532 nm. The Fourier-transform infrared (FTIR) (Thermo Fisher Scientific -Nicolet, Model: 6700, USA) spectra of the samples were recorded in the wavenumber range of 400 to 4000 cm −1 . Sample preparation was carried out by the KBr method.
Nitrogen isotherms at 77 K (Quantachrome Autosorb-iQ, Anton Paar, Graz, Austria) were used to estimate the surface properties of the samples. The specific surface area (S BET ) was determined by the Brunauer-Emmett-Teller (BET) method (relative pressure range from 0.05 to 0.2). The total pore volume was calculated from the gas adsorbed at a relative pressure of 0.99. The mesopore volume (Vm) was estimated by the Barrett-Joiner-Halenda (BJH) method.
The sample's basicity was determined by titration using 0.1 mol L −1 HCl and NaOH with phenolphthalein indicator. Typically, 1 g of the synthesized catalyst was placed in a 50 mL beaker and then 20 mL of HCl was added. The mixture was continously stirred at 250 rpm for 24 h. Then, the catalyst was filtered and the back titration process with NaOH was conducted until a color change was noticed. The titration was carried out in triplicate and the average volume was taken. The basicity was calculated in mmol/g.
On the other hand, the morphology of the catalyst was studied using a high-resolution scanning electron microscope (FE HRSEM Auriga 3916, Carl Zeiss Ultra, Jena, Germany). The equipment had a Genimi Schottky-type field emission column (range from 0.1 keV to 30 keV).

Biodiesel production
The synthesis of BD from soybean oil was carried out in a three-necked glass reactor equipped with a condenser to prevent the evaporation of methanol. A digital thermometer was used to monitor the temperature of the reaction. The reactor was stirred continuously to enhance the interaction of the reactants during the transesterification reaction. About 20 g of the soybean oil was placed in the reactor and different amounts of methanol were added into the reactor to evaluate the effect of the molar ratio from 1:1 to 18:1. Then various catalyst quantities were added to the methanol/oil reactant. The reaction was started when the setpoint temperature (temperature evaluated from 55 to 70 °C) was reached under different reaction times (30 min to 180 min) at constant stirring (450 rpm).

Catalyst recovery
At the end of the transesterification reactions, the synthesized catalyst was recovered from the mixture by applying a permanent external magnetic field to it. The excess methanol was recovered from the BD produced using a rotary evaporator (Buchi Rotavapor / R-210, Marhsall Scientific, USA) at a temperature of 80 °C for 1 h. The BD was stored in airtight jars for analysis. The catalyst was washed repeatedly with a mixture of solvents (hexane and acetone) and stirred in a vortex (Scientific Industries / Genie II Mixer, New York, USA) for 15 min to remove surface-adhered organic matter. The catalyst was then dried in a drying oven (RedLine by Binder GmbH Tuttlingen, Germany) at 105 °C for 24 h, for subsequent reuse in the transesterification reaction. 25 The same procedure was adopted in subsequent reaction cycles.

Analysis of BD properties
The density, viscosity, calorific value, and pH of the BD obtained were determined based on the American Society for Testing and Materials (ASTM) and EN standard. The biodiesel yield was calculated using Equation (2): 26

Results and discussions
Catalyst characterization

X-ray diffraction analysis
The X-ray diffraction (XRD) pattern provides information on the size and shape of unit cells from peak positions 27 High crystallinity indicates greater catalytic activity and stability, dispersion, and heterogeneity. 28 The XRD of the samples is shown in Fig. 1. The magnetite (Fe 3 O 4 ) phase (PDF 01-071-6337) was revealed in (111), (220), (311), (222), (400), (422), (511), and (400) planes ( Fig. 1(a)). The CaO sample ( Fig. 1(b)) presents a mixture of the phases of CaO (Lime PDF 37-1497) and Ca(OH) 2 (Portlandite PDF 44-1481). However, it is clear from the peak intensity that the predominant phase is CaO. The presence of Ca(OH) 2 could be due to moisture adsorption during sample cooling after the annealing process. At low temperatures, after mixing both precursors in the presence of ethanol, the Ca(OH) 2 increases due to the contributions of the hydroxyl group of the ethanol. However, the increase in the annealing temperature promotes the phase transformation of Ca(OH) 2 into CaO, which possesses stronger alkalinity than Ca(OH) 2 . Similarly, increasing the calcination temperature facilitates the oxidation of the Fe 3 O 4 phase to the maghemite phase (PDF 39-1346), which is a magnetic component that works for the purposes of this research ( Fig. 1(c)).

Raman spectra
The Raman spectra of the samples are presented in Fig. 2 (25 °C), the peak at 365 cm −1 increased in intensity, which is evidence of an increase in the quantity of Ca(OH) 2 (also observed in XRD). 32 The increase in the annealing temperature promotes the reduction of the peak at 365 cm −1 , which suggests that Ca(OH) 2 decomposes to CaO. The appearance of the peak at 970 cm −1 (*) with the increase in annealing temperature also shows the increase in the CaO phase. 32 No peaks of Fe 3 O 4 are observed in the samples, but the broadening of the peak from 678 cm −1 to >700 cm −1 can be explained due to the incorporation of the maghemite phase (*) in the structure. 33 This reveals the oxidation of magnetite into the maghemite phase.

Fourier-transfer infrared analysis
The FTIR spectra of the samples are presented in Fig. 3. In the Fe 3 O 4 sample, the characteristic peaks of the Fe-O bonds were observed at 445 cm −1 , 570-580 cm −1 , and 630 cm −1 . 34,35 The peaks at 1632 and 3425 cm −1 are due to adsorbed moisture. 36 The CaO sample and the composites of CaO/Fe 3 O 4 showed similar behavior, recording strong absorption in the region of 500 to 700 cm −1 , which is related to the Ca-O bonds. 37 The peaks due to CO 2 were observed at wavenumbers 876, 1077, 1090, 1420, 1564, and 1638 cm −1 . These bands are attributed to different C-O bonds and in some cases they are related to the presence of carbonate groups. 38 However, CaCO 3 wasn't detected in XRD and Raman analysis, so it is due to CO 2 absorption. 36 Strong absorption at 3640 cm −1 is because of the O-H bond, which is related to the presence of adsorbed water, which could lead to the formation of the Ca(OH) 2 in the samples. 39,40 Some residues of the membrane eggshell are observed at 1798 and 2516 cm −1 due to the presence of amines and amides. 36 Characterization of morphology (field emission scanning electron microscopy) Field emission scanning electron microscopy (FE-SEM) provides relevant topographical information to show the morphology and surface of the samples. 41 Figure 4 shows the FE-SEM images of the CaO, Fe 3 O 4 , and composites annealed in the range of 200-1000 °C. All samples present a highly heterogeneous appearance without a defined morphology, and have a large size, reaching values greater than 10 μm. The Fe 3 O 4 particles have a very porous appearance and have the smallest particle size among all the samples. The CaO sample presents a compact structure with hollows along the structure due to the elimination of carbonates present in the eggshell. These hollows provide porosity and surface area. However, this high level of porosity in the CaO sample is lost once the composite is formed with Fe 3 O 4 , which is due to    the other hand, an increase in annealing temperature above 600 °C resulted in the loss of the particles' roughness and porosity, with a corresponding increase in crystallinity as observed in XRD analysis of the samples annealed at 800 and 1000 °C (Fig. 1(c)). As crystallinity increased, the edges and grain boundaries of the particles were lost, promoting larger grains and reducing the available contact surface.

Brunauer-Emmett-Teller analysis
The increase in annealing temperature increased the surface properties, reaching a maximum at 500 °C. This is because, at this temperature, volatiles and moisture are eliminated from the catalyst surface and structural water is decomposed, facilitating the transformation of Ca(OH) 2 to CaO. Above 500 °C the surface area, pore volume, and pore radius decrease, which is due to an increase in particle size. For example, Syuhada et al. 42 they reported that the decrease in the surface area could be caused by particle agglomeration, resulting in a larger particle size. A catalyst that presents a larger surface is more active during the catalytic reaction because it presents more active sites so that the reaction can be carried out successfully. 43 The crystallite size inverse behavor when the temperature increases, with smallest value (47.6 Nm) at 500 °C, which validates the observed surface properties. The reduction in particle size also leads to an increase in surface area. Table 1 shows the BET analysis, crystallite sizes, and pore distribution of the samples obtained at different annealing temperatures. Although the annealing temperature of 500 °C gave a greater surface area, the basicity at this temperature was much lower than the basicity at 600 and 1000 °C. Notably, the transesterification reaction to synthesize BD is influenced mainly by the basicity of the catalyst material and not the surface area. Above 600 °C, the catalyst basicity remained constant; hence, 600 °C was selected as the best annealing temperature. At 1000 °C there is a considerable loss of catalyst surface area and catalyst sintering, which could result in a lower BD yield.

Effect of calcination temperature
The calcination temperature plays an important role in the catalytic performance because its structural properties are improved and new active sites are generated. 44,45 Various temperatures from 200 °C to 1000 °C for 4 h were evaluated to investigate the effect of calcination temperature (Fig. 5). To  select the best calcination temperature, BD was synthesized under the following reaction conditions: a methanol to soybean oil molar ratio of 6:1; catalyst loading of 1 wt%; reaction temperature of 65 °C; reaction time of 90 min. The results revealed that an increase in the calcination temperature from 200 °C to 600 °C increased the BD yield from 75.30% to 88.86%. Above 600 °C, the BD yield decreased. This can be attributed to the collapse of its structural properties. The structural collapse is mainly due to the sintering effect, which occurs at high temperatures. 46 Effect of catalyst amount on the BD yield Figure 6(a) shows the effect of the catalyst loading varied from 1 wt% to 5 wt%. The BD yields gradually increased from 83.3% to 88% when the amount of catalyst was increased from 1 wt% to 3 wt% and reached a maximum of 90.66% at 4 wt%. The initial increase in BD yield is because of an increase in the available catalytic active sites, which facilitates the abstraction of protons from methanol and the subsequent reaction of the methyl group with the free fatty acids of the triglyceride to form fatty acid methyl esters and glycerol. A further increase in the amount of the catalyst to 5 wt% affected the BD yield, reducing the yield to 79.69%. The decrease in yield is due to high viscosity, which does not benefit the mass transport of the reactants to the active sites.
It is also possible that increasing the amount of catalyst could result in catalyst particle agglomeration and emulsification. The phenomena of emulsification could result in the formation of less active phases especially when in contact with the glycerol Ca(C 3 H 7 O 3 ) 47 and methanol Ca(OCH 3 ) 2 by-products. 48 A large amount of energy is also Original Article: Fe 2 O 3 -CaO catalysts for BD production required for stirring and longer reaction time to maintain sufficient contact between the reactants and the catalyst. 49 A reported study has demonstrated that the highly viscous nature of vegetable oil could present a barrier to mass transfer and results in lower yields. 50 Effect of reaction temperature on the yield Increasing the reaction temperature during the production of BD leads the reaction kinetics of the transesterification reaction to produce FAME. 51 The reaction temperature was varied from 50 °C to 65 °C with a constant catalyst load of 4 wt%, a molar ratio of methanol/soybean oil of 6:1, and a reaction time of 90 min ( Fig. 6(b)). The results show that the BD yield increased as the temperature was increased and reached a maximum of 95.47% at 65 °C. Above this temperature, the BD decreased to 82.3% due to a considerable loss of methanol by evaporation; this remains in the vapor phase inside the reactor causing less interaction with other reactants. Hence, 65 °C was selected as the optimal temperature for the BD synthesis. Figure 6(c) shows variation in the conversion of soybean oil to BD as a function of reaction time. The transesterification reaction of triglyceride and methanol is an equilibriumbased reaction, that requires time for reactant interactions to generate products. Evidently, a short reaction time of 30 min was insufficient to achieve a higher BD yield. The maximum BD yield reached in this time was about 60.66%. When the time was increased from 60 to 120 min, the BD yield gradually increased, reaching a 95.32% yield in 120 min. Above 120 min reaction time, the BD yield suffered a decline, maintaining 75.00% at 180 min. The loss in catalytic activity is because of secondary reactions that could facilitate the hydrolysis of esters. 52,53 Hydrolysis or soap formation are undesired side products in the transesterification to produce BD. Aslan and Eryilmaz 54 revealed that above the optimal reaction time, ester hydrolysis, which is a reverse reaction, could be promoted, resulting in decreased BD yield. Likewise, soap formation due to interaction of the alkali catalyst and FAMEs could reduce the BD yield. 55 Hence, 120 min was selected as the best suitable reaction time to achieve the best BD yield.

Effect of molar relation on the yield
The transesterification reaction requires more than an equimolar amount of methanol to achieve higher triglyceride conversion successfully. The effect of the molar ratio of methanol to soybean oil was studied from 6:1 to 18:1 and the results are shown in Fig. 6(d). It was observed that the BD yield increased as the molar ratio of methanol to soybean oil increased from 6:1 to 9:1 with values of 80% and 88.78%, respectively. The optimal conversion of 96.45% was achieved when a molar ratio of 12:1 was used. An additional increase in the molar ratio did not improve the yield; rather, the yield decreased because of insufficient interaction between the catalyst sites and the reactant, which was propagated by high reaction volume. 56

Catalyst reuse cycles
Catalyst stability affects process economics and is an important criterion for measuring the suitability of heterogeneous catalysts in industrial applications. 57 The catalytic stability of the catalyst synthesized in the present research work was assessed using optimized variables (4 wt% catalyst load, temperature of 65 °C, molar ratio of 12:1, and a reaction time of 120 min). The catalyst could be reused for seven consecutive cycles as seen in Fig. 7(a). The highest conversion yield in BD production was 95.33%, which was recorded during the first run and was maintained until the fifth cycle, then it gradually decreased, sustaining a 77.2% BD yield at the seventh cycle. Figures 7(b)

Catalyst deactivation
The deterioration in the catalytic loss of the synthesized material could be attributed mainly to the loss of mass of the CaO/Fe 3 O 4 catalyst due to the use of an external magnetic field. This phenomenon has been reported previously in other works. 44 The accumulation of organic constituents of the reactants on the surface of the catalyst such as unconverted triglycerides may mask the catalyst's active site. Another possible cause of the deterioration of the catalyst activity could be the leaching of the CaO, which results in the loss of active sites. Leaching deactivates active sites and reduces the activity of the magnetic heterogeneous catalyst. The heterogeneous catalyst was evaluated again for its basicity; it was observed that it decreased considerably from 12.00 mmol g -1 to 9.3 mmol g -1 , which means that the material suffered a loss in the number of its active sites found on the surface of the catalyst. Meanwhile, BET analysis performed on the spent catalyst revealed that the synthesized catalyst suffered a decrease in its surface area from 9.839 to 6.743 m 2 g -1 . Both the decrease in basicity and the loss in the surface area of the catalyst could be caused mainly by trace contamination of unconverted triglycerides, BD, and glycerol, crystalline phase change, and leaching of CaO. 58 For instance, Yan et al. 59 reported that one of the most common problems that usually arise when working with CaO-based catalysts is Ca leaching during transesterification reactions for BD production, causing the loss of basicity and basic strength on the surface of the catalyst.
To evaluate any change in the crystal phase, an XRD analysis was conducted as shown in Fig. 8. The analysis detected the presence of three different phases in the spent catalyst. As expected, the CaO compound was poisoned by moisture and CO 2 present in atmospheric air. This process is known as rehydration and it transforms the phase of the conversion of CaO to Ca(OH) 2 due to the atmosphere that the catalyst is exposed to during storage and handling processes. The Ca(OH) 2 is characterized by a lower catalytic activity because of its lower basicity compared with CaO. 60 The existence of this compound was confirmed by PDF file: 44-1481 with the crystal planes of (1 0 0), (1 0 1), (1 1 0), (1 1 1), (2 0 0), and (2 0 1) exhibiting a hexagonal crystal structure.
The other compound detected in the catalyst was CaCO 3 or calcite with a rhombohedral crystal structure, with crystal planes of (0 1 2)    These phase changes did not affect its magnetic properties because it is still a material that is attracted by an external magnetic field. The predominant crystal planes of this compound were (1 1 1), (2 2 0), (3 1 1), (2 2 2), (4 0 0), (4 2 2), (5 1 1), and (4 4 0). Hence, the crystalline phase changes accelerated the gradual loss in the catalytic performance of the catalyst, thereby decreasing the yield of BD.

Comparison with other magnetic heterogeneous catalysts
The catalyst presented in this work has been compared with reported works on magnetic catalysts employed for the synthesis of BD using the transesterification reaction. Table 3 shows that the CaO/Fe 2 O 3 catalyst was efficient under moderate reaction conditions, was more reusable, and converted in a very similar way to the other catalysts reported in the literature. We can therefore conclude that the CaO/Fe 3 O 4 heterogeneous catalyst is an ideal candidate for the transesterification reaction at the pilot and industrial scale levels.

Physicochemical properties of BD
To determine the viability of the BD produced, it was tested using various standard methods (ASTM-D6571 and EN-14214). The physicochemical properties of the soybean oil and the BD obtained are shown in Table 4. The values indicate that the properties of the BD synthesized from the CaO/ Fe 3 O 4 catalyst satisfactorily complied with the standards of the American Society for Methods and Materials (ASTM) for BD viscosity, density, calorific value, acid number, and pH. Hence, the synthesized BD in this work could be used in all types of internal combustion engines, with good environmental benefits, including the reduction of greenhouse gases.

Conclusions
In this research work, a magnetic calcium oxide catalyst derived from eggshell and iron oxide was successfully synthesized by ball milling and subsequent annealing at high temperatures. Results reveal that, at 900 °C, for a period of 5 h, eggshell can be transformed from predominantly CaCO 3 to CaO with higher basicity. The combined CaO/ Fe 3 O 4 annealed at different temperatures (200 to 1000 °C) showed the best BD yield to occur under optimal conditions of 600 °C for 2 h. The catalyst provided sufficient basicity to tailor the transesterification reaction of triglyceride and methanol to achieve 95.33% BD yield under the best conditions of soybean oil molar ratio of 12:1, a catalyst load of 4 wt%, a reaction temperature of 65 °C, and a reaction time of 120 min. The catalyst was heterogeneous, easily separated by a magnetic field, and maintained a relatively high yield of BD of 77% after seven reuse cycles. The BD produced has a density of 0.9 g cm -3 , a pH of 6, and a calorific value of 39.99 MJ kg -1 , which is similar to the established ASTM and EN standards.