Sustainable Biodiesel Production via Biogenic Catalyzed Transesterification of Baobab Oil Methyl Ester and Optimization Process

Abstract

Biomass diesel is one of the sustainable and renewable sources of energy envisaged to hold a prominent position in the world energy infrastructure. In this study, biodiesel was produced from baobab seed oil by transesterification using biogenic heterogeneous catalyst, derived from mixed wastes of white chicken eggshells and banana fruit peels. The production process was statistically analyzed using Box-Behnken Design-Response Surface Methodology (BBD-RSM). The influential transesterification reaction parameters investigated with their ranges include reaction time (40–80 min), molar ratio of oil to methanol (1:9–1:15) and catalyst weight (3–5 wt%). The nano-catalyst (CaO-BFP-850 NPs) was prepared by calcination at high temperature of 850 °C for 4 h, and its properties were found to contain majorly the basic elements of Ca and K when investigated with analytical instruments such as SEM, EDS, DSC-TGA, FT-IR, and XRD. The regeneration test of the CaO-BFP-850 NPs conducted showed it could be reused for more than four cycles with less catalytic efficiency reduction. The ideal conditions instituted by BBD-RSM was 75 min of reaction time, 12.8:1 molar ratio of oil to methanol, and 4.08 wt% CaO-BFP-850 at 65 °C and 650 rpm constant temperature and agitation speed respectively, with the validated biodiesel yield of 96.70 wt%. The assessment of the quality of the biodiesel produced showed compliance with the standard specifications of ASTM D6751, EN 14241, and SANS 833.

1. Introduction

Biomass energy holds a great prospect towards accomplishing the energy-related targets of the Sustainable Development Goals (SDGs) on climate change and other environmental benefits. Fossil fuel, as the most extensively utilized energy, accounts for about 75% of the global emission of greenhouse gases and roughly 90% of the emissions of carbon dioxide [1]. Its rate of consumption is projected to rise to about 28–60% in 2050 [2]. The ecological community is indeed endangered with this enormous increase due to the greenhouse gases emission and environmental pollution associated with its usage. The limitation of supplies and the tendency of depletion is another aspect of concern, as the importance of energy to human society cannot be underrated. Current research focus has been propelled toward sustainable diesel fuels to address the energy security and environmental challenges that are linked with the energy system. Adoption of sustainable green fuel in the transportation system will not only minimize the importation of fossil fuel sources but will also lessen the oil and gas extraction damage [3]. The deployment of renewable fuel such as biodiesel in the transportation sector appears to be the simplest step to transition into zero emission technology [4]. In line with the strategic objectives of SDGs target on climate change and access to affordable clean energy (goal 7), biodiesel fuel is the cleanest, most reliable and accessible potential local fuel with desirable features such as biodegradability, sustainability, non-toxicity, accessibility, clean burning, and affordability [1,2]. Among several solutions implemented for greenhouse gases reduction, the utilization of biodiesel in heavy duty vehicles is considered more advantageous compared to electrification strategy [5]. Electric vehicles (EVs) are generally deployed in shorter routes and operation compared to diesel vehicles (DVs). Biodiesel can be used in all diesel vehicles because of its good resemblance with fossil diesel in terms of the physio-chemical properties [6]. Therefore, they can be applied in the same engine system without any technical adjustment [7].

Biodiesel is obtained from bio-lipids (vegetable oils and animal lipids) as long chain fatty acids called mono-alkyl ester. Currently, over 85% of biodiesel fuels come from edible oil, which leads to the conflict of energy imbalance between food crops and energy resources [8]. To promote biodiesel commercialization, the focus has been shifted to waste oil, algae, and non-edible oils as well as underutilized seed oils. Tropical oil-bearing plants such as baobab seed oil have been reported to contain a high percentage of oil, which is highly rich in mono-saturated fatty acids such as oleic (30–42%), linoleic (20–35%), and palmitic (10–20%) acid. The Baobab plant, scientifically referred to as Adansonia digitata, is a species of Bombacaceae, belonging to the humongous tree category of Malvaceae. The tree is commonly found in tropical regions such as Western Africa, Southern African, Australia, and Madagascar. In South Africa, it is common in the warm part of Limpopo. It is adaptable to regions with low rainfall, as well as semi-arid regions and dry woodland. It grows to a height of 25 m and has a distinguished large trunk diameter of 10–12 cm. A mature tree bears up to 30 kg of fruits and is rich in vitamin C. The seed oil is reported to have a wider application in traditional medicine and cosmetic industries [9]. However, the use of baobab seed oil for biodiesel production has scarcely been investigated, and therefore, requires adequate information to be established as a potential biodiesel feedstock.

The most effective and widely used method for triglyceride conversion to methyl ester is the transesterification method. This method involves a chemical reaction process where triglyceride is reacted with an alcohol (usually methanol) and a catalyst to generate esters and glycerol (Figure 1). Catalyst plays an important part in boosting the efficacy of the reaction process alongside reducing the cost of production. Catalysts employed for biodiesel synthesis can be heterogeneous or homogeneous (base, acid, bi-functional, or enzyme). While homogeneous catalysts have well defined structures and high catalytic performance, they are not stable and cannot be recycled [10]. Their use in transesterification experiments is limited due to issues related to soap formation, toxicity, and huge wastewater generation [11]. Transesterification reactions with heterogeneous catalysts is much recently preferred over homogeneous ones due to some distinct advantages, which include economic benefits such as stability, easy separation, recoverability, and recyclability for several transesterification reaction cycles. The heterogeneous catalysts developed from biomass waste materials have a porous and functional structure together with superior performance. Additionally, they are cost-effective, and their usage reduces environmental waste while promoting a circular economy [12]. Agricultural wastes materials contain a high proportion of potassium and significant amount of other mineral elements and compounds. Some of the agro-waste materials that have been investigated include coco pod, avocado peels, pineapple leaves, moringa leaves, kola nut husk, plantain peels, banana peels, palm fruit bunch, pawpaw peels, and several others. Among these, banana peels have been extensively reported for their active performance as heterogeneous catalyst for biodiesel synthesis with the yields ranging from 95–100% [13,14,15]. The concentration of K (% mass fraction) contained in banana peel ash has been reported in the range of 40–99% when subjected to high temperature of calcination range of 500–700 °C [14,16,17,18]. The corresponding crystalline phase dominant mineral compounds was reported to increase in the order of KCl > K₂O > CaMgSiO₄, in addition to other mineral elements such as Na, Ca, Mg, P, Cl, Cu, Al, Fe, Mn, and Zn [14,15,19]. However, these mineral components differ and vary in quantity based on the sourced location, mood of preparation, and applied heat treatment to generate the catalyst. On the other hand, the literature indicated that waste animal shells and bones are composed of 96–98% of CaO with traces of MgO, ZnO, SrO, SiO, etc. when calcined at high temperature of 750–1000 °C [20]. Eggshells, as an excellent source of CaO, have been extensively studied for biodiesel synthesis due to its availability and easy processing steps [21,22,23]. However, bare CaO obtained by calcination is highly alkaline and usually suffers from low surface area and low catalytic activity [6,24]. Therefore, a successful attempt has been made in this study to improve its alkalinity, stability, catalytic performance, and reusability by doping it with active bio-metals from banana peels. Thus, the obtained green and cost-effective catalyst in this study can be used to replace the chemical catalysts with their attendance environmental issues.

One of the economic measures to achieving effective biodiesel production process is through application of RSM in transesterification experiments. This involves optimization of the influential operating parameters to maximize the yield at a reduced cost and time. In this study, the RSM-BBD was selected for this purpose based on its flexibility and economical advantage such as the potential of optimizing multi-factor problems with a reduced number of experiments using only three levels of factors [25]. BBD has been used severally in modeling the transesterification process, as it allows for an effective relationship between the process variables to achieve the best conditions for the maximum yield [14,18,21,22]. Thus, the objective of this study is to investigate the biodiesel production from baobab seed oil via a sustainable transesterification approach using an active biogenic catalyst, i.e., CaO-BFP-850 NPs prepared from the mixture of wastes white chicken eggshells and banana fruit peels. RSM was used for the optimization and statistical analysis of the process input variables interaction with yield(s). The chemo-physical properties of the methyl ester obtained from the process were further examined to ensure compliance to the required standards. Overall, the study contributes to the existing studies on a cleaner approach of production of renewable fuels to decarbonize global economy and combat climate change.

2. Materials and Methods

2.1. Materials

The baobab oil used for this study was supplied by Laboquip, Springfield Park, Durban, South Africa. The chemicals used for the biodiesel production and characterization were all analytical grades, which were used without further purification. These include ethanol (99.9% purity), methanol (99.9% purity), hexane (99%), potassium hydroxide, diethyl ether, phenolphthalein, cyclohexane, hydrochloric acid (37%) was purchased from Sigma-Aldrich (St. Louis, MO, USA), potassium iodide and Wij’s solution was purchased from minema chemicals, South Africa.

2.2. Preparation of the Catalyst

This study adapted a simple catalyst preparation and synthesis approach [6]. The distilled water was used in washing the waste materials: white chicken eggshells and banana (Musa Sapientum Linn species) fruit peels, which afterward were dried in the oven at 100 °C for 24 h. The dried materials were respectively crushed and milled to power and then sieved with a 50-µm mesh. The fine powder of each material was mixed at an equal proportion (1:1) in grams and heated in the muffle furnace for 4 h at 850 °C. The resultant biochar was pulverized and packed into a well-sealed bottle and kept in a desiccator for further use. The obtained nano-catalyst was labeled CaO-BFP-850 NPs, where BFP and NPs denote banana fruit peels and nanoparticles, respectively.

2.3. Catalyst Characterization

The catalyst properties were determined by performing various characterization analysis.

An SEM machine—(AURIGA, Zeiss, Jena, Germany) fitted with an EDX detector was used to assess the structure and elemental composition of the sample. The catalyst surface functional group was determined using an FT-IR spectrometer (Nicolet, Impact 410, Oxnard, CA, USA) at 4000–500 cm⁻¹ wavelength. X-ray Diffractometer (XRD) (Bruker AXS Karlsruhe, Germany) equipped with Cu Kα radiation was used to examine the crystallinity of the calcined CaO/BFP-850 NPs. The thermal behavior of the calcined catalyst was examined using the Thermo Gravimetric Analysis (TGA) (Netzsch, TG 209 F1 Libra, Selb, Germany).

Scherrer’s equation, given in Equation (1), was used to determine the crystallite size of the particles:

$$D={\displaystyle \frac{\kappa \lambda }{\mathsf{\beta }\mathrm{c}\mathrm{o}\mathrm{s}\mathsf{\theta }}}$$

(1)

where D = Crystallite size (nm), Κ = the Scherrer’s constant (K = 0.9), λ = X-ray wavelength (0.15406 nm), β represents the full width at half maximum (FWHM) in radians, and θ represents the Peaks position (Bragg-diffraction angle) in radians.

2.4. Empirical Design of the Transesterification Process

The empirical design and the data analysis statistic were done using Design of Expert (DoE) software, version 11.1.0.1 (Stat-Ease Inc., Minneapolis, MN, USA). Three parameters investigated in the modeling of the process maximization for the biodiesel yields include the molar ratio of methanol-to-oil (9:1–15:1), reaction time (40–80 min), and catalyst loading (3–5 wt%). The stirring speed and reaction temperature were held constant at 65 °C and 650 rpm [26]. The range of values selected for each parameter was obtained after some preliminary experiments conducted based on previous studies.

Table 1. The range of operation and factor levels.

Factors	Symbol	Unit	Coded Factor and Levels
			−1	0	1
Methanol/oil ratio	A	w/w	9:1	12:1	15:1
Catalyst loading	B	wt%	3	4	5
Process Reaction time	C	min	40	60	80

displayed the BBD design, which was used to generate 15 experimental runs (

Table 2. Experimental conditions with actual and predicted BOME yields.

Standard Order	Run	A	B	C	Actual BOME Yield	Predicted BOME Yield
14	1	12	4	60	96.55 ± 0.18	96.37
4	2	15	5	60	94.64 ± 0.05	94.59
15	3	12	4	60	96.31 ± 0.06	96.37
7	4	9	4	80	95.42 ± 0.15	95.27
12	5	12	5	80	94.62 ± 0.26	94.88
9	6	12	3	40	91.42 ± 0.26	91.16
13	7	12	4	60	96.26 ± 0.11	96.37
2	8	15	3	60	92.20 ± 0.11	92.31
10	9	12	5	40	93.57 ± 0.10	93.47
11	10	12	3	80	94.60 ± 0.11	94.71
6	11	15	4	40	95.14 ± 0.15	95.29
5	12	9	4	40	89.73 ± 0.22	89.95
1	13	9	3	60	90.80 ± 0.05	90.85
8	14	15	4	80	95.15 ± 1.78	94.93
3	15	9	5	60	91.16 ± 0.11	91.05

) from the regression model for data fittings. An analysis of multiple regression was used to calculate the model terms while the significance terms of the model were determined by the analysis of variance (ANOVA). An evaluation of the equitability of the model was also performed based on the regression parameters. The diagnostic plot was used to interpret and examine the fitness effect of the model. While the three-dimensional plots were used to describe the effect of the relationship between the three parameters with the yield, the equation of differential second order that best defined the parameters correlation with the yield is expressed in Equation (2):

$$Y=K_0+\sum _{i=1}^n{K}_i{\mathrm{ƒ}}_i+\sum _{i=i}^n{K}_{ii}{\mathrm{ƒ}}_i^2+\sum _{i=1}^{n-1}\sum _{j=2}^n{K}_{ij}{\mathrm{ƒ}}_i{\mathrm{ƒ}}_j+\xi$$

(2)

where Y represent the methyl ester yield, K₀ is the cut off, ƒ_i represent the linear variables coefficient, K_ii denote the quadratic variables coefficient, K_ij is the variables interactive coefficient, ƒ_i ƒ_j are the experimental variables, and ξ is the differential error.

2.5. Biodiesel Synthesis from Baobab Oil and Characterization

A single-step transesterification reaction process was carried out due to the reduced free fatty acid content of the baobab oil < 2.5%. This was carried out in a 500 mL three-necked round bottom flask assembled with a water-cooled reflux condenser and magnetic stirrer. The temperature was controlled at 65 °C with a thermocouple probe inserted in the reaction mixture.

A measured oil quantity of 50 g, followed by a calculated quantity of methanol and heterogeneous catalyst were used based on the requirement for each experiment in the experimental design presented in Table 2. The reaction content was agitated at 650 rpm to avoid catalyst suspension and to obtain a uniform temperature distribution in the reactor. The duration of each experiment was monitored in accordance with the time allotted in the design matrix (Table 2). Upon the reaction completion, the reaction content was centrifuged at 1500 rpm to isolate the solids. The remaining liquid was transferred into the separating funnel and left to stand for 12 h for phase separation [27]. The BOME layer at the top was separated from the base layer containing the glycerol and residual solid catalyst. The biodiesel yield was quantified gravimetrically using Equation (3). The retrieved catalyst was thoroughly washed in hexane, filtered and oven dried for 24 h at 120 °C for reuse in subsequent transesterification process.

$$BOMEyield(\%)={\displaystyle \frac{weightofobtainedBOMElayer\left(g\right)}{weightofBaobaboil\left(BO\right)sampleused\left(g\right)}}\times 100$$

(3)

2.6. Characterization of BOME

The fuel properties of BOME produced were examined to determine its quality. The cetane number, calorific value, and oxidative stability were determined following the methods of Yusup [28], Krisnangkura [29], and Demirbas [30], respectively. Other properties, such as density, moisture content, kinematic viscosity, acid value, FFA, and iodine value was obtained using AOAC (Association of official analytical chemists) [31] and ASTM (American Society for Testing and materials) [32] standards. The functional groups present in BO and BOME were analyzed using FT-IR spectrometer, while the percentage fatty acid compositions of the oil and biodiesel was assessed using GC-MS, as described in our previous study [33].

3. Results and Discussion

3.1. Synthesized Catalyst Characterization

3.1.1. FT-IR Analysis

The bands of available groups function present in the developed catalyst (CaO-BFP-850 NPs) were identified using the IR spectrum and are shown in Figure 2a. A similar spectrum and available functional groups were also described by [6]. The strong absorption band at 3642 cm⁻¹ is related to the tensile vibration of -OH bonding group due to absorbed water on the surface of catalyst [34]. This strong occurrence of the O-H stretching band is also accredited to the CaO hygroscopic nature, majorly generated from the eggshells [35]. Similar cases were reported for all animal waste-based materials used for heterogeneous catalyst synthesis, such as chicken bone [36], fish bones [37], and cow horns [38]. The characteristic absorption band at 1411 cm⁻¹ can be attributed to the tensile bending vibration of -C-O stretching of the carbonate species in K₂CO₃ [38,39]. The minor bands around 873 cm⁻¹ have been described to indicate the asymmetric tension of CO₃⁻² group [40]. The distinct band at 500 cm⁻¹ could be attributable to the stretching vibratory of -Ca-O in calcium carbonate. This is also aligned with the EDS results, which clearly indicate the present of the dominant elements of Ca and K contributed as a result of doping eggshells with banana peel ash and subjected to a high heat treatment. Thus, these two elements are the important ingredients responsible for the effective performance of CaO/BFP-850 catalyst in the transesterification process of BOME production.

3.1.2. XRD Analysis

The XRD pattern of CaO-BFP-850 NPs catalyst is depicted in Figure 2b. The figure displayed the crystalline characteristics of the catalyst sample subjected to a high temperature of 850 °C for 4 h. The peaks diffraction was investigated in 2θ range of 10–80° and were compared with the Joint Committee on Powder Diffraction Standards database (JCPDS: 004-0733, 041-1478, 043-1334). The analysis indicated the decomposition of the combined waste materials at high temperature into three compounds including metal oxide, chloride, and mixed metal compound of other elements present in minute quantities: CaO, KCl, and HN₄Al₃(SO₄)₂(OH)₆. The JCPDS study reveals that the peaks at 2θ values of 17.88, 33.96, 46.97, 50.68, and 63.03 are due to CaO, and the peaks 28.36, 40.70, and 54.12 indicate the presence of KCl, while the peaks at 25.43, 29.55, and 47.16 indicate the existence of HN₄Al₃(SO₄)₂(OH)₆. The presence of CaO is majorly contributed by eggshells, which has also been observed in other calcined animal resources such as fish bones, chicken bones, seashells [37,41,42]. While KCl is contributed by banana peels [15,19]. The HN₄Al₃(SO₄)₂(OH)₆ phase is the combined minor mineral constituents present in the biogenic materials which also contribute to the catalytic strength of the catalyst. These findings are also in support with the EDS analysis as the major compounds behind the effective performance of the catalyst in BOME production. The average crystallite size of CaO-BFP-850 NPs catalyst calculated using Scherrer ‘s equation (Equation (1)) was 80.02 nm.

3.1.3. EDX Analysis

The elemental composition of catalyst was confirmed by energy-dispersive X-ray spectroscopy analysis (EDX), mounted with a detector microscope. The CaO-BFP-850 NPs catalyst is made up of three major components, including calcium, oxygen, and potassium with percentage weight compositions of 41, 36, and 21%, respectively, while other elements, such as Cl, and Si, Al, and Na, were present in trace amounts < 2% (Figure 2c). The interaction between Ca and K resulted in a highly basic and well-developed porous structured catalyst, with high catalytic potential in the transesterification process. The high percentage of oxygen suggests that the metallic compound formed consist of oxygen atom, merged to form metal oxide at high heat treatment [38]. The EDX spectrum in Figure 2c indicated that CaO-BFP-850 NPs contain mainly Ca and K, which are suggested to be responsible for the high catalytic activity of catalyst in the transesterification experiment of BOME. The high content of Ca is dominantly contributed from the animal source, which is eggshells, while K, Cl, Al, and Si are majorly contributed from the agro source, which is banana peels.

3.1.4. SEM Analysis

The SEM image displaying the surface structure of CaO-BFP-850 NPs catalyst is presented in Figure 2d. The sample revealed a spongy-like and fibrous microstructural nature of the catalyst, occurring due to high calcination temperature. The distribution of the particle size of the irregular pore structure on the catalyst surface may be ascribed to the mineral impact of the two biomass materials on high heat treatment. A similar observation was also noticed for mixed ash of eggshells and papaya peels [43]. The porous nature of the catalyst suggests high activity in biodiesel production due to increase in surface area [6]. High temperature of calcination enhances evolution of volatiles matter and increases the number of pores. However, higher temperature of calcination above 900 °C could lead to sintering of active compounds, collapse of pores, reduction in chemical component, and overall reduction in biodiesel yield(s) [6]. Several reuses of a catalyst results in an increase of pore blockage, decrease of the surface area, and hindering the mass transport of the reactant.

3.1.5. DSC-TGA Analysis of CaO-BFP-850 NPs

The DSC-TGA analysis was employed to evaluate the heat flow and thermal stability of the CaO-BFP-8500 NPs catalyst, and the result is illustrated in Figure 2e. The weight loss occurred in three stages; the CaO-BFP-850 NPs displayed a weight loss of 7 wt% due to moisture evaporation from the material layers at temperatures between 70–150 °C. At the temperature range of 150–650 °C, 3 wt% weight loss of the catalyst was obtained, which might be attributed to the degradation of most carbon compounds and dihydroxylation of the material structure [44]. A huge weight loss of 10 wt% occurred at a temperature range of 650–900 °C, which is due to the disintegration and conversion of the organic compounds present in the material to gaseous compounds [16]. A huge quantity of volatile matter is released indicating significant combustion at this stage. In general, between the temperature ranges of 70 to 900 °C, the weight of the catalyst material significantly reduced to 20 wt%. This indicates that the temperature range of 850 °C selected in this study was enough to facilitate the decomposition and transition of the carbonaceous material to metal oxide. Thus, the combined material has higher thermal stability compared to previous works using bare CaO catalyst [42,45]. Thus, high temperature treatment of the catalyst improves it chemical and mechanical stability and prevent excessive leaching of the vital elements present in a catalyst and enhance reusability [27].

3.1.6. Reusability Test of CaO-BFP-850 NPs

One of the effective attributes of heterogeneous catalyst that determines its commercial suitability is the recyclability potential. This present study investigates the recyclability of the CaO-BFP-850 catalyst up to four cycles. The transesterification reaction process was performed using the established optimum experimental condition of the BBD. The catalyst was recovered after each successive transesterification experiment by centrifugation for 5 min at 1500 rpm. The spent catalyst separated was rinsed with hexane to eradicate filths such as glycerol and then dried at 100 °C. Figure 2f showed the graphical representation of the regeneration test of the CaO-BFP-850 catalyst. The result showed a minimal decrease in percentage biodiesel yield as the number of reused cycle increases. The maximum biodiesel yield obtained at fourth cycle was 88.20% without re-calcination of the recovered catalyst used in the production. Thus, this is higher when compared to other previously reported experiments with bare CaO calcined derived catalyst from animal wastes source; Ref. [38] recorded a significant decrease in biodiesel yield of less than 50% at the fourth cycle, when chicken and fish bones were combined to produced CaO catalyst. Moreover, Ref. [46] reported a significant reduction in yield of 6.2% when bare CaO catalyst produced from eggshells was used in their study. Again, Ref. [47] reported a low biodiesel yield of 67.57% at the first cycle, when plain CaO catalyst generate from white bivalve clam shell was used to trans-esterify waste frying oil. CaO tends to deactivate easily because of its hygroscopic nature when exposed to air or moisture. The leaching of its active site leads to extensive formation of calcium glyceroxide due to its reaction with glycerol during transesterification [48]. Thus, this makes it difficult for its subsequent reused and achievement of purer products after the reaction. However, attachment of support to CaO has been considered effective to increase its catalytic activity and reusability [49,50]. The combination of its source materials of production with agro-waste sources enable attachment of certain other elements and compounds to its surface to enhance its activity and reusability. Although this has been performed using homogeneous/chemical sources, this study finds it sustainable and eco-friendly to use biomass waste materials in achieving this purpose. Moreover, the re-calcination process of the spent catalyst is encouraged to enhance reusability [27]. This process sustains the lifespan of the catalyst and prevents easy weakening of the active sites through leaching. Overall, the CaO-BFP-850 NPs produced is a cost-effective, eco-friendly, non-toxic, readily available, stable, and effective catalyst for biodiesel production.

3.2. Transesterification Process Modeling and Optimization

The transesterification process of converting BO to BOME using the CaO-BFP-850 NPs catalyst was studied and statistically analyzed by RSM-BBD. The experimental conditions generated by BBD to investigate the three influential factors of the transesterification process, which include catalyst loading, reaction time and molar ratio of oil-to-methanol, are displayed in Table 2 with the actual and predicted BOME yields. The statistical analysis of the modeled equation was evaluated by analysis of variance (ANOVA) considered at the confidence probability level of p < 0.05, which is shown in

Table 3. Test of significance for every regression model coefficient.

Source	Sum of Squares	Degree of Freedom	Mean Square	F-Value	p-Value
Model	68.59	9	7.62	101.28	<0.0001
A	12.55	1	12.55	166.78	<0.0001
B	3.09	1	3.09	41.03	0.0014
C	12.33	1	12.33	163.80	<0.0001
AB	1.08	1	1.08	14.37	0.0127
AC	8.07	1	8.07	107.19	0.0001
BC	1.13	1	1.13	15.07	0.0116
A2	13.80	1	13.80	183.33	<0.0001
B2	18.53	1	18.53	246.30	<0.0001
C2	1.24	1	1.24	16.53	0.0097
Residual	0.3762	5	0.0752
Lack of Fit	0.3282	3	0.1094	4.55	0.1854
Pure Error	0.0481	2	0.0240
Cor Total	68.96	14
Std. Dev.	0.2743
Fits Statistics dataset
R2	0.9945	Adjusted R²	0.9847	Predicted R2	0.9223
Adeq Precision	28.6953	C.V. (%)	0.2923

Where A = Methanol/oil, B = catalyst loading, C = reaction Time, C.V. = coefficient of variance.

. The model effectiveness was confirmed by an F-value of 101.28 and p-value of <0.0001. The independent terms were all significant, but the influence of the oil-to-methanol ratio (A) and reaction time (C) with an F-value of 166.78 and 163.80 were stronger on the response than that of the catalyst loading with an F-value of 41.03. The interaction term of the oil-to-methanol ratio and time (AC) was also more significant than other interaction terms. The lack of fit was not significant, which shows the model’s good fit. The accuracy and effectiveness of the model prediction was appraised further by the fit statistics dataset (Table 3). The high coefficient of determination (R² = 0.9945) shows a good model interaction prediction. Thus, this value shows that the model generated can explain 99.45% of the data variability. The low value of the deviation standard of 0.27 and the percentage coefficient of variance (CV) of 0.29 implies that the model is reproducible with high consistency level and minimal deviation between the predicted and actual data. The mean value of 93.84 is also attested for the high accuracy of the model prediction. The high values of the adjusted and predicted R squares of 0.9847 and 0.9223 are in alliance with the high R² value of the model. The diagnostic plot that evaluates the consistency of the predicted and actual data is shown in Figure 3d, which also proved the accurate prediction of model. The quadratic mathematical equation that well described the correlation between the response and variable factors is presented in Equation (4):

$$BOME\left(\%\right)=96.37+1.25A+0.6212B+1.24C+0.5200AB-1.42AC-0.5325BC-1.93A^2-2.24B^2-0.5804C^2$$

(4)

3.2.1. Graphical and Numerical Optimization of the Process Parameters

Figure 3a–d depicts the graphical plot surface of 3D the interactive effect of the three process variables on the biodiesel yield(s). The mutual relationship between the molar ratio of oil-to-methanol and CaO-BFP-850 NPs loading with the yield at constant reaction time is indicated in Figure 3a. The methanol-to-oil ratio was investigated between the ranges of 9:1 to 15:1. The conversion rate of the biodiesel was observed to increase from 89.95 wt% to 96.55 wt% at the increasing ratio of oil-to-methanol from 9:1 to 12:1. An additional increase of the molar ratio up to 15:1 resulted in a yield reduction of the biodiesel produced. Thus, an excess amount of alcohol to oil enhances the proper contacts of reactants and hastens the reaction. On the other hand, the plethora of methanol poses a negative impact on the transesterification reaction by diminishing the catalyst concentration and hinders the rate of conversion of the triglyceride, leading to soap formation [35]. The catalyst loading level was studied between the ranges of 3 wt% to 5 wt%. An increase in the loading levels up to 4 wt% increases the yield of the biodiesel. A reduction in yield was studied at additional increase of the catalyst amount up to 5 wt%. This may be assigned to poor blending of the reactants in the reaction system due to the excess catalyst, which causes the reaction mixture to become viscous, and thus, promoting a surge in mass transfer resistance [27]. The result of ANOVA in Table 3 also supports the significant effect of both terms with methanol having a higher impact than the catalyst loading on the yield. The maximum yield of the biodiesel was recorded at a 1:12 oil-to-methanol ratio and 4 wt% of catalyst loading; beyond this point, a decrease in yield was recorded. Similar trends have also been recorded in literature [38,51,52].

The influence of the reaction time and alcohol-to-oil ratio is depicted in Figure 3b. The reaction process time was considered between the ranges of 40–80 min. The result indicated that BOME conversion was significantly increased, as the time of the reaction was intensified. The plot surface showed that the BOME yield was progressively increasing at the increase levels of both terms. This is also in support with the ANOVA evaluation of the significant interaction effect of both factors (AC) in

Table 4. A summary of the mixed bio-waste heterogeneous catalysts performance in the transesterification process.

Waste Renewable Resources	Calcination Condition	Oil Type	Transesterification Conditions				BOME Yield (%)	References
			Cat. Amt (wt%)	Met/Oil Ratio	Reaction Time (min)	Reaction Temp (°C)
Cocoa/kola nut/fluted pumpkin	500 °C, 4 h	Yellow oleander-rubber	1.5	9:1	40	55	95.02	[53]
Rice husk/eggshells	800 °C, 4 h	Palm oil	7.0	9:1	240	65	91.50	[54]
Eggshells/moringa leaves	800 °C, 4 h	Soya beans	2.0	12:1	78	65	94.3	[6]
Cocoa husk/plantain peels	500 °C, 4 h	Honne oil	4.5	15:1	90	65	98.98	[55]
Eggshells/papaya peels	900 °C, 3 h	Used cooking oil	3.78	14.9:1	80	65	91.20	[43]
Waste chicken/fish bones	1000 °C, 4 h	Used cooking oil	1.98	10:1	114	65	89.50	[36]
Plantain/cocoa/kola nut	500 °C, 4 h	Neem/Honne oil, rubber	1.15	12:1	60	150 W (microwave)	98.4	[56]
White eggshells/banana peels	800 °C, 4 h	Baobab oil	4.08	12.8:1	75	65	96.70	This study

, with the significant impact of the F-value (107.19) and p-value (0.0001). The optimum BOME yield of above 96% was obtained at the highest level of reaction process time of 80 min and a ratio of oil-to-methanol of 1:15. Thus, enough time and methanol are required to prevent glycerolysis reaction [37].

The surface plot of the mutual interaction effect of the reaction process time and CaO-BFP-850 NPs catalyst loading on yield is displayed in Figure 3c. The plot explained that BOME yield gradually increases with increasing reaction duration up to the highest level of 80 min, whereas the yield increases with the gradual increase of CaO-BFP-850 NPs catalyst loading up to 4 wt% and then dropped at a further increase. The regression analysis of the mutual interaction of both factors based on ANOVA evaluation showed a less significant impact as compared to other factors interactions. The optimum yield via this interaction was at 4 wt% catalyst amount and 80 min time of reaction.

3.2.2. Model Validation and Parameters Optimization

To investigate the overall optimum yield for the transesterification process, further mathematical optimization via criteria setting was carried out by placing the methanol-to-oil ratio at maximum, catalyst loading in range, reaction time at maximum, and the BOME yield at a target of 100% (Figure 3e). The conditions at optimal was established at a ratio of oil-to-methanol of 12.8:1, with a CaO-BFP-850 catalyst amount of 4.08 wt%, and the process time of reaction of 75 min at 65 °C constant temperature, with the predicted yield of 96.90% at a desirability of 88.7%. The condition was further authenticated by performing three consecutive experiments and the average yield of 96.70 wt% was achieved. Thus, the minimal percentage differences in the actual and predicted yield obtained indicates that the model prediction is accurate and applicable for adaptation. A summary of various biomass catalysts blends and their activities in transesterification reaction of vegetable oil are shown in Table 4. The yield obtained and the reaction conditions of this present study are comparable with the present studies.

3.3. Fatty Acid Composition of BO and BOME

The fatty acid compositions of BO and BOME are presented in

Table 5. Fatty acid composition of BO and BOME.

Systematic Name	Symbol	BO	BOME
		% Composition
Myristic acid	C14:0	0.35	0.49
Palmitic acid	C16:0	30.86	32.9
Stearic acid	C18:0	5.67	6.55
Arachidic acid	C20:0	1.85	0.88
Lignoceric acid	C24:0	-	0.36
Palmitoleic acid	C16:1	0.48	0.40
Oleic acid	C18:1	42.17	44.16
Linoleic acid	C18:2	12.46	14.01
Linolenic acid	C18:3	1.30	0.23
Others		4.82	-
Total Saturated		38.73	41.18
Total Unsaturated		56.41	58.80

. The results showed that the biodiesel produced from baobab oil was composed mainly of methyl esters of oleic acid (44.16%), palmitic acid (30.90%), and linoleic acid of (14.01%) followed by stearic acid (6.55%). Others, such as arachidic acid (0.88%), myristic acid (0.49%), palmitoleic (0.40%), lignoceric acid (0.36%), and linolenic acid (0.23%), are also present in small quantities. The results also indicated that the biodiesel produced consists of a total saturated acid of 38.73%, total unsaturated acid of 58.80%, out of which the mono-unsaturated acid (44.56%) was the highest, and total poly-unsaturated acid of 14.24%. It can also be observed from the table that the fatty acid profile of BOME is closely related to the oil feedstock (BO) from which it was produced. Similar observations were also reported by Betiku [57] and Tin [58], when kariya seed oil and sea mango Cerbera odollam oils were investigated for biodiesel production.

3.4. Infrared Spectroscopy of Baobab Oil (BO) and Baobab Methyl Ester (BOME)

The transesterification reaction of the biodiesel production was determined using infrared spectroscopy. The results obtained showed that the oil properties were changed after the transesterification process. The spectra of the FT-IR of BO and BOME are revealed in Figure 4a,b. The major absorption bands are 2922, 2853, 1744, 1169, 1097, and 722 cm⁻¹. Both the oil and FAME display similar functional groups with little disparity in area observed under the corresponding wavelength.

Table 6. Absorption peak properties for BO and BOME.

Wavenumber (cm−1)	Functional Group	Vibration Type	Intensity	Ref.
3006	=C-H	stretching	Weak	[59]
2922	-C-H (CH2)	Asymmetric stretching vibration	Very strong	[33]
2853	-C-H (CH2)	Symmetric	Very strong	[55]
1744	-C=O	Stretching	Very strong
1464	-CH2	Shear-type vibration	Medium	[59]
1377	-CH3	Bending vibration, symmetric deformation	Medium	[18]
1167	-C-O-C	Asymmetric stretching vibration	Very strong	[60]
1159	-CH2	Stretching	Medium	[61]
1097	C-CH2-C	Asymmetric stretching vibration	Strong	[59]
722	-CH2	Bending out of plane, rocking vibration	Medium	[55]

presents the absorption bands and their corresponding functional groups of baobab oil and its methyl ester. The carbonyl group (C-O) can be observed as strong absorption peak at 1744 cm⁻¹. The absence of obvious change between 1463 and 1436 cm⁻¹ in the BOME spectrum implies that no soap was present in the produced Biodiesel [59]. The peaks existing around 1300–1000 cm⁻¹ show the fingerprint region of multiplex spectra exhibited in both spectra [60]. This could be found more in the BOME than BO spectrum. The diagnostic peak identified for biodiesel is found at a wavelength of 1022 cm⁻¹ [57,61].

3.5. Biodiesel Quality Characterization

Table 7. Physio-chemical properties of BOME in comparison with standards.

Property	Test Method	BOME (at Optimum Conditions)	BOME (All Product)	ASTM D6751	EN 14214	SAN 833
State/color at room temp		Liquid/golden yellow	Liquid/golden yellow	liquid	liquid	liquid
Moisture content (%)	AOAC	<0.01	<0.01	<0.05	0.050% max	N/S
Refractive index	AOAC	1.467	1.467	N/S	N/S	N/S
Density at 25 °C (g/cm3)	AOAC	0.87	0.88	0.85	0.86–0.90	0.86–0.90
Kinematic Viscosity at 40 °C (mm2/s)	AOAC	3.40	4.50	1.9–6.0	3.5–5.0	3.5–5.0
Acid Value (mg KOH/g)	AOAC	0.20	0.35	0.5 max	0.5 max	0.5 max
FFA (%)	AOAC	0.10	0.17	0.2 max	0.25 max	0.2 max
Calorific value (MJ/kg)	[30]	40.14	40.14	N/S	35 min	N/S
Cetane number	[29]	55.45	55.45	47 min	51 min	51 min
Flash (°C)	ASTM D93 [32]	189	189	130 min	120 min	120 min
Cloud point (°C)	ASTM D2500 [62]	10	-	−3 to 12	N/S	−3 to 12
CFPP (°C)	[28]	−9	-	−32 to 5	N/S	−4 to 3
Copper strip at 50 °C, 3 h (Class)	ASTM D130 [62]	1	-	No. 3 max.	Class 1 min	No. 3 max

N/S: Not Specified.

presents the summary of the fuel properties of BOME produced from baobab oil both at optimum and general condition using CaO-BFP-850 NPs catalyst. The obtained values for all the biodiesel products were established within the EN 14214 [62], ASTM D6571 [62], and SAN 833 [32] specification standards. The values were found to be satisfactory with the standards, which show that it can be exploited as a fuel substitute to power the diesel engine. The existence of moisture in biodiesel is not desirable. It can lead to several problems, such as microbial growth, tank corrosion, participation in emulsion formation, and hydrolysis [63]. The percentage moisture content was found to be <0.01% at optimum and general conditions of the biodiesel produced, which shows that they were well dried and within the limit established by local and international standards. Fuel density that exceeds the specification limit affects engine performance and combustion characteristics. High density causes large injection droplets and leads to a reduction in ignition delay. The density of the all the biodiesel fuels produced have suitable values that are well compared with the standards. The density of BOME at optimum was found to be 0.887 g/cm³, which is a bit lower than the general BOME produced (0.888 g/cm³). However, the values are well established within the EN 14214 and SAN 833 limits (0.86–0.90), but lower than the ASTM D6751 standard. The kinematic viscosity is one of the key properties of biodiesel and is slightly greater than that of petrol-diesel. It allows for fuel atomization in the engine chamber. Low viscosity leads to easier pumping, atomization, and finer droplets. High viscosity leads to a cold start up and high-energy utilization during fuel spraying [64]. The viscosity value (3.40 mm²/s) for the optimum biodiesel was lower than general BOME product (4.50 mm²/s). The optimum fuel viscosity was slightly below EN 14214, and SAN 833 standards range of 3.5–5.0 mm²/s, while the general BOME produced was within the permissible range based of all standards. The high amount of acid influences the rate of fuel aging. According to the standard limits of ASTM D6751 and EN 14214, the highest acid value for biodiesel is recorded to be 0.5 mg KOH/g. The acid value of the produced biodiesel at optimum in this study was 0.20 mg KOH/g, while the general biodiesel fuel was higher (0.35 mg KOH/g) and all are within the permissible limit, which indicates that the fuel is less corrosive and will not damage the fuel system or produce many sediments [65]. A high acid value above the standard limits will lead to scale creation in the fuel system and degrade filters and pumped life [66]. The high calorific value and cetane number above the standard’s minimum specification indicate that the produced biodiesel has faster and better ignition properties and high-energy efficiency than the petroleum fuel [33]. The flash point of the produced biodiesel at optimum and general conditions was 189 °C. Thus, this shows that the flammability of the produced biodiesel was reduced as compared to petro-diesel (68.5 °C). Cloud point is the lowest temperature at which crystal formation in biodiesel first began. The cloud point of 6 °C and the cold filter plugging point of −22 °C for optimum and general biodiesel fuel were all within the standard requirements. All other properties of the produced biodiesel were within the standard specification limits.

4. Conclusions

The study mainly focused on the green process of biodiesel production from baobab seed oil via transesterification approach using a biogenic heterogeneous catalyst. The statistical analysis of the developed RSM-BBD model indicated good reliability and high predictive capacity with an R² value of 0.995 and CV of 0.29%. The FAME yield of 96.70 wt% was realized at optimum operational condition of CaO-BFP-850 loading of 4.08 wt%, methanol-to-oil molar ratio of 12.8:1 w/w, and reaction process time of 75 min at a fixed temperature of reaction of 65 °C and stirred speed of 650 rpm. The statistical result of the process showed that the most influencing parameter was oil to methanol ratio, followed by catalyst loading.

The description of the developed catalyst, CaO-BFP-850 NPs, through scientific instrument analysis such as FTIR, XRD, SEM, EDX, and DSC-TGA clearly indicated that it predominantly composed of Ca and K in form of oxide, chloride, and mixed metal sulphates and hydroxides. These compositions in addition to high heat treatment are suspected as the reason for its excellent catalytic performance during oil conversion. The synthesized catalyst demonstrates the reduction of excessive leaching and improved reusability potential with minimal decrease in subsequent biodiesel yield of <2% and a yield of 88.20% after four cycles. This is higher compared to previous studies with bare CaO–eggshell catalysts. Thus, the catalytic strength and easy deactivation of the CaO catalyst can be improved with attachment of other elements and compounds from agro-waste materials. The produced biodiesel from baobab oil using the CaO-BFP-850 NPs biogenic catalyst conformed well to the biodiesel recommended standards. This infers that the biodiesel developed can be exploited as sustainable fuel surrogate to fossil-derived diesel. Overall, the production techniques and design model developed from this study can serve as a template for the commercial biodiesel production process.