A recyclable solid catalyst of KF/Ca-Mg-Al-O using for biodiesel production from jatropha seed oil: preparation, characterization, and methanolysis process optimization

A recyclable heterogeneous KF/Ca-Mg-Al-O catalyst was prepared by co-precipitation and calcination processes. The characteristics of the catalyst were investigated using FTIR, XRD, TG-DTG and SEM. Response surface methodology was utilized to obtain the best most extreme biodiesel production yield. The ideal biodiesel production conditions were: the amount of catalyst was 3 wt%; the reaction temperature was 65 °C; the alcohol oil molar ratio 9.8:1, the reaction time was 3.5 h. Under these amended conditions, the average biodiesel yield was 95.19%, which is well in close concurrence with the worth anticipated by the model. The repeatability of catalysts was studied. After using the catalyst seven times, the catalytic efficiency was only reduced by 2.7%. These results indicate that the catalyst has good catalytic efficiency and is recyclable.


Introduction
At present, the available petrochemical resources worldwide are decreasing every day. With the development of economy and society, the demand for energy is increasing. Petrochemical resources are gradually decreasing, and the use of fossil energy will cause emissions of toxic pollutants and carcinogens (Arumugam and Ponnusami 2019). However, biomass energy has the advantages of easy production and little pollution. As an important biomass energy source, biodiesel has received widespread attention (Quah et al 2019, Gohain et al 2020. At present, the preparation of biodiesel primarily includes physical method and chemical method (Yaakob et al 2013). The physical method includes a direct mixing method and a microemulsification method (Ayodeji et al 2018). The direct mixing method is simple to prepare, but the quality of biodiesel is not high, and carbon deposition can easily occur during combustion (Marwaha et al 2019). The micro-emulsion method has the disadvantages of poor stability and unstable storage process. Chemical methods primarily include hightemperature thermal cracking and transesterification. Although the high-temperature cracking method has a simple preparation process, the reaction process requires a reactor and consumes more energy (Lani et al 2019, Pandit andFulekar 2019). Transesterification is a preparation method with good catalytic effects, mild preparation conditions, and high-quality biodiesel (Viriya-empikul et al 2010). Biodiesel is produced through transesterification with methanol or ethanol (Roschat et al 2016).

Biodiesel quantitative analysis
The content of fatty acid methyl ester was quantitatively analyzed by gas chromatography (Essamlali et al 2019). The chromatographic conditions were as follows: flame ionization detector was used; the column was AE. FFAP (30 m×0.25 mm×0.25 μm); the inlet temperature was 260°C; the injection volume was 1 μl, and the split ratio was 15:1. The temperature programming scheme was as follows: the initial temperature was maintained at 150°C for 3 min, increased to 200°C at 10°C min −1 , and then increased to 250°C for 5 min at 5°C min −1 .
FT-IR analysis was conducted with a scanning range of 4000-400 cm −1 , instrument resolution of 4 cm −1 , and scanning 16 times. The infrared test sample was prepared by KBr tableting, and the sample was approximately 2% of the KBr mass.
TG-DTG analysis: The thermal change of the catalyst and catalyst carrier was studied at 30°C, a termination temperature of 800°C, and a heating rate of 10°C min −1 .
SEM analysis: Measurement was carried out under an electron acceleration voltage of 20 kV at room temperature. The catalyst was characterized by SEM to determine the particle size and surface structure. 2.2.5. Biodiesel production The trans-esterification reaction was performed in three of them burn bottles (150 ml) , equipped to a reflux condenser and a magnetic stirrer. Then, the preprocessed jatropha oil was first blended in with methanol in variable molar ratios (8:1-14:1), and afterward stacked into the reaction vessel with the addition of proper amount of KF/Ca-Mg-Al-O catalyst (1-7 wt %). Next, the reaction temperature was changed (40°C-70°C), and the reaction time was differed from 2.5 to 5 h to guarantee the appropriate transformation of triglycerides to unsaturated fatty acid methyl esters (FAMEs). Then, the reaction mixture was cooled down and centrifugated. After centrifugation, crude biodiesel, glycerol and catalyst were used from top to bottom. The upper layer is dumped into the split funnel, adding purified water, fully shaken, static, layered, discharging the lower water layer, and repeating the above operation many times until the upper biodiesel is neutral. Anhydrous sodium sulfate was added to the washed biodiesel and centrifuged to obtain the biodiesel. RSM is an statistical method that utilizes quantitative information from fitting test plan to decide and all the while tackle multivariate conditions. Hence, RSM is less difficult and less tedious than other methods (Yin et al 2011).
Central Composite Design (CCD) of RSM was applied to study the impact factors for biodiesel preparation. Based on the preliminary range of extraction variables, a four-level-three-factor CCD was adopted in this study. The independent variables X 1 , X 2 , X 3 , X 4 represent the catalyst dosage (wt%), alcohol oil molar ratio (mol: mol), reaction temperature (°C), reaction time (h), table 1 is design and results of CCD, the examination was done in a standard order. The conduct of the system was clarified by the next second degree polynomial equation: y is the response function, β k0 is an intercept, β ki , β kii and β kij are the coefficients of the linear, quadratic and interactive terms, individually. And accordingly X i and X j address the coded free independent factors. The fitted polynomial condition is communicated as surface plots to imagine the connection between the response and trial levels of each factor and to obtain the ideal conditions. Design-expert (Version8.0.6.1) software package was utilized to examine the trial information. P-values of less than 0.05 were viewed as genuinely critical.

Results and discussion
3.1. Physicochemical properties of raw oils and contents Table 2 shows the acid value, saponification value, water content, iodine value, and molecular weight of Jatropha seed oil.
As shown in table 2, Jatropha seed oil has a molecular weight of 951.49 g mol −1 , water content of 0.11%, acid value of 7.68 mg KOH/g, and iodine value of 101.56 g/100 g. Jatropha seed oil has good fluidity.
Jatropha seed oil has a high acid value of 7.68 mg KOH/g. Considering that the catalyst used in the transesterification reaction is a solid-base catalyst, a high acid value is not conducive to the transesterification reaction. 20.0 g of Jatropha seed oil was collected and put into a separatory funnel. Then, 100 ml of methanol was added to the separatory funnel, and the mixture was shaken thoroughly. After five reciprocations, the acid value of Jatropha seed oil dropped to 0.84 mg KOH/g.
The fatty acid of Jatropha seed oil was methylated, and the fat methyl ester content of the selected Jatropha seed oil was analyzed by gas chromatography-mass spectrometry. Table 3 shows the fatty acid methyl ester.

Investigation of catalyst preparation conditions
3.2.1. Effect of the molar ratio of magnesium to aluminum on biodiesel yield As shown in figure 1(a), the yield of biodiesel increases with the increase of the magnesium-to-aluminum molar ratio. When the ratio of magnesium to aluminum is 3:1, the yield of biodiesel is the largest. This might because that the basicity of MgO is stronger than that of Al 2 O 3 , the alkalinity of the whole catalyst is increased with the increase of magnesium to aluminum ratio, thereby its catalytic activity increased (Lawan et al 2020). On the other hand, the electronegativity of aluminum is stronger than that of magnesium. When magnesium ratio increases, the relative number of aluminum decreases, and the average electronegativity of the catalyst also decreases, the number of basic active sites in the catalyst increased with the increase of magnesium to aluminum ratio, and the conversion rate of biodiesel increased (Cheng 2011, Lawan et al 2020. When the molar ratio of magnesium to aluminum is 4:1, the alkaline enhancement of the whole catalyst is not evident and the biodiesel yield changed slightly. Gao et al reported that when the ratio of divalent cation to trivalent cation for the catalyst was high, too much divalent cation could hindered their combination with the trivalent cation to form the layer structure, and too much divalent cations were added and there were not enough trivalent cations to act with them (Gao et al 2010). Therefore, the selected magnesium-to-aluminum molar ratio is 3:1.

Effect of calcium content on biodiesel yield
As shown in figure 1(b), biodiesel yield increased when calcium content was less than 2% and decreased when calcium content was higher than 2%. When the calcium content was 2%, the biodiesel yield was the highest. When the calcium content was higher than 2%, the biodiesel yield decreased with the increase of calcium content. It suggests that Ca is essential in the creation of the active sites for transesterification. The Ca impregnated on the surface of the Mg-Al HT created active sites for the reaction to occur (Dahdah et al).
Castro et al reported that Ca content could affects the amount of base sites/base site strength, and the physical structure of the catalysts (Castro et al 2014). On the other hand, when the Ca 2+ content increased, more Al 3+ combined with it and partly destroyed the layer structure, the Mg 2+ relative molar ratio decreased instantaneously, and then less KMgF3 was formed (Gao et al 2010).

Effect of KF load on biodiesel yield
As shown in figure 1(c), the activity of the catalyst increased with the increase of the KF load. When the KF load reached 25%, the biodiesel yield was the highest. When the KF load was further increased, the biodiesel yield decreased slowly probably because 25% KF has already occupied all the surface-active sites, and further increasing KF could not be combined with the active site. Moreover, the catalytic activity per unit was lowered during the preparation of biodiesel. Thus, KF load of 25% was selected.

Effect of calcination time on biodiesel yield
As shown in figure 2(a), the biodiesel yield increases with the increase of calcination time. When the calcination time was 4 h, the biodiesel yield was the highest, but the catalytic effect was almost unchanged when longer than 4 h probably because the catalytically active structure has formed on the surface of the catalyst, and there is little change with the extension of time. Therefore, 4 h was selected as the calcination time.

Effect of calcination temperature on biodiesel yield
As shown in figure 2(b), the activity of the catalyst increases as the calcination temperature increases. When the calcination temperature reaches 700°C, the biodiesel yield was the highest, which may be due to the change of the pore structure of the catalyst surface and the occurrence of the calcination temperature. Upon chemical reaction, a new composite metal oxide is formed on the surface. However, the catalytic effect continuously increases, which may be due to the fact that the temperature is too high to destroy the active structure of the catalyst; thus, the catalytic effect is lowered. Therefore, 700°C was selected as the calcination temperature of the catalyst.

Catalyst characterization 3.3.1. Basic strength characterization
The catalyst prepared under optimized conditions was subjected to color reaction with bromo thymol blue (pK a = 7.2), phenolphthalein (pK a = 9.8), and 2,4-dinitroaniline (pK a = 15.0). It can make bromo thymol blue and phenolphthalein alkaline color, but it cannot make a 2,4-dinitroaniline alkaline color. Therefore, the basic strength of the catalyst was 9.8-15.0.

XRD analysis
As shown in figure 3, the difference among the diffraction peaks of Mg-Al-O, CaO, KF, and KF/Ca-Mg-Al-O solid-base catalysts was at diffraction angles of 10°to 80°. After modification with Ca, the corresponding diffraction peak of Mg-Al-O at 65°and the broad peak corresponding to 37°disappear, and a new diffraction  peak is formed at approximately 55°, which proves that CaO is not simply dispersed on the surface of the KF/Ca-Mg-Al-O catalyst, and forming a new composite metal oxide Ca-Mg-Al-O as a catalyst carrier by calcination at a high temperature and Mg-Al-O is possible. Compared with KF and KF/Ca-Mg-Al-O catalysts, the characteristic diffraction peaks of KF at 34°and 46°disappeared, which proved that KF was not loaded onto the surface of the catalyst carrier. This may be that after calcination at high temperature, new compounds such as KCaF 3 were formed. The following experimental results of FI-IR verified this conclusion.

FT-IR analysis
As shown in figure 4(b), compared with the infrared spectrum of uncalcined and calcined Mg-Al-O, the -OH peak of 3500 cm −1 was significantly weakened after calcination and appeared at 1700 and 1400 cm −1 before calcination. The sharp peak, stretching vibration peak of Mg-O at 1700 cm −1 , and the stretching vibration peak of Al-O at 1400 cm −1 almost disappeared after calcination. In addition, the vibration peak of Al-O-Mg was formed at 694 and 578 cm −1 ; thus, a composite metal oxide Mg-Al-O was formed. As shown in figure 4(c), the infrared pattern of Ca-Mg-Al-O before and after calcination was compared. Considering that Ca(OH) 2 was used for modification, the bound water was reduced after calcination, resulting in a significant decrease in the -OH stretching vibration peak of the bound water (3500 cm −1 ). The peak appearing at 3650 cm −1 after calcination was the stretching vibration peak of Ca-O. Comparing Ca-Mg-Al-O (uncalcined) with Mg-Al-O (calcined) in figure 4(b), significant Ca-O-Mg and Ca-O-Al stretching vibration peaks were observed at 1640 and 1460 cm −1 . After calcination, the stretching vibration peak of Ca-O-Mg at 1650 cm −1 disappeared, and a single absorption peak was formed at 1490 cm −1 , which was presumed to be a metal oxide formed by the combination of three metals, namely, Ca, Mg, and Al.
Comparing the infrared spectra before and after calcination of KF/Ca-Mg-Al-O, two new absorption peaks are formed at 1480 and 1380 cm −1 (figure 4(d)). Therefore, a stretching vibration peak is estimated to be Ca-F-K, and KCaF 3 may be formed.
In the infrared spectra of KF and CaO, the peak between 3750 and 3600 cm −1 is a stretching vibration peak belonging to Ca-O, and the infrared spectrum after calcination from KF and KF/Ca-Mg-Al-O shows the disappearance of the characteristic peak of KF from 1750 cm −1 to 1650 cm −1 and the weakening of the characteristic peak from 1850 cm −1 to 1800 cm −1 . The newly emerging peaks at 3500 and 750 cm −1 prove that KF is not only loaded onto the surface of the body, but also formed by calcination with the carrier at a high temperature.
By comparing the infrared spectra of Mg-Al-O (calcined) and Mg-Al-Ca-O (calcined), a new compound (Ca-Mg-Al) may be formed after modification of Ca at 1450 cm −1 . This finding is also demonstrated by the XRD demonstrates that KF binds to the carrier after calcination to form a new compound.

TG-DTG analysis
As shown in figure 5(a), Mg-Al-O has four distinct thermogravimetric stages at 85°C-105°C, 135°C-150°C, 220°C-260°C, and 450°C-500°C. 85°C-105°C was the weight loss caused by the removal of adsorbed water and surface-adsorbed carbon dioxide, and the weight loss rate was approximately 4%. 135°C-150°C was the weight loss caused by the removal of adsorbed water in the internal void structure of Mg-Al-O, and the weight loss rate was about 2.3%. The weight loss peak of 220°C-260°C corresponded to the removal of bound water between internal voids, and the weight loss rate was approximately 7.1%. The weight loss peak at 450°C-500°C was due to dehydroxylation or anion between the magnesium-aluminum hydrotalcite layers, and the weight loss rate was approximately 18.3%. When the temperature was higher than 500°C, the mass change of Mg-Al-O was small.
As shown in figure 5(a), the calcination temperature of Mg-Al-O should not be lower than 500°C, but too high temperature might destroy the surface structure of Mg-Al-O. Moreover, the voids might collapse, and Mg-Al-O might be agglomerated. Therefore, the selected calcination temperature was 600°C.
As shown in figure 5(b), Ca-Mg-Al-O has three distinct stages of thermal weight loss at 85°C-125°C, 320°C -400°C, and 630°C-680°C. 85°C-125°C was the weight loss caused by the removal of adsorbed water and surface-adsorbed carbon dioxide on the surface, and the weight loss rate was approximately 2.5%. The weight loss peak at 320°C-400°C was due to the removal of bound water in the internal void of Ca-Mg-Al-O and the removal of hydroxide in the modified Ca(OH) 2 , and the weight loss rate was approximately 3%. The weight loss peak of 630°C-680°C may be the weight loss caused by CaO and Mg-Al-O forming a new composite metal oxide of Ca-Mg-Al-O, and the weight loss rate was approximately 1.8%. Therefore, Ca-Mg-Al-O may be formed similar to that obtained in figure 4. The mass of the carrier Ca-Mg-Al-O slightly increases when continuously increasing the temperature.
As shown in figure 5(b), the calcination temperature was lower than 680°C to form a catalyst carrier. If the calcination temperature is too high, then the surface structure of Ca-Mg-Al-O may be destroyed; the voids may collapse, and Ca-Mg-Al-O may be agglomerated, which will reduce the loading sites of KF and decrease the catalytic efficiency of the final catalyst. Therefore, the selected optimum calcination temperature was 700°C.
As shown in figure 5(c), KF/Ca-Mg-Al-O has four distinct thermogravimetric stages at 65°C-115°C, 300°C -350°C, 360°C-420°C, and 600°C-670°C. 65°C-115°C is the weight loss caused by the removal of adsorbed water and surface-adsorbed carbon dioxide on the surface, and the weight loss rate is approximately 6%. The weight loss peak at 300°C-350°C is the removal of some of the hydroxyl groups hydrated by Mg-Al-O during impregnation, and the weight loss rate is approximately 4.2%. The weight loss peak of 360°C-430°C may be the removal of the hydroxyl group of Ca(OH) 2 formed by Ca-O during impregnation, and the weight loss rate is approximately 1.6%. The weight loss peak at 600°C-670°C, the possible loading of KF, and the new compound formed by calcination of the catalyst carrier were considered as the active components of the catalyst. This hypothesis is consistent with the conclusions obtained in figure 3. As shown in figure 5(c), when forming a KF/Ca-Mg-Al-O catalyst, the calcination temperature should not be lower than 670°C. If the calcination temperature is too high, then the surface structure of the catalyst may be destroyed; the voids may collapse, and the catalyst may agglomerate. The active site leading to the catalytic reaction is reduced, and the catalytic effect of the final catalyst is lowered. Therefore, the selected optimum calcination temperature was 700°C.

SEM analysis
As shown in figure 6(a), the surface of the calcined carrier Ca-Mg-Al-O at 700°C has many fine granular structures; the entire surface is relatively flat, and the overall structure is relatively regular.
As shown in figure 6(b), the surface of the catalyst KF/Ca-Mg-Al-O after loading KF has a large particle structure, and the lamellar structure appears on the surface of the particle probably because of the loading of KF. This finding can also explain the decrease in the specific surface area of the catalyst after loading. Moreover, a pore-like structure appears on the solid surface, which allows the catalyst to combine with the reactants and provide a good catalytic effect. Furthermore, an interaction is found between the loaded KF and carrier, which is not simply carried on the surface.
3.4. Effective parameters of biodiesel production 3.4.1. Effect of KF/Ca-Mg-Al-O catalyst dosage on FAMEs yield As shown in figure 7(a). When the catalyst dosage is less than 3 wt%, the yield of FAMEs increased, at the catalyst dosage of 3 wt%, the biodiesel yield was the highest, continue to increase the catalyst dosage, FAMEs yield did not change significantly. The reason is that when the amount of catalyst is less than 3%, increasing the catalyst amount will increase the alkalinity in the reaction system, the alkali activity center of the reaction increases with the increase of the catalyst (Lawan et al 2020), plays a role in promoting the reaction, and increases the yield of FAMEs. When the catalyst continues to increase, the whole reaction system has been saturated, and thus the FAMEs yield does not change significantly. So the optimal amount of the catalyst selected is 3 wt%.

Effect of methanol/oil molar ratio on FAMEs yield
As shown in figure 7(b), at the alcohol oil molar ratio below 10:1, the FAMEs yield ratio increases with the alcohol oil mole ratio, the peak of the FAMEs yield appears at 10:1, and at the alcohol oil molar ratio above 10:1, the FAMEs yield decreases with the alcohol oil molar ratio above 10:1. The reason for the rising stage is that increasing the reactant methanol can promote the positive movement of the ester exchange reaction balance and promote the production of FAMEs (Dahdah et al 2020). Therefore, the FAMEs yield increases with the alcohol oil molar ratio. However, when it continues to increase the methanol consumption, on the one hand, the concentration of the dilution catalyst in the system will be not conducive to the movement of the reaction to be positive, and on the other hand, it will reduce the concentration of the pretreatment mad seed oil in the reaction system. Moreover, excess methanol is dissolved in biodiesel, increasing the cost of separation and recovery, and, therefore, 10:1 was chosen as the best alcohol oil molar ratio.

Effect of reaction temperature on FAMEs yield
As shown in figure 7(c), the yield of FAMEs increases with the temperature at the reaction temperature below 65°C , the highest biodiesel yield at 65°C, and at the reaction temperature above 65°C, the FAMEs yield decreases with the reaction temperature. The reason is that as the temperature increases, the molecular movement in the reaction system accelerates, and the biodiesel yield increases with the reaction temperature. On the one hand, continuing to increase the temperature, the rapid reduction of methanol in the liquid phase leads to an increase in the volatilization amount of methanol during the reaction process, which is not conducive to the positive movement of the ester exchange reaction balance, aggravates the reverse reaction degree, and reduces the yield rate of FAMEs; on the other hand, the transesterification is chemical reaction controlled in the low temperature range as the energy provided for the reactant molecules to generate transesterification is inadequate. At the higher temperature, the transesterification is more easily generated and the higher molecular activity impels the equilibrium towards the products. At the same time, the mass transfer resistances between the heterogeneous catalyst and the liquid reactants become weak due to the reduced viscosity (Niu et al 2018, Niu et al 2020. So determining the temperature of 65°C is the optimal reaction temperature.

Effect of reaction time on FAMEs yield
The transesterification prompted for the methyl esters production with a rapid rate, where the biodiesel yield is markedly, increased through prolonging the transesterification duration (Ganesan et al 2019). Figure 7(d) shows that with longer than 3.5 h, the FAMEs yield increased at less than 3.5 h, with the highest biodiesel production at 3.5 h, and the FAMEs yield did not change significantly. The reason is that when this reaction is less than 3.5 h, the reaction is still moving forward. No equilibrium, and when reached 3.5 h, the reaction of preparing biodiesel has reached a dynamic equilibrium, continues to extend the reaction time, and will not have much impact on the reaction system. Therefore, 3.5 h was chosen as the best value of the reaction time.
3.4.5. Optimization of biodiesel production by RSM 3.4.5.1. Statistical analysis and the model fitting Response surface optimization is more favorable than the traditional single parameter optimization in that it saves time, space and unrefined material (You et al 2014). There were a sum of 29 runs for enhancing the four individual parameters in the CCD, the experimental conditions and biodiesel yield according to the factorial design was displayed in table 1. The experimental results showed that the biodiesel yield was 90.6% to 95.18%; the results were predicted using Design software and the predicted best conditions for biodiesel production is catalyst dosage of 3.01 wt%, alcohol oil molar ratio of 9.79:1, reaction temperature of 65.84°C, reaction time of 3.52 h, at this time, the FAMEs yield was 95.21%. The outcomes were fitted with a second order polynomial equation. The values of equation regression coefficients were determined, the response variable and the test factors are connected by the following second-order polynomial equation (3) The statistical significance of the polynomial regression equation was checked by F-test and p-value, and the analysis of variance (ANOVA) for the response surface quadratic equation model was displayed in table 4. The determination coefficient (R 2 =0.9874), displayed by ANOVA of the quadratic regression equation model, showing that the regression equation model was effective for prediction for expectation inside the scope of test factors. The linear coefficients (X 2 , X 3 , X 4 ), a quadratic term coefficient (X 12 , X 22 , X 32 , X 42 ) and the interaction coefficient (X 1 X 2 , X 2 X 3 , X 3 X 4 ) were seen as exceptionally critical (p < 0.04).

Optimization of the variable parameters
The RSM numerical approach was used to predict the optimum conditions in biodiesel production. Response surfaces were plotted by utilizing Design expert (version 8.0.6.1) programming to concentrate on the impacts of factors and their interactions on FAMEs yield. The full model filled equation (3) was made three dimensional graph to anticipate the connections between the dependent variables and the independent variables (Yin et al 2011).  The results of FAMEs yield impacted by catalyst dosage, alcohol oil molar ratio, reaction temperature, reaction time. These kinds of plots show impacts of two variables on the response at a time and the other variables was kept at level zero. As can be seen in figure 8(a), the FAMEs yield is greatly affected by the catalyst dosage and alcohol oil molar ratio, and the overall trend is rising first and then decreasing, indicating that the best value makes the highest FAME yield. Moreover, it can be seen that the catalyst dosage and alcohol oil molar ratio strongly interact from figure 8(a). Figures 8(b) and (c) is similar to the surface situation reflected in figure 8(a), so that the catalyst dosage strongly interacts with both the reaction temperature and the reaction time, and both have the optimal value of the FAME yield. In figure 8(d), the overall smoothness of the surface from the alcohol molar ratio shows that the alcohol molar ratio has less impact on FAMEs yield, while the FAMEs yield changes rapidly in the reaction temperature direction and has greater impact on FAME yield, with the best value interaction value. Figures 8(e) and (f) reflect a similar situation to figure 8(c), with large surface changes in the direction of the reaction time, indicating that the reflected reaction time factors have a strong impact on the experimental results.
In a word, the above four experimental factors have an interaction on FAEMs yield, but the relationship between the factors from the contour chart is: the weakest catalyst dosage, reaction temperature, reaction time, and the impact of alcohol oil mole ratio on FAMEs yield.

Verification of predictive model
The conditions for optimum FAMEs yield was obtained from the regression model equation, The suitability of the model equation for predicting the optimum response values was tested by using series of five experiments. The predicted best conditions for biodiesel production is catalyst dosage of 5.5 wt%, alcohol oil molar ratio of 8.84:1, reaction temperature of 54.27°C, reaction time of 3.89 h, at this time, the FAMEs yield was 95.34%.
Considering practical consideration, the conditions for biodiesel preparation of catalyst dosage were selected as 3 wt%, alcohol oil molar ratio 9.8:1, reaction temperature 66°C, reaction time is 3.5 h, three experiments, the average yield of FAME of 95.19% is close to the predicted theoretical value of 95.21%, indicating that the model can accurately describe the actual situation. It is shown that the model can be applied to a magnetic concave earth catalyst for biodiesel preparation.

Application and repeatability of catalysts
The biodiesel prepared by transesterification using KF/Ca-Mg-Al-O as catalyst was studied. The molar ratio of alcohol to oil was 10:1; the amount of catalyst was 3%; the reaction temperature was 65°C, and the reaction time was 3.5 h. The rate reached 94.3%.