Statistical Optimization of Biodiesel Synthesis from Waste Cooking Oil Using Bentonite Clay as Catalyst

Biodiesel was synthesized from waste cooking oil (WCO) utilizing a transesterication process with clay as a heterogeneous catalyst. The transesterication was carried out for 4 hours at reaction temperatures of 60°C using 4 wt.% catalyst and 9:1 (methanol to oil) ratio. Continuous stirring at 60°C for 4 hours and a 9:1 methanol/oil molar ratio with 4wt.% catalyst resulted in a high yield of FAMEs (91.2%). The impact of several reaction factors on biodiesel yield were explored. For the optimization of ve process parameters, this study uses regression models constructed using central composite design (CCD) of randomized response surface methodology (RSM) to forecast FAME production for the transesterication of waste cooking oil (WCO). The inuence of methanol to oil molar ratio, catalyst weight, temperature, and reaction time was investigated in 32 trials. The estimated coecient of determination (R), corrected R, and coecient of variance were determined to be 0.9956 percent, 0.9877 percent, and 663.15 percent, respectively, at a molar ratio of 9:1, catalyst concentration of 4 wt.%, and a reaction duration of 4 hrs. 3D plots were used to determine the effects of the combination of these elements. The response surface methodology (RSM) was shown to be an effective statistical technique for developing an acceptable empirical model for linking operational parameters and forecasting optimal operating conditions. The statistical analyses and the closeness of the experimental results to model predictions demonstrate the regression model's dependability, and the ndings will aid in the selection of an ecient and cost-effective biodiesel production technique from low-cost raw materials with high free fatty acid. Furthermore, according to international biodiesel specications, the fuel characteristics of the created biodiesel were within acceptable levels.


Introduction
Because of the political and economic instability of the oil market, as well as the environmental bene ts associated with reducing gaseous emissions from the combustion of non-renewable fuels, the production of sustainable alternative fuels has piqued academic and industrial interest in recent years 1,2 .Due to the expected scarcity of fossil fuels, an increase in the price of petroleum, and a large molecular comparison between biodiesel and petroleumbased diesel, biodiesel is being considered as a possible replacement and future fuel for diesel engines. This substitute fuel has a chance of realizing the technological essentials of diesel fuel 3 . It's biodegradable, renewable, safe, and environmentally friendly, with a high cetane number, greater lubrication, and a high ash point, and it has all of the physical and chemical properties of normal petroleum diesel 4 .
Biodiesel is described as monoalkyl esters of long-chain fatty acid generated from renewable fatty raw materials such as vegetable oils or animal fats, according to the American Society for Testing and Materials (ASTM). The term 'bio' refers to its recyclability and biological origin, while 'diesel' refers to its resemblance to diesel fuel and its use in diesel engines. This fuel has properties similar to fossil diesel, but it is far superior to the latter 5,6 .
Biodiesel is a renewable energy source that can be used to replace fossil-based diesel and reduce emissions 7 . Diesel is made from crude petroleum oil, which contains a mixture of pure hydrocarbon molecules (no oxygen molecules) ranging in size from 8 to 21 carbon atoms. Long-chain hydrocarbons with an ester functional group (-COOR) make up biodiesel, on the other hand. Monoalkyl esters of long-chain fatty acids derived from various feedstocks, such as plant oils, animal fats, or other lipids, are also known as triacylglycerides (TAGs), or simply triglycerides 8 .
Transesteri cation, also known as alcoholysis, is a process that produces biodiesel and is aided by acids, bases, enzymes, and other types and forms of catalysts 9 . Transesteri cation for biodiesel synthesis uses a variety of catalysts, including base, acid, and lipase, although the base-catalyzed reaction is the most popular in the industry because it is easier, faster, and less expensive to process 10 .
A heterogeneous catalyst, on average, has a greater conversion e ciency than a homogeneous catalyst, but more research is needed to establish the best time and temperature for each reaction. Calcium oxide is frequently employed as a heterogeneous catalyst due to the abundance of natural calcium sources and low production costs as well as its high basicity 11,12,13,14 . Calcination can convert natural calcium carbonate from eggshells, shells, limestone, dolomite, cuttlebone, and other sources to CaO.
A study on oil transesteri cation using CaO as a heterogeneous catalyst was conducted, and 95% conversion was achieved with a methanol to oil molar ratio of 12:01, 8% CaO in relation to oil mass, and a reaction time of 3 h 15 . A similar study was carried out, and 93% conversion was achieved with 80 minutes of reaction time and 5 percent catalyst in relation to oil (m/m) using CaO as the catalyst at a reaction temperature of 65 C. The authors employed a 6:1 molar ratio of methanol to oil 16 . Due to the larger production ratio in comparison to traditional processes based on bench reactors, all of the studies by the above-cited authors were undertaken in a xed-bed reactor on a laboratory scale for biodiesel production in continuum processes 17,18,19 .
Vegetable oil, animal fats, and microalgal oil are the current feedstocks for biodiesel manufacturing. Vegetable oil is being used as a sustainable commercial feedstock right in the middle of it all. There are a variety of edible oils available on the market that can be used to make biodiesel. However, edible oil is typically used for cooking, and when it is utilized for biodiesel production, it may result in a shortage of oil for cooking, raising the cost of oil 20 .
Recent biodiesel research has focused on nding ways to lower the high cost of biodiesel, particularly for techniques that focus on lowering the cost of raw materials. Because waste cooking oil is roughly half the price of virgin oil, using it instead of virgin oil to make biodiesel is a cost-effective solution to save money on raw materials 21 .
Furthermore, repurposing excess cooking oil could aid in the solution of the waste oil disposal problem 22 . The result of repeated frying, waste cooking oil (WCO) is generated every day from a variety of sources, including residences, restaurants, catering establishments, and industrial kitchens. If this garbage is disposed of in the environment, such as in an aquatic area, the water channel will degrade the area's quality. As a result, a different application is required.
Because it includes triglycerides, one possible application for waste cooking oil is to convert it into biodiesel.
Consumers dispose of waste cooking oil in sinks, garbage bins, drainage systems, toilets, or directly to nearby water bodies and lands, which is supposed to be processed and managed in a manner that is not harmful to human health and/or the environment. Several researchers have used WCO as a feedstock for biodiesel synthesis using a homogeneous catalyst, but only a few have looked into the combination of CaO catalyst and WCO in biodiesel production 23 .
The goal of an optimization study for biodiesel production is to help researchers design the most e cient and costeffective system in the biodiesel industry 24 . Several scholars have investigated biodiesel production optimization utilizing Response Surface Methodology (RSM) 25 . Response surface methodology is a set of mathematical and statistical methodologies that is widely used in the design of experiments, the construction of models, the determination of optimum conditions, and the evaluation of the relative importance of various elements impacting a process 26 . The goal of the research presented in this study was to nd the best production conditions for base catalyzed methyl ester transesteri cation from waste cooking oil. The goal of the study was to see how the parameters of methanol to oil molar ratio, catalyst weight, agitation speed, temperature, and reaction duration interacted during the transesteri cation of waste cooking oil. The tests were carried out using a central composite design (CCD) and response surface methodology (RSM) to assess the relationship between the parameters and nd the best conditions for producing methyl ester from waste cooking oil.

Materials And Methods
Materials. Sodium hydroxide (NaOH), methanol of analytical grade was purchased from Sigma Aldrich and used as received without further puri cation. Bentonite clay was purchased in a shop at Wuse in Abuja, Waste cooking oil, Deionized water obtained at Zaria was used throughout the research. Samples of the as-produced waste cooking oil (WCO) was collected from small and owner-operated restaurants from Onitsha, Anambra State, Nigeria.

Treatment of Waste cooking oil
The sample was poured into a beaker and heated at 110•C in an electric heater for 15 min to remove moisture. The WCO sample was allowed to cool to room temperature and subjected to vacuum ltration process to remove any food residue and other suspended solid matter in the sample. The clean WCO samples were stored in an airtight glass container The physicochemical properties of WCO such as like density, ash content, acid value, free fatty acid (FFA) value, saponi cation value, kinematic viscosity, and molecular weight were estimated as shown in Table 3 using standard procedures; American Oil Chemists Society (AOCS) while the remaining sample was kept in glass bottles to prevent contamination. The process of 38 was employed in the treatment of waste cooking oil for this research.

Preparation of catalyst
The bentonite clay was modi ed by using the impregnation method. The impregnation of bentonite clay with NaOH solution was conducted at room temperature for 24 hours under continuous stirring with ratio between bentonite clay to sodium hydroxide (NaOH) solution 1:10. After completing the impregnation process, the slurry was dried in the oven at 110 O C for 4hours to remove water. The dried slurry was then calcined in a mu e furnace at 400°C for 3 hours. The nal calcination was conducted to remove any volatile substances as a puri cation process. The catalyst was analyzed by scanning electron mocroscopy (SEM) and Fourier Transform-Infra-red (FTIR).

Transesteri cation Reaction
A batch reactor was used to carry out the transesteri cation reaction. A controlled water bath heater was used to heat the reaction. In a known volume of methanol, a known amount of catalyst (1-5 g) was added. In a controlled temperature water bath, the mixture was then heated to the desired temperature (60 °C). After that, waste cooking oil was added to the mixture while it was being vigorously stirred (350 rpm). In this investigation, the molar ratio of methanol to oil was 9:1, and the transesteri cation reaction lasted 1-5 hours. The solution was centrifuged at the end of the reaction time. The glycerin and biodiesel layers were separated after centrifugation. To remove water and methanol, the biodiesel phase was washed with water, decanted, and heated at 100 °C. By comparing the weight of layer biodiesel to the weight of waste cooking oil utilized, the percentage of biodiesel production was calculated.

Modeling and optimization by RSM
The central composite design (CCD) of the response surface technique version of design of experiment (DOE) available on the Design-Expert software was used to optimize biodiesel yield and catalyst recovery resulting from the transesteri cation of Waste cooking oil (WCO) into FAME. With full or fractional factorial points, axial points, and centre points that can be reproduced for every combination of categorical factor level, the CCD of a randomized RSM is a very popular and successful optimization technique. Several experiments are carried out in order to determine the most important elements for the greatest results when using DOE. Later, RSM was hired to create a mathematical model that would con rm the results of the studies. Catalyst concentration, reaction time, reaction temperature, methanol-to-oil molar ratio, and agitation speed were the process parameters evaluated for the optimization of the waste cooking oil (WCO) transesteri cation process, as shown in Table 1 To anticipate the reaction as a function of independent factors and their interactions, a quadratic polynomial equation was built using a central composite design 27 . The projected reaction was calculated using a mathematical model based on a second-order polynomial with interaction terms. The following is the response for quadratic polynomials 28 .
Where Y is % methyl ester yield, xi and xj are the independent study factors (coded variables), and 0, , and are constant co-e cient, regression co-e cient of the linear terms, regression co-e cient of the quadratic terms, and regression co-e cient of the interaction terms, respectively, and k is the number of factors studied and optimized in the experiment (number of independent variables). The Design-Expert 11.0.4.0 software package was used for regression analysis and analysis of variance (ANOVA).

Results And Discussion
Waste cooking oil characterization Due to several chemical events such as hydrolysis and material transfer between oil and food occur during the frying process, the chemical and physical features of the oil are altered. The physic-chemical variables of the waste cooking oil sample collected are listed in Table 3.
Characterization of the catalyst Scanning electron microscope. Fig.1a shows SEM pictures of the catalyst. The particles have irregular forms with voids, as shown in Fig.1a, as a result of the activating ingredient evaporating (NaOH). Fig.1b shows the spongy form of the catalyst. Fig.1b also shows the existence of rod-like particles and has a smoother surface than g.1a, which contains a few cavities and hair line breaks. This could be due to the activation process, which may have lled in the holes and created a smooth surface. Because triglyceride molecules react with methanol molecules absorbed on these spongy surfaces, the spongy structure of the catalyst will boost biodiesel generation.
Fourier transform infra-red analysis. The FTIR spectra of the catalyst was carried out in the range from 1000 -3500 cm-1 to study the catalytic transesteri cation effect. The FTIR spectra is shown in Fig 2. The changes in the functional groups provide the indication of the modi cations that occurred during the impregnation process. Based on the area of each peak, major absorption peaks were observed at 3893.8, 3652.8, 1114.5, 998.9, 9096.1 ,779.0 and 685.8 cm-1 . These major absorption bands and their corresponding functional groups are listed in Table 2.
Response surface method statistical analysis Table 4 shows the results of the CCD experimental design for transesteri cation of waste cooking oil (WCO) to FAME using RSM. Tocreate an acceptable and usable regression model, the actual yield was analyzed. A suitable model was chosen from among mean, linear, quadratic, cubic, quartic, and other options. The software created a cubic regression model and used it to forecast ideal parameters for the transesteri cation of WCO to biodiesel.
Equation (3) shows the best t model for FAME yield.
The trial/experimental changes are: catalyst concentration, methanol to oil molar ratio, reaction temperature, reaction time, and agitation speed, and A, B, C, D, and E show the estimates of the trial/experimental changes: catalyst concentration, methanol to oil molar ratio, reaction temperature, reaction time, and agitation speed. The positive sign in front of the words indicates that the component has a synergistic effect in increasing FAME yield, whereas the negative sign indicates that the factor has an antagonistic effect 29,30 . The positive coe cients in the model regression (equation 3) showed a linear rise in FAME yield. The quadractic word, on the other hand, had detrimental consequences on the FAME yield. Physical property of biodiesel produced 94.10% biodiesel (FAME) was obtained under ideal conditions. Based on ASTM standards, the biodiesel generated in this situation was further examined to establish its viscosity, density, pour point, and cloud point. The physical properties of the biodiesel produced are shown in Table 7. Despite the fact that the percent biodiesel (FAME) yield attained in this study was less than 96.50%, the product's viscosity, density, acid value, pour point, and ash point all met ASTM D6751 and EN 14214 standards.

As indicated in
Effects of process parameter on biodiesel yield.
Effect of methanol/oil molar ratio on biodiesel yield. Fig. 3a shows the experimental results, which show that the molar ratio of methanol to oil has a substantial impact on biodiesel yield. With the molar ratio, the biodiesel yield was enhanced. At a 9:1 molar ratio of methanol to oil, a 76.4% yield was reported. The surplus methanol is required since it can speed up the methanolysis process. The high concentration of methanol increased the production of methoxyl species on the catalyst surface, causing an equilibrium shift to the forward direction and consequently an increase in biodiesel conversion rate 31 . Increasing the methanol to oil molar ratio after the optimal 9:1 methanol to oil molar ratio would diminish the biodiesel yield. This is due to the presence of too much methanol above the optimal concentration, which stymies the reaction. The glycerol produced as a by-product of the reaction would primarily dissolve in the excess methanol and therefore block the reaction of methanol to reactants and catalyst, interfering with glycerine separation, lowering conversion by shifting the equilibrium in the other way.
Effect of catalyst concentration on biodiesel yield. The concentration of the catalyst is critical for improving the yield of the transesteri cation reaction. Fig.3b shows that as the catalyst concentration is increased from 1% w/w to 5% w/w, the biodiesel yield increases, but the yield decreases marginally as the catalyst concentration is increased further. With a biodiesel production of 86.0%, the best catalyst concentration was determined to be 4% w/w clay catalyst. Due to soap generation in the presence of a large amount of catalyst, the biodiesel yield has been marginally reduced. Furthermore, the presence of too much catalyst raises the viscosity of the reactants, decreasing the biodiesel production. The basic sites generated on the surface of the catalyst, as well as the soluble substance leached away from the catalyst, catalyze the transesteri cation 32 .
Effect of temperature on biodiesel yield. Fig.3c shows the biodiesel output from waste cooking oil transesteri cation at reaction temperatures ranging from 45 to 75°C. The biodiesel yield rises with the reaction temperature until it reaches an ideal point of 60°C, with an 89.0 percent biodiesel yield. The transesteri cation required some thermal energy at rst because the reaction was endothermic. Because the reaction mixture is a three-phase system (oil, methanol, and catalyst), enough thermal energy was required to overcome the diffusion barrier between the phases 33 . The high temperatures, on the other hand, are not ideal. When the temperature rises to the boiling point of methanol, the methanol quickly vaporizes and forms a signi cant number of bubbles, inhibiting the process at the two-phase interface and lowering the biodiesel yield.
Effect of time on biodiesel yield. Fig.3d illustrates the biodiesel production for waste cooking oil transesteri cation at various reaction times ranging from 0.5 to 5 hours. Biodiesel generation was rapid in the early phases of the transesteri cation reaction until the reaction achieved equilibrium. The reaction begins to reverse in the direction of reactants once it has passed the optimum point. The reversibility of the transesteri cation reaction caused this result 34 . As a result of the catalyst's ability to absorb the product, a long reaction time limits biodiesel yield. As a result, determining the optimal transesteri cation reaction time is critical. The best reaction time in this example was 4 hours, with a yield of 89.0%.
Effect of agitation speed on biodiesel yield. During the transesteri cation reaction of triglycerides, the agitation speed is an important reaction variable that impacts the biodiesel production. Fig.3e shows that as the agitation speed was raised, the biodiesel yield rose, reaching a maximum of 91.0 percent at 350 rpm. However, there was no substantial improvement in biodiesel yield beyond this optimum agitation speed. The current study used a 350 rpm agitation speed to achieve the highest biodiesel output. Furthermore, this demonstrated that a 350rpm agitation speed was su cient to minimize mass transfer limitations in the transesteri cation reaction.
Three dimensional response surface and the contour plots Figures 4b (1), b (2), 4c (1), c (2), 4d (1), and 4d (2) show the three-dimensional response surface and contour plots (2). Each curving contour represents an unlimited number of possible combinations of two test variables, with the other two remaining at zero. It is simple and convenient to comprehend the interactions between two components and to determine their optimum levels using contour plots. Figure 4a, depicts the link between expected and experimental biodiesel yield. It can be seen that the anticipated and experimental biodiesel yields are highly correlated (R 2 = 0.9956). The predicted and experimental values were reasonably close to one another (R 2 value near unity), indicating that the data t the model well. Figure 4b (1) and 4b (2) demonstrate the response to the interaction between methanol oil ratio and catalyst weight versus yield, as well as the related 3D response surface plot. These graphs show that better biodiesel yields occur when the methanol oil/ratio is 9:1, the catalyst weight is 4%, the reaction temperature is 60° C, and the reaction period is 4 hours. The oil to methanol molar ratio had only a little effect on synthesis at low catalyst concentrations, however at high catalyst concentrations, the oil to methanol molar ratio was signi cantly important for synthesis enhancement 35 . However, if the catalyst concentration is higher than the prescribed levels, the product will not separate. In other words, the transesteri cation reaction would be di cult to complete. As a result, the transesteri cation reaction was hampered by low catalyst concentrations and a higher methanol-to-oil ratio. When the catalyst was increased to a speci c level and the methanol to oil ratio was high, the yield merely improved. The contour plot revealed that a high biodiesel production (> 91.2%) may be achieved by using a combination of intermediate to high catalyst loading (3 to 5 wt.%) and a high methanol to oil molar ratio (6 to 12). The interaction effect of the methanol/oil ratio and temperature is shown in Figures 4c (1) and 4c (2). The maximum yield (91%) was obtained at a temperature of 60°C and a methanol/oil ratio of 9:1 according to the plots. The solubility of methanol in the oil increases as the temperature rises, as does the rate of reaction. In fact, at low temperatures, methanol is not soluble in the oil at all; when stirring begins, an emulsion appears 36 . On the other hand, a high amount of alcohol (more than 9:1) makes glycerol recovery di cult. This could be due to the stiochiometry of transesteri cation, which demands a 3:1 molar ratio of alcohol to triglycerides, and because this reaction involves the conversion of one ester and an alcohol into another, an excess of alcohol is utilized to drive the reaction to completion 37 .The contour and response surface plot of reaction time and methanol to oil ratio on production are shown in Figures 4d (1) and 4d (2). Increases in the methanol/oil ratio over 9:1 and reaction duration above 4 hours result in a better yield when the other parameters remain constant. In other words, as time passes, the methanol/oil ratio rises, resulting in a high yield. When a high molar ratio is used for a speci c period of reaction and catalyst weight, higher Methyl ester synthesis is signi cantly preferred 37 .

Conclusion
Waste cooking oil is a low-cost feedstock for methyl ester (biodiesel) production, as it has been successfully transesteri ed into biodiesel. The goal of the project was to use response surface methodology (RSM) and a central composite design to optimize methyl ester synthesis from waste cooking oil (CCD). The interaction effects of the ve primary parameters that affect methyl ester formation, methanol to oil molar ratio (A), catalyst weight (B), temperature (C), reaction duration (D), and agitation speed (E), were thoroughly investigated. According to the results of the experiments, the molar ratio of methanol to oil, temperature, and speed all have a substantial impact on the transesteri cation of WCO to methyl ester. The best conditions for base-catalyzed transesteri cation of methyl ester from WCO determined by RSM were: 4% catalyst weight, 9:1 methanol oil molar ratio, 60°C temperatures, 4-hour reaction time, and 350 rpm agitation. Under these conditions, the expected yield is 91.2 percent. The regression model was judged to be highly signi cant at a 95% con dence level since the correlation coe cients for R-Squared (0.9895), adjusted R-Squared (0.9877), and projected R-Squared (0.9221) were all extremely near to 1, indicating good correlation and predictive ability. WCO can be utilized as a feedstock because it is available in large quantities all over the world. The prospect of impregnating bentonite clay with sodium hydroxide solution as a catalyst was also highlighted in this work, and it was shown to be a strong solid-base catalyst for the transesteri cation of WCO with methanol to biodiesel.   Fourier transform infrared spectroscope of catalyst