Tungsten oxide supported on copper ferrite: a novel magnetic acid heterogeneous catalyst for biodiesel production from low quality feedstock

This study aims to synthesize a WO3/CuFe2O4 catalyst through a wet impregnation method and use it as a new magnetic acid catalyst in the transesterification process of waste cooking oil (WCO). The results of the characterization by XRD, FTIR, SEM, EDS, TG/DTG, VSM and Surface Acidity showed that the obtained bifunctional catalyst has been successfully synthesized. The study of the reaction parameters, such as reaction temperature (140–180 °C), reaction time (1–5 h), molar ratio MeOH : oil (25 : 1–45 : 1) and catalyst loading (2–10% m m−1) was performed in the conversion of WCO into biodiesel via transesterification. The reactional behavior showed the following optimal reaction conditions: reaction temperature of 180 °C, reaction time of 3 h, molar ratio MeOH : oil of 45 : 1 and catalyst loading of 6%. Based on the results, biodiesel with a maximum ester content of 95.2% was obtained using the WO3/CuFe2O4 magnetic catalyst under the optimal reaction conditions. The magnetic catalyst showed excellent catalytic and magnetic performance and it was applied in five reaction cycles with ester content above 80%. Biodiesel properties were found in accordance with ASTM limits. This research provided the development of a stable and reusable WO3/CuFe2O4 bifunctional catalyst for potential application in biodiesel production.


Introduction
In recent decades, new energy sources that are renewable and less polluting have been researched and compared to fossil fuels, widely used today. 1,2 In this sense, biodiesel emerges as one of the most promising energy alternatives, since it is a renewable, biodegradable and considerably less polluting fuel than petroleum diesel. 3,4 Chemically, biodiesel could be dened as alkyl esters, produced through the transesterication and esterication reactions of triacylglycerols in the presence of short-chain alcohols, such as methanol and ethanol. 4 In the literature, several sources of triacylglycerol have been applied in fuels production, edible and inedible oils, algae, etc. [5][6][7] In addition, the use of residual oilseed matrices has been studied in order to reduce biodiesel production costs, since the raw materials used in biodiesel production represent about 60-75% of the process total cost. 8,9 In that regard, biodiesel production is reported in the literature through the use of residual sources of triglycerides, such as animal fats, waste cooking oils, etc. 9,10 Transesterication and esterication reactions are usually performed in the presence of homogeneous, heterogeneous or enzymatic catalysts. 11,12 Heterogeneous catalysts have been attracting attention due to the possibility of being easily recovered and reused in more than one reaction cycle, which results in the reduction of the biodiesel production process nal cost. 12,13 The main types of heterogeneous catalysts applied to synthesis of compounds with high value added, such as biodiesel, reported in the literature are: zeolites, biochar carbons, mesoporous silicas, metal oxides, etc. [14][15][16][17][18][19] In addition, among the class of heterogeneous catalysts, the catalysts with acidic properties stand out because it is possible to apply them, in esterication and transesterication reactions, in the presence of residual raw materials without causing inconveniences, such as soap production, reactor corrosion, etc. 12,20,21 In this sense, the use of metal oxides, such as tungsten oxide (WO 3 ), in biodiesel production is reported in the literature due to its strong acidity of Brønsted and Lewis. 22 The work developed by Xie and Yang studied the application of a catalyst composed of WO 3 supported in AlPO 4 in the production of biodiesel from soybean oil. The results presented a ester conversion of 72.5% in the following reaction conditions: reaction temperature of 180°C, MeOH : oil molar ratio of 30 : 1, catalyst loading of 5% and reaction time of 5 h. 23 Among the heterogeneous catalysts magnetic materials have been attracting attention due to the fact that their magnetic properties facilitate the separation step of the catalyst from the reaction medium (usually using techniques such as ltration and centrifugation), through the application of a magnetic eld, which would reduce the costs employed in the separation stage and consequently the total costs of the process. 24,25 Several materials have been applied as catalysts or catalytic supports in the production of biodiesel, such as magnetites (Fe 3 O 4 ), hematites (Fe 2 O 3 ) and ferrites (MFe 2 O 4 ), where M is the transition metal. 25,26 Seffati et al. studied biodiesel production through the use of chicken fat and methanol. The catalyst used in the reaction consisted of the impregnation of calcium oxide (CaO) in copper ferrite (CuFe 2 O 4 ) and the biodiesel obtained ester content of 94.52% in the following optimal reaction conditions: reaction temperature of 70°C, MeOH : oil molar ratio of 15 : 1, 3% catalyst loading and reaction time of 4 h. 8 This study aims to study the application of a heterogeneous magnetic acid catalyst, composed of WO 3 impregnated in cooper ferrite (CuFe 2 O 4 ), in the production of biodiesel using waste cooking oil and methanol. The effects of the variables present in the transesterication reaction were investigated in the catalytic activity of the catalyst, such as reaction temperature, reaction time, molar ratio MeOH : oil and catalyst loading. It is noteworthy that biodiesel synthesis by magnetic acid catalyst using CuFe 2 O 4 as catalytic support has not been reported in the literature yet. O at molar ratio 1 : 2 (Cu : Fe) were dissolved into 150 mL of distilled water followed by mechanical agitation at room temperature during 30 min. Then, a solution of NaOH 4 mol L −1 was added drop by drop to adjust the pH of the mixture to 12. At the end of the precipitation process, the system was kept under mechanical agitation at 65°C for 4 h. Thus, the obtained product was washed with distilled water several times until the washing water achieve neutral pH (pH = 7), and nally, the material was dried in an oven at 80°C for 12 h and calcined at 500°C for 3 h (10°C min −1 ) in order to obtain the CuFe 2 O 4 .

Experimental
2.2.2. Impregnation of the active phase. In the process of synthesis of the magnetic acid catalyst, the method of wet impregnation was applied by using the same tungsten precursor and calcination temperature previously described by Kaur et al. 27 A series of WO 3 impregnated CuFe 2 O 4 catalysts was prepared (20-40% WO 3 loading), where the catalyst with 35% of WO 3 on the support was chosen as the best catalyst (Fig. S1, see ESI †). In a typical procedure, approximately 0.97 g of Na 2 WO 4 -$2H 2 O was dispersed in 10 mL of distilled water to obtain 35% of the metal W on the surface of the support. Then 1.0 g of CuFe 2 O 4 was added to the system. The mixture was kept under constant mechanical agitation for 2 h at room temperature. Then, the material was dried in an oven at 80°C/12 h and calcined at 700°C/3 h (10°C min −1 ). Fig. 1 illustrates the schematic diagram of the catalyst preparation, designated as WO 3 /CuFe 2 O 4 .

Transesterication reaction
The reactions were performed in a PARR 5000 Multireactor reactor with xed agitation at 700 rpm, the following reaction conditions were investigated: reaction temperature (140-180°C ); reaction time (1-5 h); MeOH : oil molar ratio (25 : 1-45 : 1) and catalyst loading (2-10% m m −1 ). Aer the reaction, the catalyst was separated by the application of the external magnetic eld to the reaction system. The reaction products were transferred to a funnel, separated and washed with 500 mL of distilled water (60°C) for removal of residual alcohol and glycerol. Finally, the biodiesel samples were stored for further analysis.

Determination of biodiesel properties
The methyl ester content of biodiesel samples was determined by gas chromatography according to the methodology adapted from the European standard (EN 14103) proposed by Silva et al. 29 In this method, the Varian gas chromatograph, model CP 3800, equipped with Flame Ionization Detector (FID) and capillary column CP WAX 52 CB (30 m long, 0.32 mm in diameter and 0.25 mm lm) was used. Methyl heptadecanoate was used as an internal standard and heptane as a solvent. In addition, the oven temperature programming from 170°C to 250°C (same FID temperature) was used at a rate of 10°C min −1 , helium gas was used as a mobile phase with ow of 1.0 mL min −1 , and the injection volume was 1 mL of sample. The ester content (EC) was calculated according to eqn (1): where: P A T is the sum of the peaks total area; A IS is the peak area of the internal pattern; C IS is the solution concentration of the internal standard (mg L −1 ); C B100 is biodiesel's concentration aer dilution (mg L −1 ).
The main physicochemical properties of biodiesel were determined by a standard method of the American Society for Testing and Materials (ASTM). The kinematic viscosities, analyzed at 40°C , were determined for the biodiesel samples synthesized according to the ASTM D445 method, using a viscometer model Cannon-Fenske (SCHOTT GERATE, 520 23). The density, analyzed at 20°C, was measured by ASTM D6890 method on a KEM DAS-500 automatic densimeter. The acid value was evaluated conforming to the ASTM D664 method. The ash point was estimated employing the ASTM D093 method on an automatic TANAKA APM 7 Pensky-Martens ash point. The cold lter plugging point was determined using ASTM D6371 methodology on a TANAKA equipment AFP-102 model. The corrosiveness to copper was appraised by ASTM D130 in a copper corrosion bath from Koehler. The oxidative stability was determined using a Rancimat, model 743 from Metrohm, in accordance with EN 14112.

Reuse study
The reuse study was conducted for the magnetic catalyst under optimal transesterication reaction conditions. Aer each reaction cycle, the catalyst was separated from the reaction medium by the application of the external magnetic eld, washed once with heptane (10 mL) and twice with ethyl alcohol (25 mL) and dried in an oven at 80°C for 12 h.
The process of recalcination of the magnetic catalyst was proposed by observing the decrease of the ester content values during the reaction cycles. This study was conducted in two stages. First, the catalyst was reproduced and washed aer each reaction cycle, as described above. Then, the catalyst was thermally reactivated by calcination at 500°C for 3 h, in a muffle oven, to eliminate organic components from the surface of the material.

Physical-chemical evaluation of the magnetic catalyst
The CuFe 2 O 4 support and WO 3 /CuFe 2 O 4 magnetic catalyst were characterized by XRD, FTIR, SEM/EDS, TGA/DTGA, VSM and Surface Acidity techniques.
3.1.1. XRD analysis. The X-ray diffraction patterns of CuFe 2 O 4 , WO 3 and WO 3 /CuFe 2 O 4 materials are presented in Fig. 2. The material corresponding to the CuFe 2 O 4 phase (red line) exhibited peaks at 2q = 18.44°, 29.98°, 35.67°and 62.49°, which were well correlated to crystallographic planes (111), (220), (311) and (440), with cubic system and spatial group Fd3m (ICDD 01-077-0010), conrming the formation of CuFe 2 O 4 with reverse spinel structure, in which Fe 3+ ions are situated in an equivalent way in the tetrahedral and octahedral sites, Cu 2+ is only in octahedral sites. 30,31 In addition, the diffractogram also showed peaks at 2q = 38.87°and 48.95°corresponding to the planes (111) and (−202) of CuO with monoclinic system (ICDD 03-065-2309). According to Nikolić, the formation of this impurity may be related to the oxidation of Cu 2+ ions during the synthesis process carried out under atmospheric conditions. 32   The band at 462 cm −1 is attributed to metal-oxygen elongation (M-O, M = Cu or Fe), specic to spinel ferrite. 8 34,35 Thus, the analysis of the data from the FTIR spectra presented suggest the efficiency of the ferrite synthesis process and the impregnation of the WO 3 species in the magnetic support to obtain the bifunctional character of the catalyst.
3.1.3. SEM and EDS analysis. The surface morphology of the magnetic materials CuFe 2 O 4 and WO 3 /CuFe 2 O 4 was examined by SEM analysis (Fig. 4). The SEM micrographs of the CuFe 2 O 4 magnetic support and WO 3 /CuFe 2 O 4 magnetic catalyst revealed that both materials have structures composed of particle clusters of different shapes and sizes (#5 mm). The interaction between magnetic particles is responsible for the morphological nature (clusters) of these materials as reported in previous studies. 36 The SEM micrograph referring to CuFe 2 O 4 ( Fig. 4a), evidences the rough and spongy appearance in the analyzed region. It is noteworthy that this characteristic is also reported in the study of nanocatalyst synthesis based on CuFe 2 O 4 developed by Rajput et al. 37 In addition, when analyzing the SEM micrograph of the WO 3 /CuFe 2 O 4 catalyst (Fig. 4b), the disappearance of the spongy aspect is observed, this may be related to the dispersion of the WO 3 phase over the support.  Fig. 5b, indicating the efficiency of the used synthesis process by coprecipitation. Fig. 6a represents the results of elemental composition of the WO 3 /CuFe 2 O 4 catalyst. In the analysis, the presence of the tungsten element (W) on the surface of the catalytic material   Fig. 6b, it is possible to verify that element W is relatively well dispersed on the surface of the catalytic support, and this uniform distribution of W is essential for the catalytic activity of the magnetic catalyst developed.
3.1.4. TG/DTG analysis. Fig. 7 illustrates the thermogravimetric proles of CuFe 2 O 4 and catalyst WO 3 /CuFe 2 O 4 noncalcined. The results showed two main stages of mass loss for the materials CuFe 2 O 4 and WO 3 /CuFe 2 O 4 . For the CuFe 2 O 4 magnetic support (Fig. 7a), the rst stage of thermal decomposition occurs in the temperature range of 55-167°C, corresponding to the mass loss of 17%, which is associated with the removal of physically adsorbed water and hydroxyl groups on the surface of the material, in addition to the decomposition of organic components such as the AcO − ion (acetate) from the metallic precursor. 38,39 A small mass loss, 0.6%, was observed in the temperature range of 674-694°C, and may be related to the phase transition from tetragonal structure to cubic structure, indicating the development of a more stable phase for ferrite CuFe 2 O 4 . 38,40 However, the heat treatment at 500°C made the material CuFe 2 O 4 stable.
The TG curve of the WO 3 /CuFe 2 O 4 catalyst (Fig. 7b) shows the rst mass loss event of 1.7% around 56-96°C referring to the structural water loss of the material surface. The increase in temperature leads to a reduction in material mass of approximately 1.6% in the range of 319-377°C, due to the formation of the crystalline phase of WO 3 of monoclinic structure, becoming stable above 377°C. 41 Thus, the results showed that the WO 3 / CuFe 2 O 4 catalyst has greater thermal stability when compared to the magnetic support CuFe 2 O 4 .
The magnetic properties investigated were saturation magnetization (M s ), remaining magnetization (M r ) and coercivity (H c ). Based on the results obtained, the CuFe 2 O 4 support (red curve) and the WO 3 /CuFe 2 O 4 catalyst (blue curve) showed saturation magnetization values of 22.70 emu g −1 and 12.04 emu g −1 , respectively, when a eld of ±20 000 Oe was applied at The arrangement of the WO 3 /CuFe 2 O 4 catalyst in the system before and aer magnetic separation, aer the end of the reaction process, is presented in Fig. 8b. It is possible to observe that the magnetic properties of the developed catalyst are efficient to promote the process of separation and recovery of the catalyst from the reaction products (biodiesel and glycerol) when applied an external magnetic eld, leading the process to full separation in a few minutes.

Inuence of reaction parameters on biodiesel synthesis process
The biodiesel samples synthesized from different reactional conditions of reaction temperature, reaction time, MeOH : oil molar ratio and catalyst loading were evaluated for the ester content and kinematic viscosity, results presented in Fig. 9.
The effect of reaction temperature on the catalytic performance of the WO 3 /CuFe 2 O 4 catalyst during the trans-esterication reaction was studied in the range from 140 to 180°C (Fig. 9a). The highest ester conversion, 85.6%, was achieved for biodiesel synthesized at 180°C. Thus, it is possible to infer that the transesterication process using the magnetic acid catalyst is strongly inuenced by reaction temperature due to its endothermic nature. In general, reactions using acid catalysts require high temperatures due to low diffusion and reaction speed. 47 Thus, the use of high temperatures has advantages to the reaction system, such as: increase in molecular collisions, kinetic energies and degrees of miscibility between the reagents. These factors favor the activation of the carbonyl group from the waste cooking oil triacylglycerols, allowing the nucleophilic attack of methanol, resulting in the production of methyl esters by the transesterication route. 33 This inuence of the temperature variable on the trans-esterication reaction is also reported in the studies developed by Jiménez-López et al. and Xie and Yang, in which they obtained biodiesel with ester contents of 92.0% and 72.5%, when the reactions were conducted at reaction temperatures of 200°C and 180°C, respectively. 23,48 The kinematic viscosity of the biodiesel obtained suffered direct interference from the reaction temperature applied in the system. At a reaction temperature of 180°C, the biodiesel obtained presents kinematic viscosity of 4.7 mm 2 s −1 . On the other hand, when the reaction was carried out at 140°C, biodiesel presented a kinematic viscosity of 23.3 mm 2 s −1 . Thus, kinematic viscosity values tend to decrease with the increase of temperature. This behavior occurs due to the transfer of saturated chains to the biodiesel molecule as methyl esters are formed. Thus, there is a reduction of intermediates such as monoacylglycerol and diacylglycerol, compounds that may be responsible for increasing the viscosity of biodiesel. 49 The study of the inuence of time on the biodiesel synthesis process was carried out in the interval of 1 to 5 h. Based on the data presented on Fig. 9b, an increase in the value of the ester content for the biodiesel was observed from 24.4% to 85.6%, when the reactions are performed at reaction times from 1 to  3 h, respectively. However, the use of reaction times greater than 3 h did not presented signicant changes in ester levels of biodiesel esters obtained, since biodiesel synthesized in 4 and 5 h resulted in ester contents of 85.3% and 83.2%, respectively. This slight decrease occurs due to the reversible nature of the transesterication reaction aer reaching equilibrium. 50 In addition, this inuence is strongly observed in the kinematic viscosity of biodiesel synthesized in different reactional times, since there is a signicant decrease in viscosity values from 18.4 mm 2 s −1 to 4.7 mm 2 s −1 , when the reactions are processed at times from 1 and 3 h, respectively. Therefore, the time of 3 h was chosen as an optimal parameter for the transesterication reaction using the magnetic catalyst WO 3 /CuFe 2 O 4 . Fig. 9c shows the impact of the variation of the molar ratio MeOH : oil from 25 : 1 to 45 : 1 on the transesterication reaction using the magnetic catalyst WO 3 /CuFe 2 O 4 . It is possible to observe from the results obtained that the catalytic efficiency of the WO 3 /CuFe 2 O 4 catalyst increases as the MeOH : oil molar ratio is increased in the reaction, since the biodiesel synthesized using the MeOH : oil molar ratio of 25 : 1 presents an ester content of 51.2%, while the reaction performed in the MeOH : oil molar ratio of 45 : 1 leads to a biodiesel with an ester content of 95.2%, representing an increase of about 45% in the ester content. Thus, the MeOH : oil molar ratio of 45 : 1 was chosen as the most benecial relationship for the process. In general, acidic nature catalysts require higher molar ratios to achieve a higher conversion into biodiesel, since a greater amount of methanol in the reaction medium favors the phenomena of mass transfer in the system and facilitates the access and performance of the catalyst to the substrate through the high internal pressure inside the closed reactor. 51,52 The inuence of the MeOH : oil molar ratio on the kinematic viscosity of biodiesel revealed a decrease in the values from 10.8 mm 2 s −1 to 4.7 mm 2 s −1 when using the MeOH : oil molar ratios of 25 : 1 and 45 : 1, respectively. It is noteworthy that a biodiesel with a kinematic viscosity value below 6.0 mm 2 s −1 is desirable, as it facilitates the injection and dissolution of the fuel during its use. 53 Catalyst loading is considered a key reaction parameter for the biodiesel production process. In order to evaluate the effect of the concentration of the magnetic catalyst WO 3 /CuFe 2 O 4 , catalytic tests were performed under the catalyst loading range of 2-10%. Fig. 9d shows that the efficiency of biodiesel production showed a signicant improvement in the values of serum content from 50.3% to 95.2% when using catalyst loadings from 2 to 6%, respectively. This is due to the greater availability of active sites present in the reaction system, promoting greater contact of the oil-methanol-catalyst system. 54 The use of catalyst loadings greater than 6% in the transesterication reaction causes a decrease in the ester content of the biodiesel, given that the use of 8 and 10% of catalyst in the process resulted in biodiesel with ester contents of 91.7 and 78.3%, respectively. This negative impact is related to mass transfer problems in the system due to excess catalyst, since a greater amount of catalyst increases the viscosity of the reaction mixture during the transesterication reaction of frying oil. 53 The biodiesel obtained using the optimum catalyst loading condition of 6% showed kinematic viscosity of 4.7 mm 2 s −1 .
The results obtained in the study of inuence of reaction variables applying the magnetic catalyst WO 3 /CuFe 2 O 4 , evidence the optimal reaction condition of the process: reaction temperature of 180°C, reaction time of 3 h, molar ratio MeOH : oil of 45 : 1 and catalyst loading of 6%, which results in a biodiesel with a value of maximum ester content of 95.2%.

Physicochemical properties of biodiesel
Biodiesel obtained from waste cooking oil under optimal reaction conditions, using the magnetic acid catalyst WO 3 /CuFe 2 O 4 , was evaluated for its physicochemical properties and compared with the ASTM D6751 international standard. The results are given in Table 2. The kinematic viscosity and density are important fuel properties because the rst one shows the ability of a material to ow, and both are related to the quality of fuel atomization and biodiesel's molecular structure. 13 The biodiesel exhibited kinematic viscosity and density values of 4.7 mm 2 s −1 and 0.881 g cm −3 , respectively. Based on these results, the biodiesel obtained in this research showed values within limits established by ASTM standard range. The estimated acid value of the synthesized biodiesel was 0.21 mg KOH g −1 . The low acid value is within the limit dened by ASTM as well as it means that any corrosion will be caused in engine by biodiesel. 55 The ash point (FP) is another essential fuel property which is an indirect measure of fuel volatility. 56 The FP measure of biodiesel reached 155°C, indicating security for storage and portability. The cold lter plugging point (CFPP) is a parameter used to determine the minimum temperature at which fuel lters clog in automotive engines due to partial solidication of fuel. 13 The biodiesel showed CFPP of 0°C, inferring that the biodiesel could be used in cold weather countries. In the corrosiveness to copper analysis, the biodiesel presented a value of 1a, suggesting that the biofuel will not cause damage to the engine's metallic components. Similar value was obtained by Gonçalves et al. 20 All results conrm that the waste cooking oil has been successfully converted into biodiesel using WO 3 / CuFe 2 O 4 magnetic acid catalyst and conform to ASTM D6751 standard. The biodiesel oxidative stability value of 4.80 h is greater than the minimum limit of 3 h dened by ASTM D6751.

Assessment of magnetic catalyst stability WO 3 /CuFe 2 O 4
The reuse and recovery capacity are characteristics that make the heterogeneous catalyst more economically feasible for the biodiesel production process. 55 The magnetic catalyst WO 3 / CuFe 2 O 4 was evaluated by carrying out several reaction cycles under the optimal condition of transesterication reaction, as shown in Fig. 10. It is noteworthy that aer each reaction cycle, the catalyst was recovered by applying an external magnet, washed with heptane and ethanol to eliminate possible impurities from the catalyst surface from the reaction mixture, and dried in an oven for 12 h. Fig. 10a shows the results obtained in terms of ester content and kinematic viscosity for the synthesized biodiesel. The WO 3 /CuFe 2 O 4 magnetic catalyst was reused for ve reaction cycles and showed a reduction of its catalytic efficiency of approximately 40%. Two reasons may be related to the loss of catalytic activity: (1) partial leaching of the active sites and (2) deposition of organic matter on the catalyst surface. The rst hypothesis was veried by analyzing the EDS of the catalyst recovered aer the h reaction cycle (Fig. 11a), the analysis revealed a decrease in the tungsten concentration (W) present in the catalyst from 35.4% to 12.4%. In addition, there was a reduction in surface acidity value from 7.43 mmol H + g −1 (catalyst before reaction) to 3.38 mmol H + g −1 (catalyst aer the h reaction allotment cycle), conrming the leaching of tungsten species from the magnetic catalyst surface.   Surface Acidity of the CuFe 2 O 4 support (2.71 mmol H + g −1 ) and the catalyst WO 3 /CuFe 2 O 4 (7.43 mmol H + g −1 ). The intrinsic acidity of CuFe 2 O 4 was not sufficient to promote the trans-esterication process. The impregnation with tungsten increased the surface acidity value through the new Brønsted acid sites present in the catalyst, which enable the catalytic activity of the catalyst by its interaction with the carbonyl group in the transesterication process. 27 Catalyst recovery was 88.05 ± 2.92%, which proves the efficiency of the magnetic separation process employed. Furthermore, the VSM analysis of the catalyst before and aer the h reaction cycle is displayed in Fig. 11b. When comparing the magnetic characteristics of the materials, it is possible to observe an increase in the value of saturation magnetization (M s ) from 12.04 emu g −1 to 19.12 emu g −1 . This behavior possibly occurs due to leaching of the nonmagnetic component (WO 3 ) of the magnetic support surface (CuFe 2 O 4 ) during reaction cycles, which makes the M s value of the reused catalyst closer to the M s value of the copper ferrite (22.70 emu g −1 ).
In view of the previous statements, the strategy adopted to improve the catalytic efficiency of the WO 3 /CuFe 2 O 4 catalyst was to perform a heat treatment by calcination at 500°C/3 h of the catalyst aer each reaction cycle. The reuse study of the catalyst using heat treatment is presented in Fig. 10b and demonstrated superior reuse capacity when compared to the previous results, providing a biodiesel with an 80.6% ester content in the h reaction cycle. At rst, the calcination catalyst reactivation process is a way to eliminate organic components from the catalyst surface that were not removed in the washing stage, causing a reduction in the efficiency of catalytic solids due to the blocking of active sites, which hinders the access of reagents to these sites. 57 In addition, heat treatment played a key role in assigning greater stability and maintenance of active species in the magnetic support.
The study of the oxidative stability of the biodiesel produced during the transesterication reaction cycles using WO 3 / CuFe 2 O 4 was carried out to evaluate the impact of leached species in the oxidation process of the obtained esters (Table S1 †). The results of biodiesel's oxidative stability showed values of 4.80, 4.61, 4.17, 3.89 and 3.33 h for the rst to the h reaction cycle, respectively. The decrease in induction time values implies that leached metals can affect the oxidation process of the obtained samples. However, leaching does not affect the quality of biodiesel obtained through catalyst reuse cycles, because all results are in accordance with the minimum limit of 3 h stipulated by the method ASTM D6751.
The catalytic efficiency and stability of the WO 3 /CuFe 2 O 4 catalyst are compared with various acid catalysts under the optimal transesterication reaction conditions. The magnetic acid catalyst developed in this study allowed high activity and stability throughout the reactional cycles of transesterication, without signicant loss of its catalytic efficiency aer ve uses. From the analysis of the data contained in Table 3, it is veried that most acid catalysts require higher temperatures and reaction times in order to lead to biodiesel with high ester contents. On the other hand, some works developed exhibit more severe conditions of alcohol : oil molar ratio e catalyst loading. 20,49,60,61 However, the application of WO 3 /CuFe 2 O 4 catalyst in the transesterication reaction promoted the production of a biodiesel with an ester content greater than 95%, using milder reaction parameters than most catalysts presented in Table 3 or with similar trend. Therefore, the WO 3 /CuFe 2 O 4 magnetic acid catalyst can be considered an appropriate choice for the biodiesel production process via transesterication.

Conclusions
The present work evaluated the application of a new heterogeneous magnetic acid catalyst in the transesterication of the waste cooking oil in methyl biodiesel. The WO 3 /CuFe 2 O 4 catalyst was prepared by wet impregnation of the active phase on the CuFe 2 O 4 support, synthesized by the coprecipitation method. The analyses of XRD, FTIR, SEM, EDS, TG/DTG, VSM and Surface acidity indicated that the heterogeneous catalyst was successfully formed and conrmed its bifunctional character (catalytic and magnetic activity). The magnetic catalyst developed in this study obtained a biodiesel with an ester content of 95.2% under the optimal reaction condition (reaction