**3.3 GC-MS characterization result**

The various fatty acids present in the sea almond biodiesel are presented in **Table 4** in an increasing order of their retention time. A total of 38.14% saturated fatty acid, 39.92% monounsaturated fatty acid and 12.50% polyunsaturated fatty acids were found to be contained in the biodiesel The presence of high level of


#### **Table 4.**

*Fatty acid profile of the sea almond seed oil biodiesel.*

*Sea Almond as a Promising Feedstock for Green Diesel: Statistical Optimization and Power Rate… DOI: http://dx.doi.org/10.5772/intechopen.93880*

monounsaturated fatty acids in methyl esters translates to high biodiesel quality [26]. Therefore, the high levels of monounsaturated fatty acids (39.92%) contained in the sea almond seed oil methyl ester is expected to make it possess excellent fuel qualities. Also, the higher the amount of unsaturated fatty acid in a biodiesel sample, the better the cloud point but lower the oxidation stability which implies that the higher composition of unsaturated fatty acids in the methyl ester (52.42%) would therefore enhance its cold flow properties [27]. It is reported that the high viscous nature of waste frying oils is because of their high saturated and less unsaturated fatty acids and this could cause micro-crystal formations that are dangerous to engine fuel injection units [28, 29]. Therefore, the application of biodiesel derived from the kernel seed of sea almond would possess no inherent viscosity problem. According to the present investigation, the cetane number of sea almond seed oil methyl ester is 63.39, and this shows the presence of high amount of monounsaturated fatty acids [30]. Methyl esters derived from animal sources has cloud point of about 17°C which is quite above 13°C obtained from palm oil sourced biodiesel while conversely feedstocks with lower concentrations of saturated fatty acids produces methyl esters with very low cloud point (< 0°C) [30]. Basically, biodiesel properties such as cetane number, kinematic viscosity, oxidative stability and cold flow properties are the specifications that are required to be satisfied and these have high relationship with the biodiesel fatty acid structural composition [31, 32]. Knothe [33] has reported that exhaust emission, and heat of combustion are likewise influenced by the fatty acid composition while methyl oleate is reported to be the most desired fatty acid to furnish produced biodiesel with the above expected fuel properties [1].

#### **3.4 NMR characterization result**

carbon (C=C) unsaturated bonds can cause the biodiesel samples to remain in liquid state but may be liable to poor storage stability due to oxidation. This implies that the biodiesel would not need cold flow improver for better performance. All the absorptions corresponding to C-O and C=O stretches indicate that the biodiesel product contains ester functional groups typical to any biodiesel type, while the following groups: C–H, C=H, and O–H indicated biodegradability of the oil and produced biodiesel [11]. Significant differences were effected by the ester groups. The specific peak that appeared at 892.50 cm<sup>1</sup> possesses bending type of vibrations appearing at low energy and frequency region in the spectra. It indicates the presence of = C–H functional groups [4]. It is part of fatty acid methyl ester with unsaturated bond in the seed oil and ester [23]. The specific peaks found in the region of 1088.80 cm<sup>1</sup> and 1197.20 cm<sup>1</sup> show split stretching and rocking vibrations of the carbonyl group (C–O) for the triglyceride and its methyl ester respectively [24]. The bending and rocking vibrations of methyl group in the parent oil and its methyl ester appeared between 1317.66–1500.50 cm<sup>1</sup> and 1317.66–

The various fatty acids present in the sea almond biodiesel are presented in **Table 4** in an increasing order of their retention time. A total of 38.14% saturated fatty acid, 39.92% monounsaturated fatty acid and 12.50% polyunsaturated fatty acids were found to be contained in the biodiesel The presence of high level of

**Peak No. Retention time (min.) Fatty acid methyl ester Amount (%)** 1. 3.874 Capric acid 1.06 2. 4.017 Caprylic acid 1.14 3. 4.357 Stearic acid 1.24 4. 4.866 Eicosenic acid 8.14 5. 5.289 Erucic acid 0.75 6. 5.788 Palmitic acid 8.23 7. 6.729 Lignoceric acid 3.75 8. 7.243 Oleic acid 39.34 9. 8.922 α- Linolenic acid 9.07 10. 11.044 Palmtoleic acid 0.66 11. 11.281 Elaidic acid 1.09 12. 12.999 Arachidic acid 3.30 13. 14.569 Behenic acid 3.66 14. 16.888 Myristic acid 3.88 15. 18.367 Margaroleic acid 1.18 16. 22.223 Linoleic acid 0.72 17. 22.781 Gadoliec acid 0.12 18. 23.770 Lauric acid 1.66 19. 23.995 γ-linolenic acid 3.21 20. 23.875 Vaccenic acid 2.01

1555.12 cm<sup>1</sup> respectively [25].

Prunus

**Table 4.**

**116**

*Fatty acid profile of the sea almond seed oil biodiesel.*

**3.3 GC-MS characterization result**

Nuclear magnetic resonance spectroscopy (NMR) is one of the instrumental analytical techniques used to quantify the conversion of triglycerides in vegetable oils into s [34, 35]. It is therefore, considered as one of the promising techniques for the characterization of biodiesel. The percentage conversion of the parent oil into its biodiesel using integration values for methoxy and ɑ - carbonyl methylene protons [35] was found to be 95.7%%. Experimentally, the maximum sea almond yield obtained numerically as presented in **Table 2** and by GC maximum determination after 1 hr. were 94.21% and 93.01% respectively. All results are quite in good agreement and validate each other. The slight variation in conversion could be due to incomplete separation of FAME s from glycerine (by-product) and minor system errors as in the case of experimental and GC determinations respectively. The <sup>1</sup> H NMR spectrum of biodiesel from sea almond seed oil biodiesel is shown in **Figure 3a**. The specific peaks appearing at 0.452 ppm and 0.811 ppm for terminal methyl protons (C-CH3) appears as singlet. From the <sup>1</sup> H NMR, the peak around 0.452 are from the terminal alkyl methyl in the s [36]. **Figure 3b** shows the 13C NMR spectrum of biodiesel from the sea almond seed oil. The 13C NMR shows significant aliphatic composition (CH3) at the 24–28 ppm resonance [37] and for terminal carbon methylene at 17.774 pp. The peak at 124.629 ppm is typical of polycyclic aromatics structures [38]. Also, the peak at 167.288 ppm shows the presence of carbonyl carbon (-COO-) and O-aromatics (C-O) [34]. The peaks at 17.774–28.907 ppm could be attributed to terminal methyl groups. The unsaturation characteristics of s was confirmed by peaks appearing at δ124.629 ppm [34].

#### **3.5 RSM optimization of sea almond seed oil methanolysis process**

A central composite design (CCD) was applied to develop a correlation between the factors affecting transesterification reaction and the yield. The complete design

**Figure 3.** *(a) <sup>1</sup> H NMR spectrum of the biodiesel. (b) 13C NMR spectrum of the biodiesel.*

matrix, experimental and predicted responses is presented in **Table 5**. The experimental values of the content obtained were found to be in the range of ranged from 60 > actual value >95 wt %.

#### *3.5.1 The RSM quadratic model ANOVA*

The analysis of variance (ANOVA) of the RSM models (Linear, interactive linear, quadratic and cubic) were performed by considering the significance of the Fischer's F-value, lack of fit, degree of freedom (df) and R-squared (R2 ). The result showed that the quadratic model best-satisfied the above set criteria. Other relevant appraisal methods involved the determination of coefficient of determination (R2 ), adjusted coefficient of determination as well as coefficient of variation (C.V). These were applied to ascertain the adequacy of the model [13]. **Table 6** contains the effect of parameters using the second-order polynomial model. The following parameters X1, X2, X3, X1 X2, X1 X3, X2 X4, X1 <sup>2</sup> and X2 <sup>2</sup> are found to be significant (**Table 7**). Since the parameters whose square are significant have more effect on the sea almond seed oil methanolysis [39], it implies that temperature, reaction time and catalyst had much effect on the studied response. The Model Fischer's F-value of 5.75 implies the model is significant and implies that there is only a 0.09% chance that a "Model Fischer's F-Value" this large could occur due to disturbance. The "Lack of Fit Fischer's F-value" of 0.2429 implies the Lack of Fit is not significant relative to the pure error. There is a 24.29% chance that a "Lack of Fit F-value" this

large could occur due to disturbance or noise. Non-significant lack of fit is good. It shows that the effect of most independent variables on the sea almond seed oil base methanolysis was significantly high. The non-significant lack of fit is good because it shows that the model will be well fitted [40]. The adequate precision compares the range of predicted values to the average prediction error. "Adeq Precision" measures the signal to noise ratio and a ratio greater than 4 is desirable (**Table 7**). The ratio of 8.148 obtained shows an adequate signal. The coefficient of variation is the ratio of the standard deviation of estimate to the mean value of the observed

*The design matrix, experimental and predicted values of methanolysis process.*

**Run Factor 1 X1 (°C)**

**Table 5.**

**119**

**Factor 2 X2 (wt %)**

*DOI: http://dx.doi.org/10.5772/intechopen.93880*

**Factor 3 X3 (min.)** **Factor 4 X4**

*Sea Almond as a Promising Feedstock for Green Diesel: Statistical Optimization and Power Rate…*

1 40 1.0000 50 1:4000 79.8700 79.2754 0.5946 2 60 1.0000 50 1:4000 69.6400 71.6404 2.0014 3 40 2.0000 50 1:4000 65.9400 66.3137 0.3737 4 60 2.0000 50 1:4000 68.7100 69.5587 0.8487 5 40 1.0000 60 1:4000 83.9860 86.3597 2.3737 6 60 1.0000 60 1:4000 86.7560 87.6047 0.8487 7 40 2.0000 60 1:4000 83.0560 83.2380 0.1820 8 60 2.0000 60 1:4000 85.8260 86.0330 0.2070 9 40 1.0000 50 1:6000 66.8700 67.1823 0.3123 10 60 1.0000 50 1:6000 70.6400 70.3273 0.3127 11 40 2.0000 50 1:6000 64.9400 65.9606 1.0206 12 60 2.0000 50 1:6000 68.8100 69.7556 0.9456 13 40 1.0000 60 1:6000 84.9860 84.2066 0.7794 14 60 1.0000 60 1:6000 85.7560 86.9016 1.1456 15 40 2.0000 60 1:6000 85.0560 84.5349 0.5211 16 60 2.0000 60 1:6000 85.8260 87.3499 1.5239 17 30 1.5000 55 1:5000 73.5780 76.3376 2.7596 18 70 1.5000 55 1:5000 79.1180 79.3776 0.2596 19 50 0.5000 55 1:5000 77.2780 79.7043 2.4263 20 50 2.5000 55 1:5000 85.4180 86.0110 0.4070 21 50 1.5000 45 1:5000 59.2320 61.6583 2.4263 22 50 1.5000 65 1:5000 94.3640 94.1967 0.1673 23 50 1.5000 55 1:3000 76.3480 78.8357 2.4877 24 50 1.5000 55 1:7000 76.0420 77.3425 1.3005 25 50 1.5000 55 1:5000 75.9431 76.5521 0.6090 26 50 1.5000 55 1:5000 75.9431 76.5527 0.6090 27 50 1.5000 55 1:5000 75.9431 76.5521 0.6090 28 50 1.5000 55 1:5000 75.9431 76.5521 0.6090 29 50 1.5000 55 1:5000 75.9431 76.5521 0.6090 30 50 1.5000 55 1:5000 75.9431 76.5521 0.6090

**Experimental value (%)**

**Predicted value (%)** **Residual**


*Sea Almond as a Promising Feedstock for Green Diesel: Statistical Optimization and Power Rate… DOI: http://dx.doi.org/10.5772/intechopen.93880*

**Table 5.**

matrix, experimental and predicted responses is presented in **Table 5**. The experimental values of the content obtained were found to be in the range of ranged from

The analysis of variance (ANOVA) of the RSM models (Linear, interactive linear, quadratic and cubic) were performed by considering the significance of the

showed that the quadratic model best-satisfied the above set criteria. Other relevant appraisal methods involved the determination of coefficient of determination (R2

adjusted coefficient of determination as well as coefficient of variation (C.V). These were applied to ascertain the adequacy of the model [13]. **Table 6** contains the effect of parameters using the second-order polynomial model. The following

(**Table 7**). Since the parameters whose square are significant have more effect on the sea almond seed oil methanolysis [39], it implies that temperature, reaction time and catalyst had much effect on the studied response. The Model Fischer's F-value of 5.75 implies the model is significant and implies that there is only a 0.09% chance that a "Model Fischer's F-Value" this large could occur due to disturbance. The "Lack of Fit Fischer's F-value" of 0.2429 implies the Lack of Fit is not significant relative to the pure error. There is a 24.29% chance that a "Lack of Fit F-value" this

<sup>2</sup> and X2

). The result

<sup>2</sup> are found to be significant

),

Fischer's F-value, lack of fit, degree of freedom (df) and R-squared (R2

*H NMR spectrum of the biodiesel. (b) 13C NMR spectrum of the biodiesel.*

60 > actual value >95 wt %.

**Figure 3.** *(a) <sup>1</sup>*

Prunus

**118**

*3.5.1 The RSM quadratic model ANOVA*

parameters X1, X2, X3, X1 X2, X1 X3, X2 X4, X1

*The design matrix, experimental and predicted values of methanolysis process.*

large could occur due to disturbance or noise. Non-significant lack of fit is good. It shows that the effect of most independent variables on the sea almond seed oil base methanolysis was significantly high. The non-significant lack of fit is good because it shows that the model will be well fitted [40]. The adequate precision compares the range of predicted values to the average prediction error. "Adeq Precision" measures the signal to noise ratio and a ratio greater than 4 is desirable (**Table 7**). The ratio of 8.148 obtained shows an adequate signal. The coefficient of variation is the ratio of the standard deviation of estimate to the mean value of the observed

#### Prunus


*3.5.2 The RSM model equations*

*DOI: http://dx.doi.org/10.5772/intechopen.93880*

*a*

*b*

*c*

**121**

**Table 8.**

*The present report.*

*Mehdic and Kariminia [21].*

*Esonye et al. [4].*

The chosen models based on coded, actual and significant terms are presented in Eqs. (18)–(20) respectively. The coded equation is useful for identifying the relative impact of the factors by comparing the factors coefficients, while the equation in terms of actual factors can be used to make predictions about the response for actual levels of each factor [40]. Analyzing the obtained model, it is observed that increase in the levels X1X2, X1X3, X1X4 and X2X4 results in a decrease in sea almond seed oil biodiesel yield [13].

*Optimized transesterification conditions for sea almond compared with sweet almond and Iranian bitter almond.*

**s/n Operating variables Sea almond<sup>a</sup> Sweet almondb Iranian Bitter almond<sup>c</sup>**

*Sea Almond as a Promising Feedstock for Green Diesel: Statistical Optimization and Power Rate…*

 Reaction time (min.) 58.52 65.00 60 Catalyst conc. (wt%) 2.04 1.5 1.4 3 Alcohol/oil molar ratio 4.66 5 9.7 Temperature (°C) 50.03 50 35 Predicted yield (wt%) 93.09 94.36 94.7 6 Experimental validated yield (wt%) 92.58 - 96.7

SASO FAME yield %w ð Þ¼þ *=*w 86*:*83 þ 2*:*75 ∗ A þ 2*:*75 ∗ B þ 0*:*75 ∗C þ 1*:*08 ∗ D

þ 4*:*75000 ∗ Molar ratio

� 0*:*16250 ∗ Temperature ∗Cat Conc

� 003 ∗ Temperature ∗ Molar ratio þ 0*:*15000 ∗Cat Conc ∗ Rxn Time � 1*:*75000 ∗Cat Conc ∗ Molar ratio þ 0*:*037500 ∗ Rxn Time ∗ Molar ratio

SASO FAME yield %w ð Þ¼� *=*w 85*:*75000 þ 2*:*75833 ∗ Temperature

SASO FAME yield %w ð Þ¼� *=*w 85*:*75000 þ 2*:*75833 ∗ Temperature

� 3*:*25 ∗ A ∗ B � 3*:*00 ∗ A ∗C � 0*:*25 ∗ A ∗ D þ 1*:*50 ∗ B ∗C � 3*:*50 ∗ B ∗ D þ 0*:*75 ∗C∗ D � <sup>6</sup>*:*<sup>08</sup> <sup>∗</sup> A2 � <sup>3</sup>*:*<sup>33</sup> <sup>∗</sup> <sup>B</sup><sup>2</sup> � <sup>1</sup>*:*<sup>83</sup> <sup>∗</sup>C<sup>2</sup> � <sup>1</sup>*:*<sup>33</sup> <sup>∗</sup> <sup>D</sup><sup>2</sup>

þ 21*:*37500 ∗Cat Conc þ 2*:*42917 ∗ Reactionn Time

� 0*:*015000 ∗ Temperature ∗ Rxn Time � 6*:*25000E

� <sup>0</sup>*:*<sup>015208</sup> <sup>∗</sup> Temperature<sup>2</sup> � <sup>3</sup>*:*<sup>33333</sup> <sup>∗</sup>Cat Conc<sup>2</sup> � <sup>0</sup>*:*<sup>018333</sup> <sup>∗</sup> Rxn Time<sup>2</sup> � <sup>0</sup>*:*<sup>33333</sup> <sup>∗</sup> Molar ratio<sup>2</sup>

þ 21*:*37500 ∗Cat Conc þ 2*:*42917 ∗ Rxn Time

� 0*:*16250 ∗ Temperature ∗Cat Conc � 0*:*015000 ∗ Temperature ∗ Rxn Time þ 0*:*15000 ∗Cat Conc ∗ Rxn Time � 1*:*75000 ∗Cat Conc ∗ Molar ratio þ 0*:*037500 ∗ Rxn Time ∗ Molar ratio

� <sup>0</sup>*:*<sup>015208</sup> <sup>∗</sup> Temperature<sup>2</sup>

(18)

(19)

(20)

–3*:*33333 ∗Cat Conc<sup>2</sup>

#### **Table 6.**

*Sea almond seed oils FAME yield response surface quadratic model ANOVA.*


#### **Table 7.**

*The regression model summary.*

response and a measure of reproducibility and repeatability of the models [41]. Therefore, the C.V value of 6.75 shows the model is reasonably reproducible. Also, the R-squared of 0.9429 shows that more than 94% of the overall variability can be explained by the empirical models of the Equations. A given model significance can equally be validated when the standard deviation has a lower value than mean. Also, the smaller the PRESS-value the more the adequacy and significance of the model. Therefore, the PRESS-value obtained here supports the significance of the model. The adj. R-squared and the predicted R-squared values of 0.8562 and 0.6947 respectively for the quadratic model are in close agreement [42].

*Sea Almond as a Promising Feedstock for Green Diesel: Statistical Optimization and Power Rate… DOI: http://dx.doi.org/10.5772/intechopen.93880*


*Esonye et al. [4]. c Mehdic and Kariminia [21].*

#### **Table 8.**

*Optimized transesterification conditions for sea almond compared with sweet almond and Iranian bitter almond.*

#### *3.5.2 The RSM model equations*

The chosen models based on coded, actual and significant terms are presented in Eqs. (18)–(20) respectively. The coded equation is useful for identifying the relative impact of the factors by comparing the factors coefficients, while the equation in terms of actual factors can be used to make predictions about the response for actual levels of each factor [40]. Analyzing the obtained model, it is observed that increase in the levels X1X2, X1X3, X1X4 and X2X4 results in a decrease in sea almond seed oil biodiesel yield [13].

$$\begin{aligned} \text{SASSO FAME yield} \begin{aligned} (\text{6w/w}) &= +86.83 + 2.75 \ast \text{A} + 2.75 \ast \text{B} + 0.75 \ast \text{C} + 1.08 \ast \text{D} \\ &- 3.25 \ast \text{A} \ast \text{B} - 3.00 \ast \text{A} \ast \text{C} - 0.25 \ast \text{A} \ast \text{D} \\ &+ 1.50 \ast \text{B} \ast \text{C} - 3.50 \ast \text{B} \ast \text{D} + 0.75 \ast \text{C} \ast \text{D} \\ &- 6.08 \ast \text{A}^2 - 3.33 \ast \text{B}^2 - 1.83 \ast \text{C}^2 - 1.33 \ast \text{D}^2 \end{aligned} \tag{18}$$

$$\begin{array}{l} \text{SASO } \text{FAME yield } (\% \text{w/m}) = -85, 75000 + 2.75833 \text{\*} \text{Temperature} \\ &+ 21, 37500 \text{\*} \text{ Cartons} + 2.42917 \text{\*} \text{ Rectionum Time} \\ &+ 4.75000 \text{\*} \text{ Calor ratio} \\ &- 0.16250 \text{\*} \text{ Temperatur e } \text{Rx} \text{ Time} - 62.5000 \text{E} \\ &- 0.01500 \text{\*} \text{ Temperature } \text{\*} \text{ Rarn Time} - 62.5000 \text{E} \\ &- 0.0334 \text{\*} \text{Temperature} \times \text{Rarn Time} - 62.5000 \text{E} \\ &+ 0.15000 \text{\*} \text{ Cartons} \times \text{Rarn Time} \\ &- 1.75000 \text{\*} \text{ Cartons} \times \text{Rarn Time} \\ &+ 0.037500 \text{\*} \text{ Cartons} + \text{Normal} \times \text{Rarn Time} \\ &- 0.015208 \text{\*} \text{ Temperature } ^2 - 3.33333 \text{\*} \text{ Calor ratio} \\ &- 0.013333 \text{\*} \text{ Rxn Time}^2 - 0.33333 \text{\*} \text{ Molar ratio} \\ &+ 21.37500 \text{\*} \text{ Cartons} + 2.42917 \text{\*} \text{Run Time} \\ &- 0.16250 \text{\*} \text{ Temperature } \text{\*} \text{ Cartons} \\ &- 0.16250 \text{\*} \text{ Temperature } \text{\*} \text{ Cartons} \\ &+ 0.15000 \text{\*} \text{ Cartons} + \text{Run Time} \\ &- 1.75000 \text{\*} \text{ Cartons} + \text{Normal} \text{ rzino} \\ &+ 0$$

response and a measure of reproducibility and repeatability of the models [41]. Therefore, the C.V value of 6.75 shows the model is reasonably reproducible. Also, the R-squared of 0.9429 shows that more than 94% of the overall variability can be explained by the empirical models of the Equations. A given model significance can equally be validated when the standard deviation has a lower value than mean. Also, the smaller the PRESS-value the more the adequacy and significance of the model. Therefore, the PRESS-value obtained here supports the significance of the model. The adj. R-squared and the predicted R-squared values of 0.8562 and 0.6947

**Source Sum of Squares Df Mean square F value p-value**

AD- X1 X4 1 1 1 0.037 0.8495 BC- X2 X3 36 1 36 1.34 0.2647

CD X3 X4 9 1 9 0.34 0.5709

<sup>2</sup> 92.19 1 92.19 3.44 0.0835

<sup>2</sup> 48.76 1 48.76 1.82 0.1975

Residual 402.17 15 26.81

Pure Error 82.83 5 16.57

*Sea almond seed oils FAME yield response surface quadratic model ANOVA.*

Cor Total 2559.37 29

D-Metha/oil molar ratio (X4)

Prunus

A2 - X1

B<sup>2</sup> -X2

C2 - X3

D<sup>2</sup> - X4

**Table 6.**

**Table 7.**

**120**

*The regression model summary.*

Model 2157.2 14 154.09 5.75 0.0009 *Significant* A- Temperature (X1) 181.5 1 181.5 6.77 0.0200 *Significant* B-Catalyst Conc. (X2) 181.5 1 181.5 6.77 0.0200 *Significant* C-Reaction Time (X3) 190.2 1 190.2 7.09 0.0167 *Significant*

AB- X1 X2 169 1 169 6.3 0.024 *Significant* AC- X1 X3 144 1 144 5.37 0.035 *Significant*

BD X2 X4 196 1 196 7.31 0.0163 *Significant*

<sup>2</sup> 1015.05 1 1015.05 37.86 < 0.0001 *Significant*

<sup>2</sup> 304.76 1 304.76 11.37 0.0042 *Significant*

Lack of Fit 319.33 10 31.93 1.93 0.2429 not significant

Std. Dev. 5.18 R2 0.9429 Mean 76.77 Adj R<sup>2</sup> 0.8562 C.V. % 6.75 PredR<sup>2</sup> 0.6947 PRESS 958.64 Adeq Precision 8.148 RMSE 1.177 SEP 1.50150 MSE 1.217 AAD 0.5689

28.17 1 28.17 1.05 0.3216

**Prob > F**

respectively for the quadratic model are in close agreement [42].

#### *3.5.3 The production factors interactive effects*

**Figure 4A** shows the 3D plot of interactive effects of reaction time and catalyst concentrations on sea almond biodiesel yield while keeping both the reaction temperature and methanol/oil molar ratio at constant zero (0) coded levels. The smoother curve of catalyst concentration axis on the 3D plots and its lesser quadratic coefficient p-values result clearly portrays that its quadratic is more significant than that of reaction time. It means that reaction time has less impact on the response than the catalyst amount. Optimum sea almond seed oil biodiesel yield was obtained at about of 58 minutes and 2.0 wt% catalyst amount and beyond these points the yield retarded. Similar range of reaction condition has been reported where highest yield of neem seed oil biodiesel was obtained at 60 min at all catalyst concentration [39]. The reason could be because longer reaction time and excess catalyst promotes saponification reaction and increases in biodiesel viscosity respectively (Ofoefule, 2019). The impact of oil/methanol ratio and catalyst con-

*Sea Almond as a Promising Feedstock for Green Diesel: Statistical Optimization and Power Rate…*

It was observed from **Figure 4C** that simultaneous increase in both oil/methanol molar ratio and reaction time resulted in yield increase until a certain point (6,1 and 60 min.) when it began to decrease. The smooth curves of both variables indicate that they had very significant effect on the yield of sea almond seed oil biodiesel. Both factors have almost the same impact on the biodiesel yield. Beyond these maximum points, increase in reaction time could have favored the backward reaction due to reduced concentration of the sea almond seed triglyceride while increase in molar ratio could have resulted in poor separation and recovery of glycerol [43]. This is because higher methanol content has been reported to promote high dissolution of the transesterification by-product which accelerates the reversible reaction [44]. From **Figure 4D**, the effect of reaction time and temperature while keeping other factors constant at 5.0 and 1.5 wt% for methanol/oil molar ratio and catalyst concentration respectively is presented. It shows that temperature has higher impact on the FAME yield than reaction time. The ANOVA results still show that the interactive term of temperature and reaction time was very significant while both the linear and quadratic terms of temperature were all more significant than those of reaction time similar to reports of Ofoefule et al. [13]. Basically, the higher

*(a) Normal probability plots of residuals and (b) linear correlation experimental and predicted values from*

centration while keeping other factors constant at 50°C and 55 minutes is represented in **Figure 4B**. The impact of both factors appears equal on the sea almond seed oil biodiesel yield and increase in both factors results in significant increase in the response. The response was observed to increase at all alcohol/oil molar ratios. However, below 2.5 wt% catalyst concentration showed increase effect on the response. Maximum yield was obtained at the highest catalyst concentration and molar ratio. Optimum yield was not attained by this combinations and this could be due to the fact that higher factors are required for them or the other factors kept constant at zero (0) levels requires shifts from the central points. Although literature reports that above these ranges, the yield decreased significantly even with increase in catalyst concentration, this could be that the excess catalyst (NaOH) reacts with methanol to form soap or produced emulsions that made the produced biodiesel had difficulty in the separation [43]. The feedstock studied here

could have some deviating attributes or properties.

*DOI: http://dx.doi.org/10.5772/intechopen.93880*

**Figure 5.**

**123**

*sea almond seed oil methanolysis.*

#### **Figure 4.**

*The 3D response surface plot of the effects of the variables on sea almond FAME yield. (A). Reaction time and catalyst concentration. (B). Oil/methanol ratio and catalyst concentration. (C). Oil/methanol ratio and reaction time. (D). Reaction time and temperature. (E). Catalyst concentration and temperature. (F). Oil/ methanol ratio and temperature.*

#### *Sea Almond as a Promising Feedstock for Green Diesel: Statistical Optimization and Power Rate… DOI: http://dx.doi.org/10.5772/intechopen.93880*

concentration [39]. The reason could be because longer reaction time and excess catalyst promotes saponification reaction and increases in biodiesel viscosity respectively (Ofoefule, 2019). The impact of oil/methanol ratio and catalyst concentration while keeping other factors constant at 50°C and 55 minutes is represented in **Figure 4B**. The impact of both factors appears equal on the sea almond seed oil biodiesel yield and increase in both factors results in significant increase in the response. The response was observed to increase at all alcohol/oil molar ratios. However, below 2.5 wt% catalyst concentration showed increase effect on the response. Maximum yield was obtained at the highest catalyst concentration and molar ratio. Optimum yield was not attained by this combinations and this could be due to the fact that higher factors are required for them or the other factors kept constant at zero (0) levels requires shifts from the central points. Although literature reports that above these ranges, the yield decreased significantly even with increase in catalyst concentration, this could be that the excess catalyst (NaOH) reacts with methanol to form soap or produced emulsions that made the produced biodiesel had difficulty in the separation [43]. The feedstock studied here could have some deviating attributes or properties.

It was observed from **Figure 4C** that simultaneous increase in both oil/methanol molar ratio and reaction time resulted in yield increase until a certain point (6,1 and 60 min.) when it began to decrease. The smooth curves of both variables indicate that they had very significant effect on the yield of sea almond seed oil biodiesel. Both factors have almost the same impact on the biodiesel yield. Beyond these maximum points, increase in reaction time could have favored the backward reaction due to reduced concentration of the sea almond seed triglyceride while increase in molar ratio could have resulted in poor separation and recovery of glycerol [43]. This is because higher methanol content has been reported to promote high dissolution of the transesterification by-product which accelerates the reversible reaction [44]. From **Figure 4D**, the effect of reaction time and temperature while keeping other factors constant at 5.0 and 1.5 wt% for methanol/oil molar ratio and catalyst concentration respectively is presented. It shows that temperature has higher impact on the FAME yield than reaction time. The ANOVA results still show that the interactive term of temperature and reaction time was very significant while both the linear and quadratic terms of temperature were all more significant than those of reaction time similar to reports of Ofoefule et al. [13]. Basically, the higher

#### **Figure 5.**

*(a) Normal probability plots of residuals and (b) linear correlation experimental and predicted values from sea almond seed oil methanolysis.*

*3.5.3 The production factors interactive effects*

Prunus

**Figure 4.**

**122**

*methanol ratio and temperature.*

**Figure 4A** shows the 3D plot of interactive effects of reaction time and catalyst concentrations on sea almond biodiesel yield while keeping both the reaction temperature and methanol/oil molar ratio at constant zero (0) coded levels. The smoother curve of catalyst concentration axis on the 3D plots and its lesser quadratic coefficient p-values result clearly portrays that its quadratic is more significant than that of reaction time. It means that reaction time has less impact on the response than the catalyst amount. Optimum sea almond seed oil biodiesel yield was obtained at about of 58 minutes and 2.0 wt% catalyst amount and beyond these points the yield retarded. Similar range of reaction condition has been reported where highest yield of neem seed oil biodiesel was obtained at 60 min at all catalyst

*The 3D response surface plot of the effects of the variables on sea almond FAME yield. (A). Reaction time and catalyst concentration. (B). Oil/methanol ratio and catalyst concentration. (C). Oil/methanol ratio and reaction time. (D). Reaction time and temperature. (E). Catalyst concentration and temperature. (F). Oil/*

the temperature, the higher the reaction rate due to increase in the average kinetic energy of the reacting molecules according to Arrhenius theory [44]. The optimum temperature (60°C) would entail low cost of production as energy requirement for the seed oil methanolysis is comparatively low. Likewise, beyond 60 minutes reaction time, saponification might have been favored more due to less concentration of the reactants to push the reaction in the forward direction.

sweet almond and sea almond varieties and the catalyst applied for Iranian bitter was KOH against NaOH applied for the other varieties. Although, sweet almond had the highest reaction time, 7°C above sea almond and 5°C above Iranian bitter almond, sea almond from this study has about 0.5 wt% catalyst higher. Above all, the three almond varieties irrespective of their climatic origin and chemical composition have similar optimum conditions for the base methanolysis of their seed

*Sea Almond as a Promising Feedstock for Green Diesel: Statistical Optimization and Power Rate…*

**Figure 6ai**-**aiii** shows the variation of the intermediates of the sea almond methanolysis with time. The result obtained by observing the trend is similar to that previously reported by the authors [1]. However, there is a difference between the maximum points of last intermediates. From this work, the values were 4.8 wt% at 1.0 min and 4.98 wt% at 2.0 min at 55°C, 4.65 wt% at 1.0 min and 4.82 wt % at 2.0 min at 60°C and 4.51 wt% at 1.0 min and 4.70 at 2.0 min at 65°C. The maximum points of the last intermediates (DG) previously reported on African pear seer oil were 4.59, 4.20 and 4.10 wt% at 55°C, 60°C and 65°C respectively [1]. This difference could be due to the difference in the parent oil chemical properties. However, the results compare in values. Also, **Figure 6b** shows that the effect of temperature on the FAME yield clearly follows an increasing trend. It was observed that the difference in the concentration of FAME, within the studied temperature ranges was not significant at respective reaction times. It implies that other factors other than temperature such as reaction time, mixing intensity, etc. had more effects on the seed oil TG conversion to s. This agrees with the result of the optimization where the ANOVA showed that reaction time was more significant than

Least-square approximation was applied, in fitting a straight line to the experi-

**Glyceride Temperature (T) k (wt%/min) Ea (Kcal/mol.)**

mental data according to a model developed based on TG hydrolysis and the second-order reaction rate as shown in Eq. (21) ([8, 18]). In each case the coeffi-

) was determined.

**(K**�**<sup>1</sup> )** TG!DG <sup>55</sup> 3.05 0.00960 (R<sup>2</sup> = 0.98) 12.76 60 3.00 0.01010 (R<sup>2</sup> = 0.99) 65 2.96 0.01610 (R<sup>2</sup> = 0.98) DG!MG <sup>55</sup> 3.05 0.00838 (R<sup>2</sup> = 0.98) 15.83 60 3.00 0.00845 (R2 = 0.97) 65 2.96 0.01592 (R<sup>2</sup> = 0.97) MG!Gl <sup>55</sup> 3.05 0.01650 (R<sup>2</sup> = 0.98) 22.43 60 3.00 0.02930 (R<sup>2</sup> = 0.99) 65 2.96 0.04090 (R<sup>2</sup> = 0.98)

*Summary of the kinetics result for sea almond seed oil second-order irreversible methanolysis.*

oils (**Table 9**).

temperature.

**Table 9.**

**125**

cient of determination (R2

**3.6 Chemical kinetic study results**

*DOI: http://dx.doi.org/10.5772/intechopen.93880*

*3.6.1 Second order irreversible base transesterification model*

**(°C) 1/T x10<sup>3</sup>**

**Figure 4E** shows the 3D surface plot of the effects of catalyst concentration and temperature on the biodiesel yield of sea almond seed oil while keeping the reaction time and methanol/oil molar ratio constant. It shows the same trend with what was reported by Ofoefule et al. [13] on African pear seed oil biodiesel production optimization, although the catalyst concentration for the optimum yield in this report is 0.5 wt% less than what the previous report had presented. However, the explanation for the observed trend is due to increase in viscosity of the reaction composition at high catalyst concentration [13, 45]. **Figure 4F** shows the effects of oil/methanol molar ratio and temperature on the FAME yield. The catalyst concentration and reaction time was kept constant at 1.5 wt% and 55 minutes respectively. Temperature is found to have more significant impact on the response variable than methanol/oil molar ratio (as supported by the ANOVA result in **Table 6**). The FAME yield increased with increase in temperature irrespective of the value of the methanol/oil molar ratio. A reverse observation is possible if ethanol and different factor ranges were applied [43]. Optimum temperature was observed to be between 50 and 70°C in line with previous works [46].

The response values obtained by inserting the independent values are the predicted values of the model. These values are compared to the actual and experimental values. **Figure 5a** shows the normal probability plots of the residuals for clear investigations and diagnostic analysis. As it can be seen in **Figure 5b**, the data points were closely distributed along the diagonal axis. This implies that there is a good correlation between the actual and predicted values. This further corroborates the correlation between the R<sup>2</sup> and adjusted R2 values. By implication, the CCD is well fitted into the model and has the capability of carrying out the optimization exercise for methanolysis of the seed oil.

The result of the optimized conditions for the optimum response of sea almond seed oil is presented in **Table 8** in comparison with the results previously reported by [4] and Mehdic and Kariminia [21] on sweet almond and Iranian bitter almond respectively. This was carried out using numerical optimization tool function of the Design Expert 7.0.0 version. The flexibility of the software enabled the generation of a total of 11 solutions together with their respective desirability. The selected best solution based on the best declared desirability of 1.00 represents the optimized process conditions where the sea almond seed oil FAME maximum response was obtained as 93.09 wt%. The chosen conditions were equally considered based on the economic point of view by taking into cognizance the impart of temperature on energy requirement, amount of catalyst and alcohol/oil molar ratio on the raw material cost and reaction time on the overall production cost. To confirm the model's adequacy, a replicate experiment was performed using the optimum points derived from the process variables and a validated yield of 92.58 wt% was obtained. The obtained result presents a good correlation between the predicted and actual biodiesel yield at the optimum levels. It is pertinent to compare optimized conditions with previous works in the literature. Here, the optimized *modus operandi* from *T. catappa* (sea almond) is compared with other reported biodiesel production processes on similar almond varieties: sweet almond and Iranian bitter almond. The conditions quite compared in yield, reaction time, and fairly on catalyst concentration. However, Iranian bitter almond biodiesel temperature of 35°C is found to be quite low compared with 50°C recorded for the other varieties. This could be due to the fact that its alcohol/oil molar ratio was about twice the values recorded for

*Sea Almond as a Promising Feedstock for Green Diesel: Statistical Optimization and Power Rate… DOI: http://dx.doi.org/10.5772/intechopen.93880*

sweet almond and sea almond varieties and the catalyst applied for Iranian bitter was KOH against NaOH applied for the other varieties. Although, sweet almond had the highest reaction time, 7°C above sea almond and 5°C above Iranian bitter almond, sea almond from this study has about 0.5 wt% catalyst higher. Above all, the three almond varieties irrespective of their climatic origin and chemical composition have similar optimum conditions for the base methanolysis of their seed oils (**Table 9**).

#### **3.6 Chemical kinetic study results**

the temperature, the higher the reaction rate due to increase in the average kinetic energy of the reacting molecules according to Arrhenius theory [44]. The optimum temperature (60°C) would entail low cost of production as energy requirement for the seed oil methanolysis is comparatively low. Likewise, beyond 60 minutes reaction time, saponification might have been favored more due to less concentration of

**Figure 4E** shows the 3D surface plot of the effects of catalyst concentration and temperature on the biodiesel yield of sea almond seed oil while keeping the reaction time and methanol/oil molar ratio constant. It shows the same trend with what was reported by Ofoefule et al. [13] on African pear seed oil biodiesel production optimization, although the catalyst concentration for the optimum yield in this report is 0.5 wt% less than what the previous report had presented. However, the explanation for the observed trend is due to increase in viscosity of the reaction composition at high catalyst concentration [13, 45]. **Figure 4F** shows the effects of oil/methanol molar ratio and temperature on the FAME yield. The catalyst concentration and reaction time was kept constant at 1.5 wt% and 55 minutes respectively. Temperature is found to have more significant impact on the response variable than methanol/oil molar ratio (as supported by the ANOVA result in **Table 6**). The FAME yield increased with increase in temperature irrespective of the value of the methanol/oil molar ratio. A reverse observation is possible if ethanol and different factor ranges were applied [43]. Optimum temperature was observed to be between

The response values obtained by inserting the independent values are the predicted values of the model. These values are compared to the actual and experimental values. **Figure 5a** shows the normal probability plots of the residuals for clear investigations and diagnostic analysis. As it can be seen in **Figure 5b**, the data points were closely distributed along the diagonal axis. This implies that there is a good correlation between the actual and predicted values. This further corroborates the correlation between the R<sup>2</sup> and adjusted R2 values. By implication, the CCD is well fitted into the model and has the capability of carrying out the optimization

The result of the optimized conditions for the optimum response of sea almond seed oil is presented in **Table 8** in comparison with the results previously reported by [4] and Mehdic and Kariminia [21] on sweet almond and Iranian bitter almond respectively. This was carried out using numerical optimization tool function of the Design Expert 7.0.0 version. The flexibility of the software enabled the generation of a total of 11 solutions together with their respective desirability. The selected best solution based on the best declared desirability of 1.00 represents the optimized process conditions where the sea almond seed oil FAME maximum response was obtained as 93.09 wt%. The chosen conditions were equally considered based on the economic point of view by taking into cognizance the impart of temperature on energy requirement, amount of catalyst and alcohol/oil molar ratio on the raw material cost and reaction time on the overall production cost. To confirm the model's adequacy, a replicate experiment was performed using the optimum points derived from the process variables and a validated yield of 92.58 wt% was obtained. The obtained result presents a good correlation between the predicted and actual biodiesel yield at the optimum levels. It is pertinent to compare optimized conditions with previous works in the literature. Here, the optimized *modus operandi* from *T. catappa* (sea almond) is compared with other reported biodiesel production processes on similar almond varieties: sweet almond and Iranian bitter almond. The conditions quite compared in yield, reaction time, and fairly on catalyst concentration. However, Iranian bitter almond biodiesel temperature of 35°C is found to be quite low compared with 50°C recorded for the other varieties. This could be due to the fact that its alcohol/oil molar ratio was about twice the values recorded for

the reactants to push the reaction in the forward direction.

Prunus

50 and 70°C in line with previous works [46].

exercise for methanolysis of the seed oil.

**124**

**Figure 6ai**-**aiii** shows the variation of the intermediates of the sea almond methanolysis with time. The result obtained by observing the trend is similar to that previously reported by the authors [1]. However, there is a difference between the maximum points of last intermediates. From this work, the values were 4.8 wt% at 1.0 min and 4.98 wt% at 2.0 min at 55°C, 4.65 wt% at 1.0 min and 4.82 wt % at 2.0 min at 60°C and 4.51 wt% at 1.0 min and 4.70 at 2.0 min at 65°C. The maximum points of the last intermediates (DG) previously reported on African pear seer oil were 4.59, 4.20 and 4.10 wt% at 55°C, 60°C and 65°C respectively [1]. This difference could be due to the difference in the parent oil chemical properties. However, the results compare in values. Also, **Figure 6b** shows that the effect of temperature on the FAME yield clearly follows an increasing trend. It was observed that the difference in the concentration of FAME, within the studied temperature ranges was not significant at respective reaction times. It implies that other factors other than temperature such as reaction time, mixing intensity, etc. had more effects on the seed oil TG conversion to s. This agrees with the result of the optimization where the ANOVA showed that reaction time was more significant than temperature.

#### *3.6.1 Second order irreversible base transesterification model*

Least-square approximation was applied, in fitting a straight line to the experimental data according to a model developed based on TG hydrolysis and the second-order reaction rate as shown in Eq. (21) ([8, 18]). In each case the coefficient of determination (R2 ) was determined.


**Table 9.** *Summary of the kinetics result for sea almond seed oil second-order irreversible methanolysis.*

**Figure 6.**

*(a) Progress of intermediates at various temperatures at the initial stage. (b) Effect of reaction temperature on the seed oil methanolysis.*

$$\frac{-\mathbf{d}[\mathbf{TG}]}{\mathbf{dt}} = \mathbf{k}[\mathbf{TG}]^2 \tag{21}$$

Integration of Eq. (21) gives Eq. (22).

$$k\_{\rm Tg}t = \frac{1}{[\rm TG]} - \frac{1}{[\rm TG0]}\tag{22}$$

all the steps have positive activation energy and this supports the endothermic char-

Triglyceride 55 0.0429 0.81

*Second-order reaction irreversible model of (a) triglycerides, (b) diglycerides and (c) monoglycerides hydrolysis.*

*Sea Almond as a Promising Feedstock for Green Diesel: Statistical Optimization and Power Rate…*

*DOI: http://dx.doi.org/10.5772/intechopen.93880*

60 0.0476 0.80 65 0.0458 0.77

**) R2**

By ignoring the intermediate reactions of diglyceride and monoglyceride, the three steps have been combined in a single step [47]. However, due to the high molar ratio of methanol to oil, the change in methanol concentration can be considered as constant during reaction. This means that by taking methanol in excess, its concentration does not change the reaction order and it behaves as a first order chemical reaction [19]. The overall pseudo rate constants obtained from the slopes of the straight line plots of ln [TG] against t as shown in **Figure 9** are contained in **Table 10** for sea almond biodiesel. As can be seen from **Figure 9**, in the reactions conducted at 55, 60 and 65°C, there was a decrease in the coefficient of determination for the pseudo first-order kinetic model. **Figure 10** shows that the reaction at these temperatures does not fit the pseudo first-order reaction kinetic model better. This is supported by the lower values of coefficient of determination obtained from the first-order fitted plots (R<sup>2</sup> < 0.80) against high coefficient of determination obtained on the second-order irreversible kinetic model (R2 > 0.97). Similar results have been reported on the kinetics of hydrolysis of *Nigella sativa* (*black cumin*) seed

acteristics of conventional transesterification process (**Figure 8**) [1].

*Summary of the rate constants for the first-order irreversible methanolysis.*

**Glyceride Temperature (°C) Reaction rate constant (min<sup>1</sup>**

*3.6.2 First-order irreversible model*

**Table 10.**

**127**

**Figure 7.**

Where k is the overall pseudo-rate constant, t is the reaction time, TG0 is the initial triglyceride concentration.

A plot of reaction time (t) against <sup>1</sup> ½ � TG gave a straight line as shown in **Figure 7** with high values of coefficient (R<sup>2</sup> ) (**Table 9**) to show that the model is valid. The plot for the three temperatures (55, 60 and 65°C) is shown in **Figure 7a**, the slope is kTG (wt%�<sup>1</sup> min). It is observed that k fairly increased with temperature. Finally, activation energies of the reaction taking place were estimated using the calculated rate constants and temperatures at which they were observed in Arrhenius equation (Eq. (17)).

The DG and MG relationship with time followed the same trend (**Figure 7b** and **c**) with that of TG. There appears to be a very close similarity in the values of activation energy obtained in this study to the previous works [8] more especially in the Triglyceride and Diglycerides hydrolysis. However, the rate constants were found to be four (4) times higher and two (2) times lower than those reported by Darnoko and Cheryan [8] on palm oil base methanolysis and Reyero et al. [6] on sun flower baseethanolysis. The choice of feedstocks, alcohol and other factors like temperature could have resulted in the slight differences. Also, the ratio constants increase with temperature follows a trend of kTG < kDG < kMG in values. After 60 mins reaction time, the highest conversion was above 90% and it is found to be in the same range with what many other researchers have reported [1]. The hydrolysis of TG to DG is observed to be the rate determining step since it is the slowest (with smallest k) while the DG conversion to glycerol is most favored by high temperature. It is observed that *Sea Almond as a Promising Feedstock for Green Diesel: Statistical Optimization and Power Rate… DOI: http://dx.doi.org/10.5772/intechopen.93880*

**Figure 7.** *Second-order reaction irreversible model of (a) triglycerides, (b) diglycerides and (c) monoglycerides hydrolysis.*


#### **Table 10.**

�d TG ½ �

*(a) Progress of intermediates at various temperatures at the initial stage. (b) Effect of reaction temperature on*

*kTg <sup>t</sup>* <sup>¼</sup> <sup>1</sup>

½ � TG � <sup>1</sup>

Where k is the overall pseudo-rate constant, t is the reaction time, TG0 is the

plot for the three temperatures (55, 60 and 65°C) is shown in **Figure 7a**, the slope is

activation energies of the reaction taking place were estimated using the calculated rate constants and temperatures at which they were observed in Arrhenius equation

The DG and MG relationship with time followed the same trend (**Figure 7b** and **c**) with that of TG. There appears to be a very close similarity in the values of activation energy obtained in this study to the previous works [8] more especially in the Triglyceride and Diglycerides hydrolysis. However, the rate constants were found to be four (4) times higher and two (2) times lower than those reported by Darnoko and Cheryan [8] on palm oil base methanolysis and Reyero et al. [6] on sun flower baseethanolysis. The choice of feedstocks, alcohol and other factors like temperature could have resulted in the slight differences. Also, the ratio constants increase with temperature follows a trend of kTG < kDG < kMG in values. After 60 mins reaction time, the highest conversion was above 90% and it is found to be in the same range with what many other researchers have reported [1]. The hydrolysis of TG to DG is observed to be the rate determining step since it is the slowest (with smallest k) while the DG conversion to glycerol is most favored by high temperature. It is observed that

min). It is observed that k fairly increased with temperature. Finally,

Integration of Eq. (21) gives Eq. (22).

A plot of reaction time (t) against <sup>1</sup>

initial triglyceride concentration.

with high values of coefficient (R<sup>2</sup>

kTG (wt%�<sup>1</sup>

**Figure 6.**

Prunus

*the seed oil methanolysis.*

(Eq. (17)).

**126**

dt <sup>¼</sup> k TG ½ �<sup>2</sup> (21)

½ � TG gave a straight line as shown in **Figure 7**

) (**Table 9**) to show that the model is valid. The

½ � TG0 (22)

*Summary of the rate constants for the first-order irreversible methanolysis.*

all the steps have positive activation energy and this supports the endothermic characteristics of conventional transesterification process (**Figure 8**) [1].

#### *3.6.2 First-order irreversible model*

By ignoring the intermediate reactions of diglyceride and monoglyceride, the three steps have been combined in a single step [47]. However, due to the high molar ratio of methanol to oil, the change in methanol concentration can be considered as constant during reaction. This means that by taking methanol in excess, its concentration does not change the reaction order and it behaves as a first order chemical reaction [19]. The overall pseudo rate constants obtained from the slopes of the straight line plots of ln [TG] against t as shown in **Figure 9** are contained in **Table 10** for sea almond biodiesel. As can be seen from **Figure 9**, in the reactions conducted at 55, 60 and 65°C, there was a decrease in the coefficient of determination for the pseudo first-order kinetic model. **Figure 10** shows that the reaction at these temperatures does not fit the pseudo first-order reaction kinetic model better. This is supported by the lower values of coefficient of determination obtained from the first-order fitted plots (R<sup>2</sup> < 0.80) against high coefficient of determination obtained on the second-order irreversible kinetic model (R2 > 0.97). Similar results have been reported on the kinetics of hydrolysis of *Nigella sativa* (*black cumin*) seed

**Figure 8.** *Arrhenius plot of irreversible second order model reaction rate versus temperature.*

**4. Conclusion**

*First-order plot of the triglycerides hydrolysis.*

*DOI: http://dx.doi.org/10.5772/intechopen.93880*

**Figure 10.**

biodiesel from *T. cattapa* on industrial scale.

public, commercial or not-for-profit-sectors.

**Acknowledgements**

**Conflict of interest**

equipment.

**Funding**

**129**

The statistical optimization and chemical reaction kinetics of consecutive irreversible second order alkali- transesterification of *terminalia cattapa* seed oil has been successfully achieved and reported. RSM from Design Expert 7.0.0 version software was used for optimizing and predicting the process conditions in line with standard methodologies. The optimum conditions of base methanolysis process of the sea almond seed oil was obtained at favorable economic standpoint considering cheap materials requirement, low energy consumption and fast production rate. At low temperatures and latter stages, the methanolysis progresses very slowly and followed first order kinetic model but the irreversible second-order model of the power rate law best described the conversion of triglycerides with time at all stages. The data generated from the statistical optimization and chemical kinetics evaluations are found to be complimentary. The 's unsaturated characteristics would enhance its cold flow properties. The fuel properties of the biodiesel produced compared well with international standards. This research would help in commercial production of

*Sea Almond as a Promising Feedstock for Green Diesel: Statistical Optimization and Power Rate…*

The authors would like to thank the staff and management of the PZ/NOTAP Chemical Engineering laboratory of Alex Ekwueme Federal University, Abakaliki, Nigeria for the availability of the laboratory facilities, apparatus and analytical

This research did not receive any specific grant from any funding agent in

The authors hereby declare no competing financial interest.

**Figure 9.** *First-order plot of the latter stage (from 20 minutes) triglycerides hydrolysis.*

oil catalyzed by native lipase in ground seed where pseudo first-order rate equation at 20, 30 and 40°C; and the pseudo second-order equation at 50, 60 and 70°C [48]. Therefore, it could be that hydrolysis of some oils to s follows first-order irreversible kinetic models at low temperature ranges (20–40°C). The low temperature ranges is reported to favor the activity of native lipase better than at higher temperatures and this resulted in different mechanisms. But such low temperatures would not favor maximum ester yield in this study because they are far below the reported optimum temperature (Darnako and Cheryan, 2000). Darnako and Cheryan, 2000, has observed that at latter reaction stages (beyond 30 mins) of palm oil hydrolysis to, the first-order or zero-order reaction model is the best fitted. Similar observation was made on this study whereas from 20 minutes reaction, the reaction follows first-order model with high coefficient of determination (R2 > 0.94). This is shown in **Figure 10**. These stages showed low reaction rate due to reduction in the reactants concentration. It implies that at low temperatures and latter stages of methanolysis of the vegeatble oils progesses very slowly and follow first-order kinetic model.

*Sea Almond as a Promising Feedstock for Green Diesel: Statistical Optimization and Power Rate… DOI: http://dx.doi.org/10.5772/intechopen.93880*

**Figure 10.** *First-order plot of the triglycerides hydrolysis.*

## **4. Conclusion**

The statistical optimization and chemical reaction kinetics of consecutive irreversible second order alkali- transesterification of *terminalia cattapa* seed oil has been successfully achieved and reported. RSM from Design Expert 7.0.0 version software was used for optimizing and predicting the process conditions in line with standard methodologies. The optimum conditions of base methanolysis process of the sea almond seed oil was obtained at favorable economic standpoint considering cheap materials requirement, low energy consumption and fast production rate. At low temperatures and latter stages, the methanolysis progresses very slowly and followed first order kinetic model but the irreversible second-order model of the power rate law best described the conversion of triglycerides with time at all stages. The data generated from the statistical optimization and chemical kinetics evaluations are found to be complimentary. The 's unsaturated characteristics would enhance its cold flow properties. The fuel properties of the biodiesel produced compared well with international standards. This research would help in commercial production of biodiesel from *T. cattapa* on industrial scale.

#### **Acknowledgements**

The authors would like to thank the staff and management of the PZ/NOTAP Chemical Engineering laboratory of Alex Ekwueme Federal University, Abakaliki, Nigeria for the availability of the laboratory facilities, apparatus and analytical equipment.

#### **Conflict of interest**

The authors hereby declare no competing financial interest.

#### **Funding**

This research did not receive any specific grant from any funding agent in public, commercial or not-for-profit-sectors.

oil catalyzed by native lipase in ground seed where pseudo first-order rate equation at 20, 30 and 40°C; and the pseudo second-order equation at 50, 60 and 70°C [48]. Therefore, it could be that hydrolysis of some oils to s follows first-order irreversible kinetic models at low temperature ranges (20–40°C). The low temperature ranges is reported to favor the activity of native lipase better than at higher temperatures and this resulted in different mechanisms. But such low temperatures would not favor maximum ester yield in this study because they are far below the reported optimum temperature (Darnako and Cheryan, 2000). Darnako and Cheryan, 2000, has observed that at latter reaction stages (beyond 30 mins) of palm oil hydrolysis to, the first-order or zero-order reaction model is the best fitted. Similar observation was made on this study whereas from 20 minutes reaction, the

*First-order plot of the latter stage (from 20 minutes) triglycerides hydrolysis.*

*Arrhenius plot of irreversible second order model reaction rate versus temperature.*

reaction follows first-order model with high coefficient of determination

first-order kinetic model.

**128**

**Figure 9.**

**Figure 8.**

Prunus

(R2 > 0.94). This is shown in **Figure 10**. These stages showed low reaction rate due to reduction in the reactants concentration. It implies that at low temperatures and latter stages of methanolysis of the vegeatble oils progesses very slowly and follow

Prunus
