**3.1 Material and equipment**

This study focus on the extraction of oil from moringa oleifera seed by means of solvent extraction and production of bio-ethanol from rice husk using alkali as the hydrolising agent and zymomonas for fermentation. The entire chemicals used in this study are of analytical grade (98-99.5%). They include hexane, ethanol, iodine, sodium hydroxide, calcium oxide, potasium iodide and potasium hydroxide. The equipments used are mortar and pestle, sieve, electronic weighing balance, thimble, measuring cylinder, stop watch, pH meter, electric oven and distillation column. The moringa oleifera seeds and rice husk used in this study were collected in Bosso Estate, Minna, Niger State, Nigeria.

#### **3.2 The 2k factorial experimental design**

250 Sustainable Growth and Applications in Renewable Energy Sources

Fig. 2.2. Moringa oleifera (a) Dried pods (b) Seed kernel with husks (c) seed kernel without

The refined oil is clear, odourless and resists rancidity like any other botanical oil. The seed cake remaining after the oil extraction can be used as fertilizer or as flocculants to treat turbid water. The leaves are highly nutritious, being a significant source of beta-carotene, Vitamin C, protein, iron and potassium; it is consumed mostly among the Hausas in Northern Nigeria. In addition to being used fresh as a substitute for spinach, the leaves are

This study focus on the extraction of oil from moringa oleifera seed by means of solvent extraction and production of bio-ethanol from rice husk using alkali as the hydrolising agent and zymomonas for fermentation. The entire chemicals used in this study are of analytical grade (98-99.5%). They include hexane, ethanol, iodine, sodium hydroxide, calcium oxide, potasium iodide and potasium hydroxide. The equipments used are mortar and pestle, sieve, electronic weighing balance, thimble, measuring cylinder, stop watch, pH meter,

commonly dried and processed into powder and used in soups and sauces.

husk

**3. Methodology** 

**3.1 Material and equipment** 

When several factors are of interest in an experiment a factorial method of analysis is used in order to study the effect of individual factor and its interaction with other factors to economize the experimental resources (Azeez, 2005;Zhang and Huang, 2011; Wang et al., 2011). In this study, three factors namely temperature, particle size and resident time are of interest while agitation was kept constant. This gives rise to three-factor factorial experiment; the factors are tested at high and low levels. When three factors are tested at two levels as applicable in this study, it is denoted by 23 factorial; thus there exist eight (23) treatment combinations as shown in Table 3.1. The table indicates how the individual effect and interactions are calculated. It was assumed that A,B and C are the fixed factors where there are 'a' levels of A, 'b' levels of B and 'c' levels of C arranged in the factorial experiment. Generally there will be abc…..n total observations if there are n replicates of the complete experiment. The analysis variance is shown in Table 3.2.


Table 3.1. Design matrix for a 23 Factorial Design

Consider a three factors experiment, with underlying model as shown in Equation1, before the model equation can be fitted, it is important to conduct some statistical tests such as Gtest, T-test and F-test, which involves calculation of these statistical parameters with the aid of certain formulae shown in Equations 2-4 and compare them with those given in the statistical tables. G-test is used to check if the output has the maximum accuracy of replication. T-test is used to check the significance of regression coefficient, and F-test is used to test for the adequacy of the model. Equations 2-4 represent the formulae to calculate G-test, T-test and F-test respectively.

Extraction and Optimization of Oil from

Su2 = sum of dispersion

 N = number of runs =8 Y = experimental yield

S2ad = the dispersion of adequacy

 bj = coefficient of equation variable λ = insignificant coefficient = 2

Yr = response yield of a replicate

**3.3 Production of bio-ethanol from rice husk** 

**3.3.1 Fermentation of hydrolysed rice husk** 

r = number of replicates for a particular run = 2

Yi = average response yield of the replicate for a run

Where

Moringa Oleifera Seed as an Alternative Feedstock for the Production of Biodiesel 253

Prior to the production of bio-ethanol, the rice was treated to confirm the presence of starch. Paddy rice was milled sieved and the residue was collected and weighed. 2cm3 of sample was measured from the bulk sample and transferred into the test tube. Potassium iodide reagent was then added drop wise into the sample in the test tube and stirred until colour was changed from yellow to black, which confirm the presence of starch. 500g of the husk was collected and soaked in 750cm3 of water for a period of 24 hours after which it was filtered with the aid of a filter cloth, 600cm3 of the filtrate was collected and made up to 1000cm3 with boiled water, the mixture was stirred continuously to avoid formation of lumps, it was then allowed to cool and on cooling, a thick-jelly mass was formed, gelatinized mixture was then poured into a 2000cm3 flask for hydrolysis. 200cm3 of 0.5m potassium hydroxide was added to the sample and immersed in the water bath for hydrolysis and the temperature was maintained at 75oC for 60 minutes. 100cm3 of 50% ethanoic acid was then added to serve as a terminator of the hydrolysis reaction after which the mixture was set aside to cool. 4cm3 of hydrolyzed sample, , few drops of Fehling's solution was added in a conical flask and heated, colour change was observed and recorded, sample changes to brick red precipitate, which confirm the presence of simple sugars.

*Zymomonas mobilis* "Local strain" was isolated from palm wine using standard solid medium. Media constituents include 5.0g of yeast extract, 20g of agar and 1000cm3 of distilled water with pH 6.8. Medium was treated with actidione (cycloheximide) to inhibit *Zymomonas mobilis* growth before autoclaving at 121oC for 15 minutes. *Zymomonas mobilis* was then inoculated into the medium and incubated an aerobically at 3oC for 24 hours. Working close to the flame (creating aseptic environment), *Zymomonas mobilis* was introduced into the conical flask containing the substrate, the flask were then shaken (agitation process) and the mouths of the conical flasks were flamed before corking back and incubating at room temperature, they were shaken at various intervals in order to produce a homogenous paste and even distribution of the organisms in the substrates. After fermentation process, the substrates were then filtered using filter cloth and collected in a conical flask, in order to separate the desired product (the filtrate) from the residue. The filtrates were then distilled at 78.3oC using alcohol distillation apparatus, round bottom flask containing the filtrate was placed in the heating mantle and the mouth fixed to the condenser, a beaker for distillate collection was placed at the end of the set up, rubber pipes

Ycal = response yield calculated using the appropriate model equation


Table 3.2. Variance (ANOVA) analysis

$$Y\_{\rm ijk} = \mu + \tau\_i + \langle \mathfrak{k} \rangle + \chi\_k + (\mathfrak{r}\mathfrak{k})\_{\vec{\eta}} + (\tau\chi)\_{\vec{\mathsf{k}}} + (\langle \mathfrak{k}\rangle)\_{\vec{\mathsf{k}}} + (\mathfrak{r}\langle \mathfrak{b}\upsilon\rangle\_{\vec{\mathsf{i}}\vec{\mathsf{k}}} + \mathcal{E}\_{\vec{\mathsf{i}}\mathsf{k}} \tag{1}$$

i = 1,2,---a

j = 1,2,---b

k =1,2,---c l = 1,2,---n

Where µ is the overall mean effect,

τi is the effect of the ith level of factor A

βj is the effect of jth level of factor B

γk is the effect of kth level of factor C

(τβ)ij is the effect of the interaction between A and C

(βγ)ik is the effect of the interaction between B and C

(τβv)ijk is the effect of the interaction between A, B and C

Eijkl is the random error component having a normal distribution with zero and variance δ<sup>2</sup>

$$\mathbf{G}\_{\text{-cal}} = \frac{\text{Su}\_{\text{max}}^2}{\Sigma \,\text{Su}^2} \tag{2}$$

$$\mathbf{T\_{ccal}} = \frac{|\mathbf{b\_{l}}|}{\mathbf{S\_{b}}} \tag{3}$$

$$\mathbf{Sb} = \frac{\sqrt{\text{(Su^2)} }}{\sqrt{\text{(N.r)}}}$$

$$F\_{-cal} = \frac{S\_d^2 d}{Su^2} \tag{4}$$

$$\mathbf{S}\_{\mathbf{a}}^{2}\mathbf{d} = \frac{\mathbf{r}}{\mathbf{N} - \lambda} \boldsymbol{\Sigma} (\mathbf{Y} - \mathbf{Y}\_{\text{cal}})^{2}$$

$$\mathbf{S}\mathbf{u}^{2} = \frac{1}{\mathbf{r} - \mathbf{1}} \boldsymbol{\Sigma} (\mathbf{Y}\_{\text{r}} - \mathbf{Y}\_{\text{l}})$$

### Where

252 Sustainable Growth and Applications in Renewable Energy Sources

B SSB (b-1) MSB δ2 +(can Σ β2 j)/(b-1) MSB/MSE

BC SSBC (b-1)(c-1) MSBC δ2+(anΣΣ(βγ)2jk)/(b-1)(c-1) MSBC/MSE

Yijkl = µ + τi + βj + γk + (τβ)ij + (τγ)ik + (βγ)jk + (τβv)ijk + Eijk (1)

Eijkl is the random error component having a normal distribution with zero and variance δ<sup>2</sup>

G���� � ����� �

T���� � ����

S� � ������ ������ 

����� � ��

 S� �� � �

Su� � �

��

��� ∑�Y�Y�����

��� ∑�Y� � Y��

Square Expected Mean Squares Fo

δ2+(nΣΣΣ(τβγ)2ijk)/(a-1)

i )/ (a-1) MSA/MSE

k)/(c-1) MSC/MSE

ij)/(a-1)(b-1) MSAB/MSE

ik)/(a-1)(c-1) MSAC/MSE

(b-1)(c-1) MSABC/MSE

<sup>∑</sup> ��� (2)

�� (3)

��� (4)

Mean

Sources of Variation

i = 1,2,---a j = 1,2,---b k =1,2,---c l = 1,2,---n

Sum of Squares

Total SST

Table 3.2. Variance (ANOVA) analysis

Where µ is the overall mean effect,

τi is the effect of the ith level of factor A βj is the effect of jth level of factor B γk is the effect of kth level of factor C

(τβ)ij is the effect of the interaction between A and C (βγ)ik is the effect of the interaction between B and C (τβv)ijk is the effect of the interaction between A, B and C

Degree of Freedom

A SSA (a-1) MSA δ2 +(bcn Σ τ<sup>2</sup>

C SSC (c-1) MSC δ2 +(abn Σ γ<sup>2</sup>

AB SSAB (a-1)(b-1) MSAB δ2+(cnΣΣ(τβ)2

AC SSAC (a-1)(c-1) MSAC δ2+(bnΣΣ(τγ)2

Error SSE abc(n-1) MSE δ2

ABC SSABC (a-1)(b-1)(c-1) MSABC

 S2 ad = the dispersion of adequacy

Su2 = sum of dispersion

bj = coefficient of equation variable

λ = insignificant coefficient = 2

r = number of replicates for a particular run = 2

N = number of runs =8

Y = experimental yield

Ycal = response yield calculated using the appropriate model equation

Yr = response yield of a replicate

Yi = average response yield of the replicate for a run
