i
OPTIMIZATION OF BIODIESEL PRODUCTION FROM JATROPHA OIL
USING IMMOBILIZED LIPASE
ALIREZA ZAREI
A report submitted in partial fulfillment of the
requirements for the award of the degree of
Master of Engineering (Chemical)
Faculty of Chemical Engineering
University Teknology Malaysia
APRIL 2012
iii
To my beloved family
iv
ACKNOWLEDGEMENT
Praises to Allah for giving me the strength, perseverance and intention to go
through and complete my study.
In preparing this thesis, I was in contact with many people, researchers,
academicians, and practitioners. They have contributed towards my understanding
and thoughts. In particular, I wish to express my sincere appreciation to my main
thesis supervisor, Professor Dr. Nor Aishah Saidina Amin, for encouragement,
guidance, critics and friendship. I am also very thankful to my co-supervisors Dr.
Nor Azimah Mohd Zain for her guidance, advices and motivation. Without their
continued support and interest, this thesis would not have been the same as presented
here.
v
ABSTRACT
The objective of this study was to produce biodiesel through an enzymatic
transesterification process. In this process, jatropha oil was used as the feedstock to
react with methanol in the presence of Rhizopus oryzae lipase catalyst. Rhizopus
oryzae lipase was immobilized with an innovated method to avoid inhibition and
deactivation effects of glycerol and alcohol. The Lipase immobilization, in turn,
made the process of separation much easier in comparison with utilizing free lipase.
The effect of four important variables (time, temperature, alcohol/oil molar ratio, and
oil moisture content) on the yield of biodiesel was studied. The biodiesel production
process was modeled and optimized using Artificial Neural Network (ANN),
Response Surface Methodology (RSM), and Genetic Algorithm (GA). Maximum
yield percentage of 87.07 and 86.62% was predicted by ANN and RSM,
respectively. The experimental amount of 87.1% was achieved at optimum
parameters: reaction time of 17 h, temperature of 40 °C, 70 % (wt %) water content
and the alcohol/oil molar ratio of 5 %. The physico-chemical properties of the
jatropha oil and obtained biodiesel were investigated and compared to ASTM D6751
standard. It was observed that the quality of the produced biodiesel is in an
acceptable range. The results of the present study show that the immobilization
method is reliable and can enhance the performance of the lipase as the catalyst for
biodiesel production with moderate conditions.
vi
ABSTRAK
Dalam kajian ini, kaedah yang digunakan ialah kaedah transesterifikasi dan
bahan mentah yang digunakan ialah minyak jatropha. Minyak jatropha bukan minyak
masak dan ini dapat mengurangkan masalah yang timbul dalam perindustrian
pembuatan biodisel. Biomangkin digunakan untuk mengurangkan masalah yang
timbul akibat penggunaan mangkin asid dan alkali, serta membuatkan proses ini
lebih mesra alam. Lipase rhizopus oryzae telah digunakan sebagai mangkin; lipase
ini dinyahaktifkan untuk membuatkannya lebih efisien dan membolehkan proses
pengasingan mangkin menjadi lebih mudah. Kualiti hasil turut meningkat disebabkan
penyahaktifan ini. Dalam kaedah transesterifikasi, alcohol digunakan untuk proses
pengalkoholan dan methanol telah digunakan. Empat pembolehubah telah dipilih
untuk kajian ini, iaitu masa, suhu, nisbah mol alcohol kepada minyak dan kandungan
air. Proses penghasilan biodisel sangat memakan masa dan memerlukan kos yang
tinggi. Oleh itu, pencarian keadaan yang optimum untuk proses ini dijalankan amat
penting untuk mengurangkan kos serta masa. Metodologi permukaan bertindakbalas
serta rangkaian neural tiruan digunakan untuk membuat simulasi proses pembuatan
biodisel. Algoritma genetik turut digunakan untuk mengenalpasti keadaan yang
optimum. Dalam kajian ini, hasil proses yang maksimum dapat dicapai sebanyak
87.10% pada keadaan optimum 40°C, 17 jam, 5% nisbah metanol/minyak dan 70 (wt
%) kandungan air. Keputusan daripada ANN dan RSM masing masing adalah
86.62% dan 87.07%.
vii
TABLES OF CONTENTS
CHAPTER
CONTENT
TITLE
PAGE
I
DECLERATION
II
DEDICATION
III
ACKNOWLEDGMENT
IV
ABSTRACT
V
ABSTRAK
VI
TABLES OF CONTENTS
VII
LIST OF TABLES
XI
LIST OF FIGURES
XII
LIST OF ABBREVIATIONS
1
XIV
INTRODUCTION
1
1.1
Background of Research
1
1.2
Biodiesel feedstock
3
1.3
Advantages of biodiesel
5
1.4
Biodiesel production processes
6
1.5 Reaction
8
1.5.1 Transesterification process
8
1.6
Problem statement
10
1.7
Research objectives
12
viii
1.8 Scope of research
13
1
2
LITERATURE REVIEW
15
2.1
Introduction
15
2.2
Starting oils for biodiesel production
16
2.3
Jatropha oil as the proper source for biodiesel production
19
2.4 Transesterification of jatropha oil
21
2.5
22
Different catalysts for transesterification
2.5.1
Alcali catalyst
22
2.5.2
Acid catalyst
25
2.5.3
Biocatalyst
25
2.5.3.1
Extracellular lipase
28
2.5.3.2
Effective methanolysis using extracellular lipase
30
2.5.3.3 Intracellular lipase
2.6
36
2.7
Non-catalystic transesterification
36
2.8
Biodiesel purification and separation
39
2.8.1
Conventional techniques for biodiesel separation
40
2.8.2
Conventional techniques for biodiesel purification
41
2.8
3
Homogenous and hetrogeneous catalysts
32
Optimization
41
METHODOLOGY
43
3.1
Research methodology approach
43
3.2
Materials
44
3.3
Experimental
46
ix
3.3.1
Preparation of catalyst
46
3.3.2
Determination of physical stability
47
3.3.3
Determination of chemical stability
47
3.3.4
Catalyst characterization
47
3.3.5
Mechanical stability test
48
3.3.6
Enzyme assay
49
3.3.6.1 Lipase activity assay
3.3.7
3.4
4
Biodiesel production
49
49
3.3.8 Sample analysis
50
3.3.9 Biodiesel physico-chemical properties
51
3.3.9.1 Density
51
3.3.9.2 Kinematic viscosity
52
3.3.9.3 Water content
53
3.3.9.4 Flash point
55
3.3.9.5 Pour point
56
3.3.9.6 Cloud point
58
3.3.9.7 Acid value
59
Experimental result and process optimization
60
3.4.1 Response surface methodology
60
3.4.2 Artificial neural network
63
3.4.3 Genetic algorithm
65
RESULS AND DISCUSSION
66
4.1
Innovation in immobilization
66
4.2
Surface morphologies
67
4.3 Lipase activityt assay
68
x
4.4 PVA-Alginate beads stability
69
4.4.1 Chemical stability test
69
4.4.2 Mechanical stability test
70
4.5 Sample analysis
5
71
4.5.1 Kinematic viscosity
71
4.5.2 Density
72
4.5.3 Acid Value
72
4.5.4 Water content
73
4.5.5 Flash point
73
4.5.6 Pour and cloud points
74
4.6
Experimental results
55
4.7
Regretion model and statistical analysis
77
4.8
Influence of reaction temperature and time
84
4.9 Methanol/oil ratio and water content effects on FAME yield
86
4.10 Prediction with ANN
87
4.11 Process optimization
88
CONCLUSION AND RECOMMENDATION
89
5.1
Conclusion
89
5.2
Recommendation
90
REFRENCES
92
APPENDIX A
104
APPENDIX B
105
APPENDIX C
106
xi
LIST OF TABLES
TABLE NO.
TITLE
PAGE
2.1
Oil yield (l/ha) from oleaginous species and microalgae
17
2.2
Fatty acid profile of some vegetable oils
18
2.3
Biodiesel specificities for vehicle use
19
2.4
The potential advantages and disadvantages of jatropha plant
20
‎2.5
Composition of crude jatropha oil
21
‎2.6
Comparison of enzymatic catalyst versus alkaline catalyst
26
2‎ .7
Impurities effect on biodiesel and engines
40
‎3.1
The physicochemical properties of crude Jatropha oil
45
3.2
Experimental design of biodiesel production from jatropha oil
61
4.1
Biodiesel properties
74
4.2
Experimental design and experimental results of the response
76
4.3
Analysis of variance (ANOVA) for model regression
78
4.4
ANOVA for Response Surface Reduced Quadratic Model
79
4.5
Weights and biases of BP network
87
4.6
Optimization using GA with RSM and ANN
88
xii
LIST OF FIGURES
FIGURE NO
TITLE
PAGE
1.1
Biodiesel production sequence by transesterification
2
1.2
The various uses of jatropha curcas components
5
1.3
Transesterification of triglycerides with alcohols
8
1.4
Three consecutive reversible reactions during transesterification
9
2.1
Comparison of different biodiesels efficiency
2.2
Biodiesel production by alkali catalyst- separating and implying
22
washing method
23
2.3
Enzymatic transesterification method
27
2.4
Comparison of intracellular and extracellular enzymes
33
3.1
Different steps of research methodology
44
3.2
PVA-alginate beads preparation set up
46
3.3
The equipments for finding density
51
3.4
The Kinematic viscometer equipment
53
3.5
Water content determination equipment
54
3.6
Open cup method for flash point determination
56
3.7
The Pour and cloud point determination equipment
59
3.8
BP network model
64
4.1
Inner surface of the bead
67
4.2
Outer layer of a bead
69
4.3
Normal % probability and studentized residual plot
81
4.4
The studentized residuals and predicted response plot
82
xiii
4.5
The actual and predicted plot
83
4.6
The Outlier t plot
84
4.7
The effect of reaction temperature and reaction time
85
4.8
The effect of methanol/oil ratio and water content
86
xiv
LIST OF ABBREVIATIONS
ANOVA
Analysis of Variance
ANN
Artificial Neural Network
ASTM
American Society of Testing and Materials
BDF
Biodiesel Fuel
CCD
Central Composite Design
DG
Diglycerides
E
Ester
FAME
Fatty Acid Methyl Ester
FESEM
Field Emission Scanning Electron Microscopy
FFA
Free Fatty Acid
GL
Glycerol
GC
Gas Chromatograph
GC-MS
Gas Chromatograph-Mass Spectroscopy
ME
Methyl Ester
MeOH
Methanol
MG
Mono Glyceride
ROL
Rhizopus Oryzae Lipase
RSM
Response Surface Methodology
TG
Triglyceride
WCO
Waste Cooking Oil
1
CHAPTRE 1
INTRODUCTION
1.1
Background of research
There are variety of reasons which are encouraged the researchers to find the
renewable and alternative sources of energy. As the most important reasons, it can be
mentioned to environmental concerns, fossil fuel depletion, increasing the price of
petroleum and rising energy demand (Demirbas, 2009).
Fatty acid methyl ester (FAME), known as biodiesel, has attracted vast
attention in recent years and is potentially one of the main sources of fuel in the near
future. Biofuel and biomass-based energy have potential to become major
contributors of energy in the next century. Currently about 90% of the biofuel market
is captured by bioethanol and biodiesel. Biodiesel is a fuel derived from renewable
resources, such as vegetable oils (Sharma et al. 2008; Ma and Hanna 1999).
However, the industrial scale production of biodiesel is limited due to the unwanted
byproducts, restoration of glycerol, withdrawal of inorganic water and salts,
wastewater treatment and high energy requirement.
2
Biodiesel production developed by Rudolph Diesel in the 1890s. He worked on
vegetable oil fuels and shows that pure vegetable oils could be used as the fuel for
early diesel engines in agriculture and some areas of world where at that time they
did not have access to the petroleum. The kind of modern fuel that is known as
biodiesel is produced by conversion of oils such as vegetable oils into FAMEs.
The focus of industrial production is on reconstructing vegetable oils into a
mixture of fatty acid esters by a process commonly described as being akin to the
cracking of petroleum which is triglycerides transesterification with low molecular
weight alcohols. This reaction is triglyseric esters alcoholysis which produces a
mixture with higher volatility (biodiesel) that its physical properties are much more
alike to those of conventional diesel fluids. The sequence of transesterification
process is illustrated in Figure 1.1 (Srivathsan et al., 2008).
Figur 1.1. Biodiesel production sequence by transesterification.
Biodiesel is not , as such , a biotechnological creation being manufactured with
any appropriate vegetable oil from harvests with no history of plant biotechnology
(or even from fat of animals ) by an totally chemical procedure but interpreters
3
include biodiesel in the portfolio of materialized bio fuels because of its biological
creation as a plant seed oil. In 2005, the evaluated biodiesel world production was
2.91 million tons of oil equivalents, of which 87% was made in European Union
(62% in Germany), with only the United States (7.5%) and Brazil (1.7%) as other
major supplier; this total manufacture amounted to less than 20% of that of world
fuel ethanol production. On the other hand, world biodiesel supply grew by threefold
between 2000 and 2005 and grew from 3.2 to about 4.9 million tons from 2005 to
2006 and it is growing continuously and a marked development in the Europe as well
as in United State is supposed by the International Energy Agency up to 2030.
1.2
Biodiesel feedstock
Many sources exist as the feedstock of biodiesel production, such as animal
fats and vegetable oils. Vegetable oils are more widely used than other sources and
are divided into two categories: (1) edible oils, such as soybean, palm, olive,
sunflower, rapeseed, and corn oils and (2) inedible oils like waste cooking oil, algae,
greases (float grease and trap grease), and some bean oils, such as Jatropha curcas
(Jatropha), Madhua indica, Pongamina pinnata (Karanja), Hevea brasiliensis
(Rubber), and Calophyllum inophyllum (Polanga) (Pinzi et al., 2009).
The use of edible oils in biodiesel production has been questioned because it
can lead to the reduction of one of the main sources of human food worldwide. This
issue will increase the price of edible oils and consequently, the price of biodiesel.
The price of edible oil may increase because of the competition between the
biodiesel market and human consumption. (Jain and Sharma, 2010). The raw
materials price in production of biodiesel is around 60–80%.
4
It is predicted that environmental issues may possibly arise as the mass
propagation of plants producing edible oil may lead to deforestation (Leung et al.,
2010).
For overcoming these disadvantages, non-edible oils have attracted the
researchers. Between non-edible bean oils, Jatropha is more beneficial and efficient
for production of biodiesel in case of sociological, environmental implications and
economical value (Juan et al., 2010). Jatropha curcas is utilized for many years for
hedges and medicines and also for protecting gardens and fields because the animals
do not eat it (Mampane et al., 1987; Gubitz et al., 1999; Staubamann, et al., 1999;
Joubert et al., 1984). The root, bark and leaves also can be used for many other
pharmaceutical and industrial uses as depicted in Figure 1.3
5
Seed oil
-Biodiesel production
-Soap production
-Medicinal uses
-Cooking and lightening
Fruits coats
-Medicinal
-Fertilizer
Fruits
-Fertilizer
Seed Cake
-Organic
fertilizer
\
-Biogas production
-Fodder
(after detoxification or
lower toxic accessions)
Seed shells
-Combustibles
-Organic fertilizer
Seeds
-Insecticide
-Medicinal uses
Leaves
-Medicinal uses
-Fertilizer
Source of dark
blue dye
Latex
-Biocide uses
-Medicinal uses
Roots Oil
-Antihelmintic
properties
Jatropha
Figure 1.2. The various uses of jatropha components (adapted from Jones and
Miller, 1991)
1.3
Advantages of biodiesel
In comparison with conventional diesel, biodiesel has several advantages.
These advantages include:
(1) CO2 and other pollutants emission which are
generated from the engines, can be reduced with help of biodiesel, (2) It is not
6
required to modify the engines because its properties are similar to common fuels
(diesel fuel) , (3) In biodiesel with high purity, it is not needed to use lubricant, (4)
Its high cetane number cause better performance in diesel engine, (5) Renewable
sources are the origins of biodiesel and people can grow their own fuel, (6) It is more
efficient to produce biodiesel in comparison with fossil fuels, because there will be
no refinery, drilling and underwater plantation, (7) Biodiesel can be produced
locally, thus it can make an area become independent of the amount of energy that is
needed (Angina and Ram, 2000; Gerpen, 2005; Jain and Sharma, 2010a; RoblesMedina et al., 2009 ; Jayed et al., 2009).
1.4
Biodiesel production processes
It is not possible to use vegetable oils directly as a biodiesel except that mix it
with diesel fuels in an acceptable ratio and this mixture of ester is stable for the using
in short term. Process of mixing is not complicated and it involves blending alone
and so the price of equipment is low. Due to high viscosity, using these triglyceric
esters (oils) directly is impractical and unsatisfactory for using them for the long time
in the engines which are working with diesel, FFA formation and acid contamination
resulting in formation of gum by carbon deposition and polymerization and
oxidation. Therefore oil of vegetables is processed somehow to get properties
(volatility and viscosity) similar to fossil fuels properties and the fuel which is
processed is possible to be used directly in the diesel engines.
To convert vegetable oil to the fuel three processing techniques are used and
they are micro emulsification, transesterification and pyrolysis. Pyrolysis means
change in chemical that caused by using heat to get the more uncomplicated
compounds from the complex compound. This process is known as cracking. Oil of
7
vegetables can be cracked for improving cetane number and also decreasing in
viscosity. The cracking products include carboxylic acids, alkenes and alkanes.
Rapeseed, cottonseed, soyabean and other kind of oils are cracked successfully with
suitable catalysts for biodiesel production (Ranganathan et al., 2008). Cracking
ended to good flow characteristics that achieved because of viscosity minimization.
Drawbacks of this technique are high cost of equipment and also separate
equipments are needed for various fractions separation.
Also the product that
produced was like gasoline and they contained sulfur and it cause to make the
product less environmental-friendly (Ma and Hanna, 1999).
Next technique for biodiesel production is micro emulsification and it has
been suggested for biodiesel production and micro emulsion biodiesel components
include vegetable oil, diesel fuel, alcohol, cetane and surfactant improver in
appropriate proportions (Ma and Hanna, 1999). Some alcohols like ethanol, propanol
and methanol are utilized as the additives for reducing the viscosity, as the
surfactants, higher alcohols are implied and as the cetane improvers alkyl nitrates are
added. Reducing viscosity, good spray and cetane number increasing characters
promote to use the micro emulsions but on the other hand prolong usage causes some
problems such as carbon deposit formation, partial combustion and injector needle
sticking (Ma and Hanna, 1999).
Another technique is transesterification which is the most common technique
in biodiesel manufacturing. Biodiesel that is produced by transesterification is a
liquid mixture of mono-alkyl esters of higher fatty acids. Removing component with
high viscosity, glycerol, cause to reduction of product viscosity and make it similar
to fossil fuels.
8
1.5
Reaction
1.5.1 Transesterification Process
Biodiesel fuel which is produced by alcoholysising or transesterification
reaction is the substitution of ester alcohol by another alcohol in a technique much
the same to hydrolysis, except that instead of water alcohol is used.
Transesterification of triglycerides with alcohols is illustrated in Figure 1.4. As can
be seen, at least 3 mol methanol is needed to react with 1 mol triglyceride for
producing 3 mol methyl ester as the main product and 1 mol glycerol which is the
byproduct (Palligarnai and Vasudevan, 2010).
Figure 1.3. Transesterification of triglycerides with alcohols
Transesterification consists of a succession of three sequential reversible
reactions. In the first step triglycerides should be converted to diglycerides, followed
9
by the conversion of diglycerides to monoglycerides and at the end monoglycerides
to glycerol, resulting in for each glyceride, one molecule of ester at each step. The
reactions are totally reversible, although the equilibrium lies in the direction of the
production of glycerol and fatty acids esters. These three steps are shown in Figure
1.5 (Palligarnai and Vasudevan, 2010).
Figure 1.4. Three consecutive reversible reactions during transesterification.
The alcohols that are suitable for transesterification are methanol, butanol,
propanol, amyl alcohol and ethanol. Ethanol and methanol are the most ordinary
which are used, notably methanol because of its physical and chemical advantages
and its low cost. This method has been widely used to decrease the viscosity of
10
triglycerides, thereby improving the physical properties of renewable diesel to
enhance engine performance. Thus, FAMEs, known as biodiesel fuel (BDF), acquire
by transesterification can be used as the alternative fuel for engines that are working
by diesel.
1.6
Problem statement
An important obstacle for commercializing biodiesel is that the feedstocks for
biodiesel production is substantially expensive than the cost of petroleum diesel.
Production of biodiesel will become economically visible only if there was reduction
in the material cost. In this case, many researchers such as Jacobson et al. (2007) and
Chin et al. (2009) employed some oils such as waste cooking oils which their price is
low, but still many problems exist to utilize this kind of oil, such as difficulty in oil
collection. Also some times the quality of the waste cooking oil is not good enough
and some pretreatment are required which this ,in turn, will increase the costs. In this
research, among different sources of biodiesel, jatropha oil, an inedible oil, is chosen
because of its low price and also pretreatment is not required for this kind of oil.
Jatropha oil transesterification by using alkali catalyst as well as acidic
catalysts has caused many problems in biodiesel production. These are because of
soap formation that easily takes place when water is produced in this reaction and
also some problems such as equipment corrosion. Saponification not only lowers
down the biodiesel purity also can deactivate the catalyst. Apart from that removal
of these kinds of catalysts after reaction, is technically difficult and the large amount
of water is needed to separate and clean the catalyst from the products. In current
work, lipase catalyst is used to ignore acid and alkali catalysts disadvantages. In
11
addition by using biocatalyst the moderate conditions will be needed during the
reaction and it helps to decreasing the process cost.
The problem with biocatalyst is inhibitor and deactivation effects of enzyme
which is caused due to glycerol existence and alcohol utilization during the process.
To overcome these problems and also to enhance the catalyst performance, the
immobilization method is suggested.
In this study, PVA was used as the entrapment material since it has been shown
to possess desirable characteristics. Due to PVA’s excellent properties, PVA would
serves as an excellent material to be applied in the immobilization of various
microorganisms and biocatalysts. In this particular research, PVA will be used as an
immobilization matrix in bead form to immobilize lipase. There are different ways of
PVA-alginate preparation as immobilization matrix reported by previous studies
(Dalla-Vecchia et al., 2005; Imai et al., 1985; Szczesna-Antczak and Galas, 2001),
but such a technique cannot be used directly in this study due to the solubility of the
resulting matrix in our solution.
To overcome this problem, sodium sulfate was introduced to decrease the
matrix solubility after PVA-alginate beads were treated with saturated boric acid in
this research. Pillay et al. (2005) used sodium sulfate coupled with boric acid as
cross linking agent to crosslink PVA. To date, in literatures, the use of sodium sulfate
has never been reported as cross-linking agent for such an immobilization matrix to
immobilize Rhizopus oryzae lipase.
Process optimization study is important and helpful for the industrialization
and development of biodiesel production (Lee et al., 2011). The traditional technique
of optimization studies in biodiesel manufacturing is one variable at a time, which is
12
based on the variation of one component at a time, with the responses serving as a
function of a single parameter (Bezzera et al., 2008). This method, however, is
tedious and time-consuming and could not depict the total effect of the related
parameters on the procedure (Bas et al., 2007).
Some researchers such as Jitputti et al. (2006) could not find the optimum
variables for maximizing biodiesel production. The reason is that they were only
conducted the experiments ordinary based on the range of variables and plotted the
graphs by Microsoft Excel. So the result only can show the good values and effect of
variables in the reaction. Therefore, some methods such as response surface
methodology (RSM), artificial neural network (ANN) and genetic algorithm (GA)
could be employed in order to design and model the experimental work and also to
determine the optimum condition of biodiesel production.
1.7
Research Objectives
i.
To immobilize Rhizopus oryzae lipase into PVA-alginate beads
ii.
To produce biodiesel from jatropha oil and study the effect of different
variables (reaction time and temperature, alcohol/oil molar ratio and water
content of the oil ) on the yield of biodiesel
iii.
To determine optimized range of variables on biodiesel production by
utilizing RSM, GA and ANN.
iv.
To determine the properties of the obtained biodiesel
13
1.8 Scope of research
The scopes of this research involve the following steps:
i.
Preparation and immobilization of Rhizopus oryzae lipase into the
PVA-alginate bead
ii.
Test the catalyst activity and characterize the immobilized catalyst
iii.
Determination of the jatropha oil properties
iv.
Design of experiments based on the range of variables by design expert
software
v.
Production of biodiesel through the batch process and investigate the
effect of different variables on the biodiesel yield
vi.
Modeling the process and finding the optimum conditions for our
process using RSM, ANN and GA
vii.
Analysis of produced biodiesel to find different physico-chemical
properties of biodiesel
14
15
CHAPTER 2
LITERATURE REVIEW
2.1
Introduction
An important problem confronting biodiesel formation today is the presence of
water and free fatty acids in most starter oils.
These molecules result in the
formation of soap during the transesterification of the triglyceride, which decreases
the yield and requires refinement to remove the soap. This is seen especially in large
scale operations where an alkali catalyst is utilized. While it is possible to remove
these molecules from the oils themselves, it is an expensive process and one which
outweighs the benefits of using biodiesel. Along with the choice of a feedstock,
research is also focusing on the selection of a catalyst and the possible use of
acceptor molecules (Sheedlo, 2008).
The petroleum fuels continuing dependency is the issue which is widely
considered environmentally unsustainable. On the other hand, biodiesel has been
accepted generally as a suitable replacement for usual fossil fuels because it is
nontoxic, renewable and biodegradable.
While the chemical approaches for
biodiesel synthesis needs the use of either alkali or acid catalyst, enzymatic
transesterification method by using methanol (MeOH) has interested much notices,
16
because it is more environmental friendly and also it has the potential of industrial
implementation.
However, the loss of enzyme activity which is caused by glycerol (a byproduct) and alcohol (a co-substrate) and high cost of enzymes still are the obstacles
to biocatalystic method for industrial biodiesel production (Bajaj et al., 2010).
Different sources of lipase have been studied in the production of biodiesel, but
lipase B from Candida antarctica (CALB) is the most commonly that is used, which
by Novozymes has been commercialized as an immobilized form and is known as
the Novoayme 435. It is believed that insoluble MeOH can be partly responsible for
inhibition of enzyme, and using tert-butanol as a solvent has been recognized as a
solution for inhibitory effect reduction, mostly because of its ability in solubilizing
MeOH (Ranganathan et al., 2008). But the most efficient way for utilizing enzyme
catalyst is immobilization method which avoids the enzyme deactivation properly.
2.2
Starting oils for production of biodiesel
The main feedstocks which are used for biodiesel production are the oil of
vegetables that are extracted from oleaginous plants. These materials cost, represent
about 70% of the total cost of production, so it can be concluded that the most
appropriate vegetable oils are those from crops that have the highest output per
hectare or oils with low price such as waste cooking oil and some crude oils. In Table
2 oil yield of different species are compared. For biodiesel production it is not
necessary to purify crude oil perfectly. Thus, the several steps that should be
involved in the process of edible oils refining (for instance, bleaching, deodorization,
etc) can be avoided (Robles-Medina et al., 2009).
17
Table 2.1 Oil yield (l/ha) from oleaginous species and microalgae
Vegetable
Oil yield (l/ha)
Palm
2400
Jatropha
1300
Rapeseed
1100
Sunflower
690
Soybean
400
Microalgae
18750
These days, the food collapses for initiatives of biodiesel, increasing in the
levels of CO emissions in atmosphere and price of high petroleum based sourced fuel
have all created attention to the need for the alternative fuel solutions. As the
sources that have optimistically emerged are inedible oils and they are concerned as
the one of the lowest cost row materials for production of biodiesel. One of the
choices as the inedible oil is algae oil, but the cost of production of huge amount of
algae oil can be one of the obstacles in the short term. As the main reason it can be
mentioned to the operational conditions which cause the high grade oil in microalgae
usually provide low growth rates (nitrogen deficiency, low light intensity and low
temperature). Currently, it is rather difficult to get algae biomass with lipid content
of 20% less than €1/kg, and even the new advances in photo-bioreactors (open and
close) put cost of algae oil less than €5/kg.
Oils and fats could be characterized based on their chemical (saponification
index, peroxide index, iodine index, acidity, etc) or physical properties (refractive
index, density, melting point, viscosity, etc); these are the parameters that will affect
the quality of biodiesel. For instance, the iodine index is associated with the grade of
oil unsaturation and generally, biodiesel that is produced from oil with high
unsaturated fatty acids is less viscous and it indicates greater pour points (i.e.
18
temperature at which fuel stops flowing) and cloud point (i.e. the temperature at
which fuel becomes cloudy due to solidification) which they make the biodiesel more
appropriate for cold weather conditions. However, it is also prone to oxidation that
has a lower level of cetane index (related to efficiency of reaction within the engine)
and lower heat of combustion.
Comparatively, the biodiesel which is produced from high proportion oils in
long chain fatty acids (N18C) have a higher combustion heat and cetane index, but in
addition lower pour and cloud points and greater viscosity (Knothe, 2005). By other
words, the oil fatty acid profile has an effect on the quality of biodiesel that is
produced. The profile of fatty acid of some of the feedstocks employed in production
of biodiesel can be seen in Table 2.2.
Table 2.2 Fatty acid profile (%moles) of some vegetable oils used for biodiesel
production
Oil
C16:0
C16:1
C18:0
C18:1
C18:2
Almond
Borage
Corn
Cotton
Jatropha
Olive
Palm
Canola
Soybean
Sunflower
Microalgae
P. tricom.
6.5
12.9
11.7
28.3
16.4
11.8
42.6
3.5
11.4
7.1
15.5
0.5
0.2
1.4
4.3
1.9
0.9
6.2
2.7
4.4
0.9
4.4
4.7
0.3
70.7
19.1
25.2
13.3
37.0
74.1
40.5
64.4
20.8
25.5
1.3
20.0
39.0
60.5
57.5
39.2
8.5
10.1
22.3
53.8
62.4
2.2
1.0
1.5
0.3
17.3
C18:3
C20:0
C20:1
Others
18.7
0.5
0.3
0.2
3.5
0.9
2.0
0.7
0.2
8.2
9.3
0.9
0.2
0.4
0.3
1.9
0.7
0.3
0.3
62.5
Ratio SFA/
UFA
7.9/9.12
17.5/82.5
13.8/86.2
29.2/70.8
22.8/77.2
14.9/85.1
47/51.1
4.4/94.9
16.1/83.9
12.1/87.9
21.2/78.8
Between the variety of vegetable oils that are available, those that content of
oleic acid is higher are the most appropriate because of their better characteristics as
fuels and the greater stability of their alkyl esters (Knothe, 2005). Based on existing
legislation in US and EU, in any case, biodiesel that is produced will need to follow
19
the existing standards: the Standard EN 14214 in EU and ASTM Biodiesel Standard
D6751 in US which are depicted in Table2.3.
Table 2.3 Biodiesel specificities for vehicle use according to American Standard
(ASTM D-6751) and European Standard (EN 14214). (Knothe, 2005)
Property
Kinematic
Viscosity (40 ̊ C)
Density (15 ̊ C)
Ester content
Cetan number
Flash point
Cloud point
Water
Sulphated ash
Sulphur
Copper strip
corrosion
Carbon residue
Acid number
Free glycerol
Total glycerol
Phosphorous
content
Iodin number
Oxidative stability
(110 ̊ C)
Monoacylglycerols
Monoacylglycerols
Triacylglycerols
Distillation temp.
2.3
ASTM D-6751
Limits
Test method
1.9-6.0
D 445
EN 14214
Units
mm2/S
Limits
3.5-5.0
860-900
96.5 min.
51 min.
120.0 min.
Test method
EN ISO 3104
Units
mm2/S
EN ISO 3675/EN
ISO 12185
prEN 14103
EN ISO 5165
ISO/CD 3679
kg/m3
EN ISO 12937
ISO 3987
DIN 51680
EN ISO 2160
mg/kg
mass%
mg/kg
-
IN ISO 10370
prEN 14104
mass%
̊C
47 min.
130.0 min.
Report
0.050 max.
0.020 max.
0.05 max.
NO. 3 max.
D 613
D 93
D 2500
D 2709
D 874
D 5453
D 130
̊C
̊C
Volume %
mass %
mass %
-
0.050 max.
0.80 max.
D 4530
D 664
mass %
mg KOH/g
500 max.
0.02 max.
10.0 max.
No. 1 (3 h
at 50 ̊ C )
0.3 max.
0.5 max.
0.020 max.
0.240 max.
0.001 max.
D 6584
D 6584
D 4951
mass %
mass %
mass %
0.02 max.
0.25 max.
10 max.
prEN 14105
prEN 14105
prEN 14107
mass%
mg
KOH/g
mass%
mass%
mg/kg
120 max.
6 min.
prEN 14111
prEN14112
h
0.8 max.
0.2 max.
0.2 max.
prEN 14105
prEN 14105
prEN 14105
mass%
mass%
mass%
360 max.
D 1160
̊C
Jatropha Curcas oil as the proper source for production of biodiesel
Jatropha curcas belongs to Euphorbiaceous category. Jatropha name comes
from Latin words Jatros that means doctor and trophe which means food because it
has so many medicinal values. The jatropha plants natively are from American
20
tropic, but they usually grow in subtropical and tropical countries, such as South East
Asia, sub Sahara Africa, China and India. Jatropha seeds mature 3 to 4 months after
flowering and when plant grew enough and becomes adult, it will produce seeds for
50 years. About 40% of the seed content is oil.
Because of leaf shedding activity, jatropha plant becomes very adaptable in
harsh environment due to decomposition of the shed leaves that is nutrients source
for plant and it decreases loss of water in the seasons which are dry. So, it is suited
to many different types of soil, such as soils which are not rich in nutrition like stony
soils, sandy and saline, but in waterlogged land it cannot grow. Thus as a plant that is
resistance against drought, one of the best candidates is jatropha and also it is ecorestoration in wastelands. Also cultivation of jatropha in wastelands will assist the
soil to recover and would be able to help in restoration of carbon and sequestration.
The summary of disadvantages and advantages of jatropha production is in Table
2.4. As it is clear, its benefits outweigh the disadvantages, thus it worth to use that as
a source for industrial purposes.
Table 2.4. The potential advantages and disadvantages of jatropha curcas plants.
The advantages of Jatropha plant
Good agronomic traits
1. Hardy shrub which grows in semi-arid conditions and poor soils
2. Can be intercropped with high value crops such a sugar, coconut palm, various fruits and
vegetables, providing protection from grazing livestock and phyto-protection action against pests and
pathogens
3. It is easy to establish and growth relatively quickly
4. Yields around 4 tons of seed per hectare in unkept hedges are achievable
5. Has low nutrient requirements
6. Requires low labor inputs
Multi-purpose plant
1. Protective hedges around fields
2. Reclaims marginal soils
3. Non-edible and therefore does not compete with food supply when used for biodiesel production
4. Is energy crop that produce seeds with high oil yields
The disadvantages of Jatropha
Seeds and leaves are toxic to human beings and animals
Toxicity is based on several components (phorbol esters, curcains, trypsin inhibitors and others) which
make complete detoxification a complicated and difficult process.
Competes with food production for land use
21
2.4
Transesterification of jatropha oil
Transesterification can be considered as the most common method for
production of biodiesel because of its simplicity; therefore this method has been
usually implied for converting vegetable oils to biodiesel. In general, jatropha oil or
vegetable oil is consisted of unsaturated and saturated monocarboxylic acids with
trihydric alcohol glycerides. As is implied in Table 2.5, main groups of fatty acids
are formed jatropha oil, which the most important ones are stearic acid, linoleic acid,
oleic acid and palmitic acid. Those are the major fatty acids which can be
transesterified to alkyl esters by alcohol or other acyl acceptor. (Hanny et. al., 2008):
Table 2.5. Composition of crude jatropha oila
Fatty acid
Myristic
Palmetic
Palmitoleic
Stearic
Oleic
Linoleic
Linolenic
Formula
C14H28O2
C16H32O2
C16H30O2
C18H36O2
C18H34O2
C18H32O2
C18H30O2
Systemic name
Tetradecanoic
Hexadecanoic
cis-9-Hexadecenoic
Octadecanoic
cis-9-Octadecenoic
cis-9,cis-12 Octadecatrienoic
cis-6,cis-9,cis-12Octadecatrienoic
Eicosanoic
Docosanoic
Arachidic
C20H40O2
Behenic
C22H44O2
a
Adapted from Gubitz et al. (1999).
b
xx:y indicates xx carbons in the fatty acid chain with y double bonds.
Structureb
14:0
16:0
16:1
18:0
18:1
18:2
18:3
Wt%
0-0.1
14.1-15.3
0-1.3
3.7-9.8
34.3-45.8
29.0-44.2
0-0.3
20:0
22:0
0-0.3
0-0.2
Different biodiesel efficiencies are compared in the Figure 2.1. As can be
seen from the picture efficiency of biodiesel from jatropha is in second place by 95%
after biomethane.
22
Figure 2.1. Comparison of different biodiesels efficiency
2.5
2.5.1
Different catalysts for transesterification
Alkali Catalyst
Alkali catalysts have been centered in the most of the recent research to
mediate the fatty acids transesterification.
The processes which are used alkali
catalysts have been shown to be faster in comparison with those that used the acid
catalysts and biocatalysts, but determines more reasonable reaction conditions. This
process needs less alcohol for using in the reaction than acid catalysis, but from the
other aspect, when the content of free fatty acids in the oil is high, it produces more
soap commonly than those of acid catalysts. This occurrence is because of the water
and free fatty acids which are neutralized more easily by bases than by acids and
biocatalysts. These reactions then, in turn, consume the catalyst. Because of this
23
formation of soap properties, alkali catalysts in large commercial operations are
inefficient, where the levels of free fatty acids and water are difficult to regulate
(Michael et al, 2008).
Among the all methods that mentioned for biodiesel production, only the
alkali process is industrialized. Other problems occur in the downstream operations
such as unreacted methanol and catalyst separation from biodiesel. The catalyst
removal involves many efforts and biodiesel needs repeated washing for reaching the
needed purity or using other methods of separation. In Figure 2-2 different steps of
washing method are implied.
Figure 2.2. Biodiesel production by alkali catalyst- separating and employing
washing method
24
2.5.2
Acid catalyst
An acid-catalyzed transesterification processor is used for biodiesel
manufacturing because of the ability of acids to decrease the amount of water and
free fatty acid involved in the reaction and in addition the low price. Commercially,
this method is useful, because the mass preparation of biodiesel usually arises in the
presence of a comparatively huge amount of free fatty acids. The acid catalyst usage
is studied to be more efficient than alkali catalysts when the free fatty acids
concentration is high, greater than 1%. These reactions need post reaction clean-up
though, since the acids that are involved produce a great amount of salt throughout
the reaction which can be corrosive. Because of these salts existence, maintenance
of equipment also plays a crucial role during this process. Another disadvantage of
this reaction is that it is slower than alkali catalysts (Michael et al, 2008).
This catalyst achieves very high yield in biodiesel but the reaction is slow so
it needs higher temperature and also more time to reach complete reaction. The
problem with homogenous acid process is the discontinuous operation that causes a
costly separation and also has corrosion of liquid acid. In contrast, heterogeneous
acid catalyst is easily removed from the reaction. It simplifies continuous operation
and purification (Vicente et al, 2004). When free fatty acids concentration is greater
than 1% the acid catalyst usage will be more efficient than basic catalysts. From
literature it is reported that for transesterification of edible oil strong acids are used.
Some usual inorganic acids that are utilized for transesterification are hydrochloric,
sulfuric and phosphoric acid.
In acid catalyzed transesterification when the alcohol and oil are mixed
directly transesterification and separation will be conducted in one step. In this
condition, alcohol will act similar to esterification reagent and as a solvent. Another
technique is two steps catalyzed transesterification process which acid catalyzed and
25
based catalyzed are involved (Wang et al, 2007). In this method, in first step the free
fatty acid reacted with methanol by using acid catalyst. In the second step final
product from step one or un-reacted triglycerides, transesterified with methanol by
base catalyst. Two-step acid catalyzed process shows that the performance and yield
of this type of process is better, but the process requires more equipment and is more
complex than one-step process (Cayli et al, 2007, Zulaikkah et al, 2005).
2.5.3
Biocatalyst
Far greater attention has however, been paid to develope a biotechnical approach
to biodiesel production, employ enzyme catalysts, usually lipases, and utilize their
catalytic ability to carry out transesterification rather than straightforward hydrolyses
of triglycerides to liberate free fatty acids and glycerol. The principle process
advantage of the enzyme based method is the ability to use low to moderate
temperature and atmospheric pressure in the reaction vessel while ensuring little or
no chemical decomposition. The cons and pros of lipases usage as biocatalysts as
opposed to alkaline catalysts for production of biodiesel are summarized in Table
2.6.
26
Table 2.6. Comparison of enzymatic catalyst versus conventional alkaline catalyst
for producing biodiesel.
Key issue
Enzymatic Process
Alkaline process
Presence of free fatty acid
in the starting oil
Free fatty acids are
transformed to biodiesel
Free fatty acids are
transformed to soaps
Water content of the
starting oil
It is not deleterious for
lipase
Impact on the catalyst by
forming soaps. It may
hydrolyze the oil and
ultimately more soaps are
formed
Biodiesel yield
High, usually around 90%
High usually >96%
Glycerol recovery
Easy, high grade glycerol
Complex, low grade glycerol
Catalyst recovery and
reusage
Easy or not necessary when
operating in a PBR
Reusability not sufficiently
studied
Difficult or not profitable,
usually it is neutralized by
adding an acid after
transesterification. It is
partially lost as soaps or in
the successive washing steps
Energy cost
Low, temperature range 2050 ̊ C
Medium, temperature range
60-80 ̊ C
Catalyst cost
High
Low
Environmental impact
Low, waste water treatment
not needed
Medium, alkaline and saline
effluents are generated
Process productivity
Low
High
It is supposed that these kinds of catalysts have the capability to surpass
chemical catalysts. Biocatalysts are inherently occurring lipases that they have been
recognized as having the potential to carry out the transesterification reactions which
are vital to biodiesel preparation. These lipases are isolated from the number of
bacterial species: Rhizopus oryzae, Pseudomonas fluorescens, Pseudomonas cepacia,
Candida Antarctica, Thermomyces anuginosus, Rhizomucor miehei and Candida
rugosa. Most of these lipases have since become commercially available; such as
Novozym 435 which is so popular. Throughout the time of experimentation, it has
also been shown that the overall yields can be increased by adjusting the pH of
solution.
27
It is found that some enzymes like lipase can be applied to catalyze
transesterification and they can be immobilized in an appropriate support. One of the
advantages of enzyme immobilization is that enzyme can be recycled and reused
without separation. Also, in comparison with the other techniques, temperature of
the process is lower. As the drawbacks of the enzyme catalyst can be mentioned to
inhibition effect that is happened due to using of methanol and the other is the price
of enzyme (Shimada et al., 2002).
The biodiesel production with a biocatalyst withdraws the weaknesses of
alkali and acid processes by obtaining the high quality products and with less
downstream operations (Ranganathan et al., 2007). Different steps of enzymatic
production of biodiesel are simply illustrated in Figure 2.4.
Figure 2.3. Enzymatic transesterification method
This method of using a biocatalyst for production of biodiesel was patented
by Haas (1997). But the process has not industrialized yet because of some
28
restrictions such as inhibition of enzyme by methanol, high cost of enzymes and
enzyme activity exhaustion.
Enzymatic production of biodiesel is achievable employing both intracellular
and extracellular lipases. In both cases the enzyme is immobilized and after that is
used which eliminates operation problems, such as recycling and separation. Thus in
all the works that are reported either immobilized whole cells (Intracellular enzymes)
or immobilized (extracellular) enzymes are used as the catalyst. Both processes are
more efficient compared to using free enzymes.
2.5.3.1 Extracellular lipase
It is reported by Mittelbach (1990) that he had done sunflower oil
transesterification with primary alcohols such as ethanol, butanol and methanol using
C. antarctica (Novozym 435) and M. miehei in the absence and presence of the
solvent. The yields that he caught for butanol and ethanol were notable, even without
using the solvent. It is found by him that methanol produces only traces of methyl
esters without the solvent. Batch experiments are conducted by Nelson et al. (1996)
and he has found that C. antarctica was appropriate for secondary alcohols, by 80 %
conversion, such as 2-butanol and iso-propanol and M. miehei was efficient for
primary short chain alcohols, by 95% conversion, like butanol, propanol, ethanol and
methanol with hexane as a solvent. However without using the solvent, the least
efficiency was related to methanol with the 19.4% yield of methyl ester.
The low yield was obtained due to the inhibitory effects that caused by
methanol on the immobilized enzyme. Abigor et al. (2000) again confirmed this
29
result and he reported the palm kernel oil conversion by using ethanol and methanol
as 72% and 15%, respectively. Noureddini et al. (2001) implied ethanol and
methanol for soyabean oil transesterification with using immobilized enzyme that is
produced from Pseudomonas fluorescens and reported conversions 65% and 67% for
ethanol and methanol respectively. But it is demonstrated by Linko et al. (1998) that
he could get the conversion of 97% and he was using 2-ethyl-1-hexanol for the
rapeseed oil transesterification.
In the same way, instead of methanol, using other alcohols as the alternate
acyl acceptors have been experimented and conversion of 90% was gained. 90%
conversion of vegetable oil is reported by Iso et al. (2001) by using P. fluorescens
enzyme and butanol as the acyl acceptor. This reaction was conducted in a solvent
free medium with optimum condition of 60°C and 0.3% water. Modi et al. (2006)
used propane-2-ol as an acyl acceptor for sunflower, Karanj and Jatropha oil
transesterification and his maximum conversion was 93.4%, 91.7% and 92.8%,
respectively. The reusability of lipase was continued for 12 cycles with propane-2-ol
and it falls to zero after seven cycles in condition that methanol was used.
For achieving high conversions with methanol as the acyl acceptor many
efforts have been done by researchers to reduce the methanol inhibitory effects.
Using the usual solvent for oil and methanol was found as the alternative method by
many researchers since inhibition is because of methanol.
High conversion with methanol reported by Iso et al. (2001) by using 1-4
dioxane as a solvent. In the case of acyl acceptor, using tert-butanol was suggested
rather than using butanol, as a solvent for oil methanolysis. Li et al. (2006) used tertbutanol as a solvent for the rapeseed oil transesterification. 95% conversion was
obtained using Novozyme 435 and Liposome TI LM in an appropriate ratio (3:1)
under optimum conditions. Tert- butanol was again used by Royon et al. (2007) as a
30
new and effective solvent for the enzymatic transesterification and he used
Novozyme 435 for the cotton seed oil transesterification and the yield that he
achieved was 97% and the condition was 24 hours at 55 °C. By employing the
continuous fixed bed reactor (FBR) conversion of 95% was reported. Using tertbutanol as a solvent is a viable solution for moderating the methanol inhibitory
effects and also for industrializing the process.
Immobilized enzyme production was considered to reduce the enzyme
deactivation and Samukawa et al. (2000) showed this by enzyme preincubation in
methyl oleate for 30 minute and consequently in soyabean oil for 12 hours.
Inhibitory effects were noticeably decreased and high conversions were achieved.
The usage of 2-butanol and tert-butanol for restoring the activity of enzymes that
were deactivated was suggested by Chen and Wu (2003). They experienced complete
deactivation of Novozyme 435 with methanol and restoring by washing the enzyme
with 2-butanol and tert-butanol. The enzyme activity rose by 10 times in comparison
with the enzymes which was untreated and the totally deactivated enzyme was
regenerated to 56% and 75% of its original value when they washed with 2-butanol
and tert-butanol, respectively.
2.5.3.2 Effective methanolysis utilizing extracellular lipase
Yield of methanolysis was low without employing suitable solvent and even
by utilizing some solvents high conversion obtained by other alcohols is impossible.
As the enzyme deactivation was because of methanol insolubility, any methanol and
oil molar ratio more than 1.5:1 caused an important inhibition. For solving this
problem, stepwise addition of methanol was suggested by Shimada et al. (1999) and
95% conversion was observed through that, even after 50 cycles. Like their work,
31
Watanabe et al. (2000) worked on two step batch wise methanol addition and three
steps uninterrupted of methanol addition. More than 90% conversion reported by
them after 100 repeated operation cycles.
Samukawa et al. (2000) also demonstrated stepwise addition and preincubated the enzyme before usage, were successful in getting a conversion as high
as 97% through three stepwise additions of 0.33 molar equivalents of methanol at
0.25–0.4 hour period.
Kaieda et al. (1999) worked on the difference in using non-regiospecific and
regiospecific lipases. Non-regiospecific lipases such as P. fluorescens, P. cepacia
and Candida rugosa are tolerant to the inhibition of methanol. High conversion was
achieved by P. cepacia even 2-3 molar equivalents of methanol.
By using
regiospecific lipases such as Rhizopus oryzae, 80 to 90% conversion was gained with
stepwise addition of alcohol with 4–30% water. Shimada et al. (2002) could convert
WCO to biodiesel by rate of more than 90% with employing stepwise addition of
methanol. They also used continuous systems and catch high conversions.
So it was shown that methanol inhibitory effects can be reduced by stepwise
addition of methanol and high conversions can be obtained even in a solvent free
system. Bako et al. (2002) believed that the inhibitory effects are mostly because of
glycerol formation in the time of reaction and for reducing these effects; they
suggested using dialysis for situ glycerol removal. So many works were carried out
with 2 and 3 stepwise addition of alcohol and glycerol was removed constantly by
dialysis. Method that mentioned was found for developing the process
productiveness and it was mentioned that at least flow rate of 85 milliliter of glycerol
per liter of reacting mixture was needed for efficient separation with a conversion of
around 97% at 50 °C.
32
Xu et al. (2004) recommended that glycerol extracting was also feasible by
using iso propanol in a process hiring stepwise methanolysis. Thermomyces
lanuginosus (Lipozyme TL IM) was involved catalyze for soyabean oil
transesterification and a maximum conversion of 98% was reported as the maximum
yield at temperature of 40°C and a 94% conversion was obtained even after 15
repeated operation cycles. Nie et al. (2006) involved optimization experiments on
continuous and batch transesterification process and he recounted 96% as the
maximum conversion in the batch process with three stepwise methanolysis and also
utilizing immobilized Novozyme 435. The lipase was found to keep its activity for
continuous operations in more than twenty days. During the continuous process 93%
and 92% conversions were gained for WCO and vegetable oil respectively and this
process was suggested for industrial scale production of biodiesel (Ranganathan et
al., 2007).
2.5.3.3 Intracellular lipase
The process efficiency can grew by employing intracellular lipase or whole
cell immobilization as a substitute of extracellular lipase that before immobilization
needed very complex purification steps. A comparison of the different
immobilization process is illustrated in Figure 2.4.
33
Figure 2.4. Comparison of steps involved in the immobilization of intracellular and
extracellular enzymes: (a) extracellular lipase, (b) intracellular lipase
From Figure 2.4 you can see that cost will reduce with using intracellular
lipase. Matsumoto et al. (2001) developed whole cell biocatalyst by immobilizing
Rhizopus oryzae cells and permeabilizing them by air drying. After that it was
implied for the methyl esters production by three stepwise addition of methanol in
the water containing and solvent free system. It was reported then that the content of
methyl ester in the reaction was around 71 wt% after the 165 hours reaction at the
temperature of 37 °C with stepwise methanolysis. Ban et al. (2001) utilized
immobilized whole cell Rhizopus oryzae for the vegetable oils transesterification and
he is investigated the effect of some parameters such as water content and cell
pretreatment on the process of biodiesel production. For improving the methanolysis
activity of the immobilized cells, different substrates related compounds were added
to the medium, of which oleic acid and olive oil were found to be effective
(Ranganathan et al., 2007).
34
It was reported that, with stepwise methanolising and with 15% muister
content 90% conversion was gained and it was comparative with the extracellular
process. For stabilizing cells of Rhizopus oryzae, Ban et al. (2002) conducted crosslinking treatment with 0.1% glutaraldehyhde. He reported that without treatment of
glutaraldehyhde, in the stepwise methanolising process, conversion level reduced to
50% after sixth batch cycle while with glutaraldehyhde treatment, even after six
batch cycles, conversion can be retain at 72 to 83%. Fukuda and Kondo (2003)
utilized whole cell biocatalyst by pretreatment of the cells with lower alcohols in the
process of biodiesel production. The 350–600 times grow in the rate of reaction is
claimed by employing cells treated with lower alcohols in comparison with untreated
cells.
For extending the efficiency of enzymatic transesterification many
researchers used whole cell immobilization method. Hama et al. (2004) worked on
the stability of enzyme and tried to increase that with the fatty acid composite
variation of the cell membrane. By adding different
fatty acids to the culture
medium, fatty acid composition of the cells membrane could be controlled. It was
mentioned by some researchers that higher enzymatic activity is indicated by linoleic
acid and oleic acid enriched cells than palmitic acid enriched cells and saturated fatty
acid enriched cells demonstrated more stability than unsaturated fatty acid enriched
cells. Consequently an optimum ratio of unsaturated to total fatty acids was
determined as a compensation for both stability and activity and it was found to be
0.67.
The high consistency of methanolysing yield was reported and it was more
than 55% even after 10 continuous cycles. It is reported by Hama et al. (2006) that
there are two types of lipases, one bound to the cell membrane (ROL 31) and the
other to the cell wall (ROL 34). They reported that the grow in the activity of enzyme
with addition of oleic acid or olive oil was because of raising in the production of
35
membrane bound lipase (ROL 31) recommending that ROL 31 has the main role in
the methanolysis activity (Ranganathan et al., 2007).
Hama et al. (2007) worked on biodiesel production by using a packed bed
reactor (PBR) utilizing Rhizopus oryzae whole cell biocatalyst by vegetable oil
methanolysis. In contrast with methanolysis reaction in a shaken bottle, the PBR
improved repeated batch methanolysis by protecting immobilized cells from excess
amounts of methanol and physical damage. Lipase producing Rhizopus oryzae cells
were immobilized within 6 mm · 6 mm· 3 mm cuboidal polyurethane foam biomass
support particles (BSPs) throughout the batch cultivation in a 20-l air-lift bioreactor.
It is reported by Hama et al. (2007) that the emulsification of the reaction
mixture ended in increased yield, around 75.5%, owing to the raise in interfacial
surface area while 63% conversion was obtained without emulsification. The
reaction mixture flow rate was varied between 5 and 55 l/h during the investigation
and higher flow rates caused immobilized enzyme exfoliation while low flow rates
ended to decline activity of enzyme because of weakness in mixing. To produce a
maximum conversion (90%) an optimum of 25 l/h was suggested. Noda and Fukuda
(2006) patented the process of implying whole cell biocatalyst and using waste oil
containing water claiming no reduction in the process efficiency (Ranganathan et al.,
2007). This promotes the process using intracellular lipase for industrialization and
commercialization.
36
2.6. Homogeneous and heterogeneous catalysts
With considering phase of catalyst, another categorization is based on catalyst
mobility, considers the catalysts as a two groups: homogenous and heterogeneous.
Homogeneous catalysis is the kind of reactions that engage a catalyst and
reactants in the same phase. Most frequently, a homogeneous catalyst is co dissolved
in a solvent with the reactants.
Heterogeneous catalysis, in chemistry, passes on some catalysis where the
catalyst phase differs from the reactants phase. Phase here refers not only to liquid,
gas and solid, but also liquids which are immiscible, e.g. water and oil. Most of
heterogeneous catalysts that are practical are solids and the great majority of
reactants are liquids or gases. This kind of catalyst, in many areas of the energy
industries and chemical, is of paramount importance. Heterogeneous catalysts are
important and significant for biodiesel manufacturing in industrial scale because of
their easy conversion at sensible temperatures (40-65 °C).
2.7. Non-catalytic transesterification
Indeed, the requirement for the catalyst can be eliminated if high temperature
and pressures are used to generate supercritical fluid conditions, under which
alcohols can either react directly with triglycerides or (in two stage procedures) with
fatty acids liberated from triglycerides. Typically there are three kinds of material in
case of specific physical states (i.e.: liquid, solid and gas), however, it can be in a
37
fluid state above the critical temperature in which condensation does not happen with
rising the pressure. This is point to situation as the supercritical state in which a fluid
that has different properties than those of a gas or liquid. The approaches of density
that of a liquid while its transport properties and viscosity are closer to that of a gas.
Exhibition of supercritical fluids outstanding transport properties coupled with highly
tunable solvent properties. The innovative work in the supercritical fluids application
to production of biodiesel is documented a sharp rise in the rate of reaction at
pressures and temperatures above the critical point of methanol (239 °C and 8.09
MPa) without the use of a catalyst.
In supercritical transesterification, alcohol is changed to the supercritical fluid
state by implying extreme temperature and pressure. The common temperature of
reaction is above 250 °C, because critical temperature of methanol is 239 °C
(D’Ippolito et al., 2007). In this condition, liquid methanol will move to the critical
point where both liquid and gas become indiscernible fluids that in which it will
shows both gas and liquid properties. Penetrating into solid like gas and dissolving
other material into them like liquid is possible (Leung et al., 2010). In this method
for pushing the reaction forward higher molar ratio is needed. The prospect of using
the supercritical method on ethanol and methanol for production of biodiesel from
crude jatropha oil studied by Madras and Rathore (2007). In their work, a 50-1
alcohol to oil molar ratio was confirmed as the best molar ratio in 20 MPa and 300
°C. The maximum conversion (70%) of jatropha oil to FAMEs was successfully
done in 10 minute and the rate of conversion continuously increased to 85% after 40
minute under the same conditions. The conversion percentage was higher by around
2.5% at the same conditions. In different conditions, they managed to get a better
percentage of conversion up to 95% at temperature of 400 °C for both ethyl and
methyl esters. Also Hawash et al. (2009) could obtain a 100% yield of methyl esters
under more moderate reaction conditions.
Supercritical method was used for converting jatropha oil to biodiesel within 4
minutes at 320 °C under 8.4 MPa and the jatropha oil to methanol molar ratio was 1-
38
43. An important improvement is because of the FFA content in jatropha oil. The
FFA content of the jatropha oil used by Hawash et al. (2009) was 2%, while for
Rathore and Madras (2007) the content of free fatty acid was above 10%. By other
words, higher free fatty acid content results to produce more water with esterification
reaction and so, the water will hydrolyze the esters which are generated from
transesterification reaction.
The major disadvantage of supercritical method is that ethyl/methyl esters are
degraded in a very high temperature (300 °C). Recently, jatropha oil was extracted
by employing supercritical carbon dioxide and then subjected to subcritical
hydrolysis. The hydrolyzed fatty acid acquired was, in addition, reacted with
methanol in supercritical esterification (methylation) condition. 99% biodiesel from
jatropha oil was produced during 15 minute just under 11 MPa and 290 °C with 33%
v/v of hydrolyzed oil to methanol by using supercritical methylation. Saka and
Ilham (2010) also used the two step process for producing biodiesel from jatropha
oil, but instead of methanol, dimethyl carbonate was utilized in the second step.
They could produce 97% of methyl esters in 15 minute at 300 °C and under 9 MPa.
More profitable glyoxal as the by product is produced instead of glycerol during this
process and high content of FFA does not affect this process. However, the dimethyl
carbonate cost is higher in comparison with ethanol/ methanol (Juan et al., 2007).
The special benefit of utilizing one or two step supercritical method is that
purification step is not needed to remove the catalyst. Even so, an extremely high
pressure usage, high temperature and using huge amount of alcohol are drawbacks to
industrialization of biodiesel production (Kulkarni and Dalai, 2006). Thus, more
studies in the process of biodiesel production and economic evaluations are required.
Advantages of supercritical method are: it is more environmentally friendly,
it has a lower reaction time, it is non catalytic process and it involves a much simpler
39
purification of transesterified products. Though, temperature which is needed for the
reaction is 150 to 300°C and the range of pressure is 35 to 60 MPa.
High methanol and energy consumptions and large amount of alkaline
wastewater could be considered as the disadvantages of using chemical catalysts.
Employing the enzymes like lipase is introduced as the useful method to overcome
these problems specially, the separation of glycerol without critical treatment. Even
though, again, cost is the major bottle neck associated with enzymatic catalysis.
2.8. Biodiesel purification and separation
Ineffective biodiesel purification and separation leads to several problems in
diesel engines like coking on injectors, plugging of filters, gelling of lubricating oil,
excessive engine wear, more carbon deposits, oil ring sticking, thickening and engine
knocking.
(Aroua et al., 2007). Table 2.8 demonstrates some effects of some
impurities (Berrios and Skeltonb, 2008).
40
Table 2.7. Impurities effect on biodiesel and engines
Impurity
Effect
FFA
Corrosion
Low oxidation stability
Water
Hydrolysis ( FFA formation)
Corrosion
Bacterial growth (filter blockage)
Methanol
Low values of density and viscosity
Low flash point (transport, storage and use problems)
Corrosion of Al and Zn pieces
Glycerides
High viscosity
Deposits in the injectors (carbon residue)
Crystallization
Metals (soap, catalyst)
Deposits in the injectors (carbon residue)
Filter blockage (sulphated ashes)
Engine weakening
Glycerol
Settling problems
Increase aldehydes and acrolein emissions
2.8.1
Conventional techniques for biodiesel separation:
It is reported that biodiesel with high quality which is viable in term of
economic can be produced when the appropriate method for biodiesel separation is
used. Commonly, separation of by-products, glycerol and biodiesel is conducted after
transesterification. Biodiesel separation process is based on these facts that the
glycerol and biodiesel produced are typically sparingly mutually soluble, and that
there is palpable difference in density between glycerol (1050 kg/m 3, or more) and
biodiesel (880 kg/m3) phases respectively. Furthermore, the difference in density is
sufficient for applying simpler techniques like centrifugation or gravitational settling
for the glycerol and biodiesel phases separation. Moreover, the separation rate of
41
biodiesel mixture is affected by many factors like biodiesel solubility in glycerol,
intense mixing, glycerol in biodiesel and formation of emulsion (Aroua et al., 2007).
2.8.2
Conventional techniques for biodiesel purification:
One method for purification of biodiesel is washing and removing free
glycerol and the other by products such as soap, unreacted catalyst and excess
alcohol from products is the major objective in this method. Alkyl ester drying is
required to overcome the limits of biodiesel specification on content of water in the
biodiesel product which is purified as well. However, other treatments are employed
to remove phosphorus, sulfur and glycerides from the biodiesel and reduce color of
biodiesel. By using water, acid can be added to neutralize the residual alkaline
catalyst. With this process removal of the salt products is simpler. After
transesterification and before the washing stage, the residual methanol should be
removed to reduce content of alcohol in the wastewater effluent (Aroua et al., 2007).
2.9. Optimization
Although there are complex steps in producing biodiesel, it is possible to
improve the quality of the final product by optimizing the production process. For a
certain biodiesel company which has a stable feedstock supply it is possible to design
a standard operation procedure to control the quality of the biodiesel production for a
set period. There is little research published on the optimization of the biodiesel
42
production method for quality control in industrial practice. This is the main rationale
in carrying out this research.
43
CHAPTER 3
METHODOLOGY
3.1
Research Methodology Approach
The aim of this study is optimization and production of biodiesel from
jatropha oil employing biocatalyst. The methodology could be divided into different
areas include catalyst preparation, test the prepared catalyst, catalyst and jatropha oil
characterization, biodiesel production, process optimization and product analysis.
Initially, the catalyst was immobilized to reduce catalyst deactivation and inhance the
performance of the catalyst. After lipase immobilization the performance of
immobilized lipase was tested. The phisico-chemical properties of jatropha oil was
determined and in next step, the experimental work was designed using design expert
softwer and then the batch production of biodiesel was conducted based on our
design. The biodiesel production process was modeled and optimized employing
RSM, ANN and GA and the optimized results was compared to the experimental
results. Subsequently, the produced biodiesel was analysed and the quality of
biodiesel was evaluated.
The schematic diagram of the different parts of methodology are illustrated in
Figure 3.1.
44
Catalyst preparation
Test and characterize the
catalyst
Jatropha oil
characterization
Biodiesel
Production
Product analysis
Process
optimization
Figure 3.1 Different steps of research methodology
3.2
Materials
R. oryzae lipase and methanol (98%) were obtained from Sigma Aldrich,
whereas sodium hydroxide (99%), hydrochloride acid (95%), boric acids (95%) and
PVA 72000 were supplied by Merck Schuchardt OHG, Darmstadt, Germany. Other
45
necessary chemicals, such as calcium chloride and sodium sulphate, were procured
from GCE laboratory Chemicals, UK. The sodium alginate was from Fluka,
Switzerland. Jatropha oil was obtained from a local supplier in Melaka, Malaysia.
The fatty acid composition of the oil sample was determined using GC-MS, also
different properties such as kinematic viscosity, water content, density, iodine
number and acid value were analyzed as well. These physicochemical properties are
shown in Table 3.1:
Table 3.1. The physico-chemical properties of Jatropha oil
Fatty
acid
compositions
Acid, common name
Capric
Lauric
Myristic
Palmitic
Stearic
Arachidic
Behenic
Palmitoleic
Oleic
Linoleic
Structure
C10:0
C12:0
C14:0
C16:0
C18:0
C20:0
C22:0
C16:1
C18:1
C18:2
Linolenic
C18:3
Saturated
Unsaturated
Property
Density at 25 °C
Viscosity (mm2/s) at 40 °C
Acid value (mg KOH/g)
Iodine number (mg KOH/g)
Moisture content %
Acid, IUPAC name
Decanoic acid
Dodecanoic acid
Tetradecanoic acid
Hexadecanoic acid
Octadecanoic acid
Eicosanoic acid
Docosanoic acid
Cis-9-hexadecenoic acid
Cis-9-octadecenoic acid
Cis-9,cis-12-octadecadienoic
acid
Cis-9,cis-12,cis-15octadecatrienoic acid
Weight [%]
0.1
0.1
18
7.1
0.3
0.3
0.9
41.6
31.4
0.2
25.8
74.2
Jatropha oil
0.90
33.20
4.71
101.70
0.06
46
3.3
Experimental
3.3.1 Preparation of Catalyst
A specific amount of lipase was mixed with phosphate and stirred at 140 rpm
for 10 min. The mixture was then kept at 4 °C before filtration under sterile
conditions. PVA and sodium alginate were utilized for enzyme immobilization. After
preparing the PVA and sodium alginate solution, lipase was added to the solution.
The mixture was then injected drop-by-drop into a solution containing boric acid and
calcium chloride, which was continuously stirred for 30 min to 50 min to produce
beads. The beads were kept at 4 °C for a day. Subsequently, 10% boric acid solution
was added while the beads were stirred for half an hour and treated with 0.5 M
sodium sulphate solution for another 30 min. The beads were then stored at 4 °C for
further use. This experiment was conducted under sterile conditions (Azimah et al.,
2011). The immobilization set up is illustrated in Figure 3.2.
Figure 3.2. PVA-alginate beads preparation set up
47
3.3.2
Determination of Physical Stability
In order to ensure the ability of PVA-alginate beads to remain insoluble in
water, physical stability experiment was done. For this test, 1 g of prepared beads
was immersed in 100 ml of distilled water for two days at room temperature. The
beads were monitored after two days to check for possible solubilization. This test
was triplicated.
3.3.3
Chemical Stability Test
Acid solution with PH 1-6 was prepared using hydrochloric acid (HCl) whilst
alkali solutions with PH 8-13 were prepared using sodium hydroxide (NaOH). 0.5 g
of immobilized beads were soaked in 20 ml of these solutions for 48 hours at room
temperature, thoroughly rinsed with deionized water and dried in a desiccator until
no further change in weight was detected (Khoo and Ting, 2001). Control was
prepared by soaking 0.5 g of immobilized beads in 20 ml of deionized water (Ph 7).
3.3.4
Catalyst characterization
Preparation of catalyst for field emission scanning electron microscopy
(FESEM) was as follows:
48
The moisture of catalyst beads were dried with a tissue and after that with the
purpose of obtaining the cross section of the beads surgical knife employed to cut
them. Then FESEM (model Ziess SUPRA 35 VP FE-SEM) used for taking the cross
section image of the beads (Jianlong and Wei, 2009). As for dry samples, the beads
were cut at first and later dried in desiccator until stable weight readings were caught.
Then the samples were placed on a sample stub and coated four times with platinum
coating by auto fine coater JFC-1600 (Joel, USA Inc, USA).
3.3.5
Mechanical Stability Test
The 200 ml beaker with 67.4 mm diameter and 107.7 mm height was divided
into four equal regions by the steel baffles (11 mm wide each). Then 1 g PVAalginate beads and 200 mL deionized water were added to the beaker and it was
placed on a stirring machine.
The agitation speed of 500 rpm was achieved employing stirring hotplate
machine (Cimarec 2, Thermolyne, USA). The PVA beads were agitated in the beaker
for 72 hours and then dried in desiccator until no further change in weight was
detected and the final weight was recorded (Khoo and Ting, 2001). In order to ensure
reproducibility each experiment was performed in triplicate.
49
3.3.6 Enzyme Assay
3.3.6.1 Lipase activity assay
One gram wet weight of lipase loaded beads or 0.1 ml of enzyme solution
was added to 10 ml of jatropha oil solution and incubated at 40 °C and 15 hr. This
experiment was done in sterile condition by filtering the solution used through 0.45
µm nylon filter. The lipase activity assay was done by utilizing dinitrosalicylic acid
(DNS) method proposed by Miller (1959). The color formation was determined by
measuring absorbance at 540 nm using UV mini 1240 UV VIS spectrophotometer
(Shimadzu Corporation, Japan). One international unit (IU) of activity was defined as
0.1 µmol of transesterified joatropha oil per minute under the assay conditions. The
beads were stored for 60 days in phosphate buffer pH 5 and the lipase activity was
assayed again for storage stability test by using the DNS method. The calculation on
the activity of the immobilized enzyme is shown in Appendix A.
3.3.7
Biodiesel production
The biodiesel manufacturing process was carried out using a 2 mm thick
glass vial as the reactor vessel. The vial was sealed tightly with a cap of silicon
rubber to preserve the vaporized compositions. The reactor vessel was placed in an
incubator shaker for agitation and warming. The reaction mixture consisted of
Jatropha oil, immobilized lipase, methanol and water.
50
At a certain reaction interval, a sample reaction mixture was taken and
centrifuged. A 100 µL sample was then withdrawn from the supernatant and
analyzed for biodiesel yield using gas chromatograph-mass spectroscopy (GC-MS).
Each test was carried out in triplicate to ensure reproducibility.
3.3.8
Sample analysis
Agilent technologies 6890N GC-MS was used to analyze the product with an
inert mass selective detector 5975. The capillary column was an Agilent 19091S-433
HP-5MS (30 mm × 250 µm × 0.25 µm), whereas the carrier gas was helium. The
initial temperature of the oven was 80 °C for 30 s and subsequently increased at the
rate of 10 °C/min until the temperature rose to 250 °C which was held constant for 5
min. The total run time of the analysis was 42 min. The temperature of the detector
was 250 °C, whereas the temperature of the injector was 325 °C. FAME yield was
calculated using the equation (1):
Yield % 
( FAME area from GCMS  weight of product )
( weight of oil sample)
 100%
(1)
51
3.3.9 Biodiesel physico-chemical properties
The physico-chemical properties of the biodiesel were determined using a
standard methods as follows:
3.3.9.1 Density (ASTM D4052)
A small volume (approximately 1 to 2 ml) of liquid sample is introduced into an
oscillating sample tube and the change in oscillating frequency caused by the change
in the mass of the tube is used in conjunction with calibration data to determine the
density of the sample.
Figure 3.3 The equipments for finding density
52
The procedure is:
i.
Pour a sample into the graduated cylinder. Prevented any formation of
bubble. Put the hydrometer into the sample.
ii.
Use thermometer for finding the sample temperature.
iii.
Take out the thermometer and leave the hydrometer freely floating in the
sample. Push the hydrometer into the sample for about 3 scale unit then
release it. Wait until the hydrometer is exactly stationary. Read the scale to
the nearest 0.0001 for SG, and to the nearest 0.5 for °API.
3.3.9.2
Kinematic Viscosity (ASTM D445)
Thomson - Mercer viscosity meter was used for finding the viscosity of biodiesel
sample. The time is measured for a fixed volume of liquid to flow under gravity
through the capillary of a calibrated viscometer under a reproducible driving head
and at a closely controlled and known temperature. The kinematic viscosity
(determined value) is the product of the measured flow time and the calibration
constant of the viscometer. Two such determinations are needed from which to
calculate a kinematic viscosity result that is the average of two acceptable
determined values. The procedure is:
i.
Use water as the standard liquid.
ii.
Use pipette to pour 10 ml of water into the capillary glass viscometer. Put the
capillary glass viscometer into the bathe. Wait until the temperature of water
and instrument is equal.
iii.
Using a vacuumed pump, suck water in the capillary glass viscometer until
the water level is about 5 mm above the level in the viscometer.
53
iv.
Measure the time taken for the water to flow from the high level to the low
level of the viscometer. If the following time is less than 200 seconds, repeat
the test by using a smaller capillary.
v.
Repeat the above experiment at different temperatures.
vi.
When the kinematic viscosity, µk of water is known, the viscometer constant
can then be calculated.
Figure 3.4
3.3.9.3
The Kinematic viscometer equipment
Water Content (ASTM D2709)
The electrical heater of MS-E104 (TOPO) was used for finding the water
content. This test method covers the determination of the volume of free water and
54
sediment in middle distillate fuels having viscosities at 40 °C (104 °F) in the range of
1.0 to 4.1 mm2/s (1.0 to 4.1 cst) and densities in the range of 770 to 900 kg/m3.
Figure 3.5
The water content determination equipment
The procedure is:
i.
Measure 100 ml of sample and 100 ml solvent using graduated cylinder.
ii.
Mix the sample and the solvent into the dean-stark apparatus.
iii.
Put in 5-10 glass beads into the apparatus.
iv.
Put the apparatus on an electric heater.
v.
Use retort stand to hold the apparatus.
vi.
Assemble the apparatus as in Figure 3.5.
vii.
Flow water through the condenser.
viii.
Heat the sample and regulate the heat such that the liquid drop in around 3
drops per second.
ix.
Heat until the collected water volume does not change with time.
55
x.
Rinse the condenser with solvent to collect the water in the inner condenser
wall.
xi.
Take the reading of the water volume in the graduated tube water collector.
3.3.9.4 Flash point (ASTM D93)
The flash point is determined using open cup method. The set up includes
natural gas cylinder, 50ml cup, thermocouple, Bunsen burner, and a gas nozzle. The
sample is filled in the cup till it reach 1cm from the top of the cup. The gas from the
cylinder splits into to the nozzle and to the Bunsen burner. The cup is fixed on a
stand with the thermocouple inserted. Adjust the Bunsen burner below the cup. After
opening the valves that control the gas flow both the Bunsen burner and the nozzle
were ignited. While kept on passing the fire nozzle 1cm above the cup the
temperature at which the surface catches fire was recorded as the flash point of the
biodiesel.
56
Figure 3.6
Open cup method for flash point determination
3.3.9.5 Pour point (ASTM D 97)
After preliminary heating, the sample is cooled at a specified rate and examined
at intervals of 3°C for flow characteristics. The lowest temperature at which
movement of the specimen is observed is recorded as the pour point. The pour point
of a petroleum specimen is an index of the lowest temperature of its utility for certain
applications. The procedure is:
i.
Pour the specimen into the test jar to the level mark.
ii.
Samples of residual fuels, black oils, and cylinder stocks which have been
heated to a temperature higher than 45°C during the preceding 24 h, or when
the thermal history of these sample types is not known, shall be kept at room
57
temperature for 24 h before testing. Samples which are known by the operator
not to be sensitive to thermal history need not be kept at room temperature for
24 h before testing.
iii.
Close the test jar with the cork carrying the high-pour thermometer. Adjust
the position of the cork and thermometer so the cork fit tightly, the
thermometer and the jar are coaxial, and the thermometer bulb is immersed so
the beginning of the capillary is 3 mm below the surface of the specimen.
iv.
Place the gasket around the test jar, 25 mm from the bottom. Insert the test jar
in the jacket. Never place a jar directly into the cooling medium.
v.
Pour points are expressed in integers that are positive or negative multiples of
3°C. Begin to examine the appearance of the specimen when the temperature
of the specimen is 9°C above the expected pour point (estimated as a multiple
of 3°C). At each test thermometer reading that is a multiple of 3°C below the
starting temperature remove the test jar from the jacket. To remove
condensed moisture that limits visibility wipe the surface with a clean cloth
moistened in alcohol (ethanol or methanol). Tilt the jar just enough to
ascertain whether there is a movement of the specimen in the test jar. If
movement of specimen in the test jar is noted, then replace the test jar
immediately in the jacket and repeat a test for flow at the next temperature,
3°C lower. Typically, the complete operation of removal, wiping, and
replacement shall require not more than 3 s.
vi.
If the specimen has not ceased to flow when its temperature has reached
27°C, transfer the test jar to a jacket in a cooling bath maintained at 0 ±1.5°C.
As the specimen continues to get colder, transfer the test jar to a jacket in the
next lower temperature cooling bath.
vii.
Continue in this manner until a point is reached at which the specimen shows
no movement when the test jar is held in a horizontal position for 5 s. Record
the observed reading of the test thermometer.
58
[[[
3.3.9.6 Cloud Point (ASTM D2500)
The specimen is cooled at a specified rate and examined periodically. The
temperature at which a cloud is first observed at the bottom of the test jar is recorded
as the cloud point. For petroleum products and biodiesel fuels, cloud point of a
petroleum product is an index of the lowest temperature of their utility for certain
applications. The 93511- Seta automatic frigistal equipment was used for
determination of the pour point and cloud point. The procedure is:
i.
Pour the sample into the test jar to the level mark.
ii.
Close the test jar tightly by the cork carrying the test thermometer. Use the
high cloud and pour thermometer if the expected cloud point is above −36°C
and the low cloud and pour thermometer if the expected cloud point is below
−36°C. Adjust the position of the cork and the thermometer so that the cork
fits tightly, the thermometer and the jar are coaxial, and the thermometer bulb
is resting on the bottom of the jar.
iii.
See that the disk, gasket, and the inside of the jacket are clean and dry. Place
the disk in the bottom of the jacket. The disk and jacket shall have been
placed in the cooling medium a minimum of 10 min before the test jar is
inserted. The use of a jacket cover while the empty jacket is cooling is
permitted. Place the gasket around the test jar, 25 mm from the bottom. Insert
the test jar in the jacket. Never place a jar directly into the cooling medium.
iv.
Maintain the temperature of the cooling bath at 0 ±1.5°C.
v.
At each test thermometer reading that is a multiple of 1°C, remove the test jar
from the jacket quickly but without disturbing the specimen, inspect for
cloud, and replace in the jacket. This complete operation shall require not
more than 3 s. If the oil does not show a cloud when it has been cooled to
9°C, transfer the test jar to a jacket in a second bath maintained at a
temperature of −18 ± 1.5°C. Do not transfer the jacket. If the specimen does
not show a cloud when it has been cooled to −6°C, transfer the test jar to a
jacket in a third bath maintained at a temperature of −33 ± 1.5°C. For the
determination of very low cloud points, additional baths are required. In each
59
case, transfer the jar to the next bath, if the specimen does not exhibit cloud
point and the temperature of the specimen reaches the lowest specimen
temperature in the range identified for the current bath in use.
vi.
Report the cloud point, to the nearest 1°C, at which any cloud is observed at
the bottom of the test jar, which is confirmed by continued cooling.
Figure 3.7
The Pour and cloud point determination equipment
3.3.9.7 Acid value (ASTM D664)
A standard NaOH solution 0.1 M was prepared by dissolving a standard 25ml
1M solution of NaOH by 250ml of distilled water. 25ml of diethyl ether and 25ml of
ethanol was mixed in a 250ml beaker for each sample. Then a weighted 10g of
sample was mixed with the 50ml diethyl ether and ethanol mixture in a 250 ml
conical flask followed by putting 2 drops of phenolphthalein indicator. Once the
60
burette was filled (note initial volume, V0) with the 0.1M NaOH solution the titration
began. The titration was accompanied by a constant shaking and the end point was
observed when a dark pink color appears (note volume, Vf).
AV 
(V f  V0 )  N  56.1
W0
(2)
where, V0 is initial volume of NaOH (ml), Vf shows the final volume of NaOH (ml)
and N is the concentration of NaOH (molar).
3.4 Experimental design and process optimization:
Transesterification of Jatropha oil with R. oryzae lipase was optimized for
biodiesel production through the RSM, ANN, and GA methods. These three methods
adopted the experimental design for modeling and optimizing FAME yield as a
response (Lee et al., 2011).
3.4.1
Response surface methodology
In a multi-variable system, response surface methodology (RSM) which
utilizes statistical methods can provide a research strategy for studying parameter
interaction. RSM can simulate biodiesel production in different conditions with
61
minor estimation error. By utilizing this technique, existing data can be used to avoid
trial runs by clarifying the reaction conditions and the type of catalyst (Bezerra et al.,
2008).
In RSM, four factors or independent variables with two levels were
employed to determine the effect of variables on FAME yield: reaction temperature
(A), reaction time (B), methanol/oil ratio (C), and water content (D). These
independent variables, were coded at two levels between -1 and +1, where -1
corresponds to the minimum and +1 corresponds to the maximum value of each
variable, as noted in Table 3.2. Prior to experimental design we have performed
preliminary experiments to know the best range of the four variables (A-D). Based
on the preliminary results, we have chosen the appropriate range for the experimental
design.
Table 3.2: Experimental design of biodiesel production from jatropha oil
Low actual value
(-1)
High actual value
(+1)
A: Time (h)
1
17
B: Temperature (°C)
30
50
C: Methanol/oil Ratio
3:1
7:1
E: Water content (wt. %)
10
130
Independent variable
The total number of experimental runs was calculated using (2K+2K+6),
where K is the number of variables. For the current study, the number of experiments
was 30. Twenty-four experiments and six imitation tests at the design center were
conducted at random to estimate the error. The coding of test factors was based on
Eq. (3) in the regression equation as follows:
62
Xi 
U i  U i
U i
(3)
where Xi is the coded value of the independent variable, Ui is the real value of the
independent variable, Uiᴼ is the real value of the independent variable at the center
point of the independent variable, and ∆𝑈𝑖 demonstrates the step change in Ui .
Equation (3) is a suitable and simple model for optimization, with its
response based on chosen variables by quadratic and linear terms as follows:
k
k
k
j 1
j 1
      j x j    jj x 2j    ij xi x j  ei
(4)
i  j 2
where 𝜂 is the response, 𝛽0 is the constant coefficient, 𝑥𝑖 and 𝑥𝑗 are the independent
factors, 𝛽𝑗 , 𝛽𝑗𝑗 , and 𝛽𝑖𝑗 are the coefficients for linear, quadratic, and interaction
effects, respectively, and 𝑒𝑖 is the error. The fit of polynomial quality was indicated
by determination coefficients R2 and R2adj in Eqs. (4) and (5), whereas the F-value
and p-value, as well as Eqs. (6) and (7), were used to measure statistical significance
(Ozer et al., 2009).
R2  1 
SS resedual
SS mod el  SS residual
(5)
SS residual
2
Radj
 1
( SS mod el  SS residual)
DFresidual
( DFmod el  DFresidual)

Adequate precision 
(6)

Max. (Y) - Min. (Y )

V (Y )
(7)
63

1 n
P 2
V (Y )  V (Y ) 
n 1
n

(8)
In these equations, DF is the degree of freedom, P is the number of model
parameters, SS is the sum of squares, 𝜎 2 is the mean square of the residual from
ANOVA results, and n is the quantity of experiments. Subsequently, the response
yield was measured at the end of the experiment.
3.4.2
Artificial neural network
The artificial neural network (ANN) model is used to simulate the process in
the current study. ANN is the information processing model simulated from the
biological neurons system. The interrelated network construction of straightforward
processing basics conducts the processing of data with parallel computation (Antonio
and Pilar, 2006; Salamatinia et al., 2010).
The back propagation (BP) network was used for the selected simulation
model with Marquadt Levenberg Algorithm (MLA) owing to its fast training in
ANN. As presented in Figure 3.8, the BP network includes input, hidden, and output
layers. Input variable values are forwarded to hidden layers by input nodes, followed
by the estimation of the final model by output nodes.
64
Figure 3.8. BP network model
In ANN, several important parameters exist in network training. One is
learning rate, which is a factor for the speed of training. A moderate learning rate can
help obtain a good network in a short time. The number of neurons in a hidden layer
is another option for a good network, and the performance can be increased with
more neurons. However, a constantly high number of neurons does not indicate
better prediction capability.
The data in the training process are divided into three parts, namely, training,
validation, and testing data. The training data are employed in the network during
training and the weights and biases are adjusted according to their error. Validation
data are used to measure network generalization. The training can be stopped when
generalization stops improving. These two sets of data dictate the performance of the
approximation function and the prediction capacity. Testing data do not affect the
training and just provide an independent measure of network performance during and
after training. A training algorithm, which is a rule of ANN, is used to update the
weights and biases of the network to reduce the error caused by the contrast between
the values of the simulated data and the target (experimental data). In the current
work, MLA is adopted for the purpose of assisting these parameters.
65
The applications of ANN in some fields, such as function approximation,
classification, and so on, are restricted. Thus, applications are developed by
combining ANN with other methods, including statistical method, expert systems,
wavelet transform, genetic algorithm (GA), and fuzzy logic. Among these
techniques, GA collaborating with ANN can be an efficient method for the prediction
and optimization of complicated process parameters.
With these considerations, to obtain a higher yield of biodiesel as the final
product, this study investigates the best operating conditions by employing
appropriate modeling and optimization techniques, such as RSM, ANN, and GA and
by designing experiments using Design-Expert (DE) 8.0.6 and MATLAB 7.11.0
(R2010b) software programs.
3.4.3
Genetic algorithm
GA utilizes the population of individuals to explore all areas of the solution.
Each set of independent variables refers to an individual. In the present study, these
individuals are reaction temperature and time, methanol/oil ratio, and water content.
At first, individual population is randomly organized. By employing the fitness
function, individual fitness is assessed. After fitness evaluation, genetic operations
like crossover and mutation are used to make further individual generations to
evaluate fitness, which continues until the optimum solution is discovered. The
present research utilized RSM and ANN as the fitness function for GA to optimize
the process of biodiesel manufacturing (Jianlong and Wei, 2009).
66
CHAPTER 4
RESULTS AND DISSCUTION
4.1 Innovation in immobilization
The technique which was used in current research for immobilization was an
innovated method which first time used by Azimah M. Z. and her colleagues in 2008
for immobilizing invertase (Ani et al., 2008). The difference of this method from the
other works is using sodium sulfate as a treatment solution at the end of the
immobilization process. Also the best amount of sodium alginate concentration has
founded by them. In addition they have proved that the enzyme activity can improve
by this immobilization. After evaluating the immobilization method in previous work
by Azimah et al. (2011), in current work R. oryzae lipase was immobilized with this
technique for the first time.
67
4.2 Surface morphologies
Field emission scanning electron microscopy (FESEM) was used to examine the
surface morphologies and to analyze the trace element in the PVA-alginate beads, as
well as the cross-sectional structure of the beads. Figure 4.1 clearly portrays the
formed layers of cross-linked polymers in the inner surface of the beads, where the
microstructure and the stability of the beads with uniformly distributed layers can be
distinguished. The stability of the beads can be attributed to the use of highly
concentrated PVA. As a result, the beads had more compact layers with much denser
structure and because of that beads did not solve in a warm and shaking mixture
during the process (Azimah et al., 2011).
Figure 4.1. Inner surface of the bead
The formation of the pores on the outer layer of the beads is revealed in
Figure 4.2. The pores are uniformly distributed with average sizes. Moreover, the
surface morphology of the support shows that the enzyme is attached to the support
layers (Ani et al,. 2008).
68
Figure 4.2. Outer layer of a bead
4.3 Lipase activity assay
The free enzyme activity was 1098.3 IU of activity while the activity for
immobilized enzyme was 1204.4 IU. The activity of immobilized enzyme is more
than free enzyme probably because of the right amount of boric acid used to
crosslinke PVA resulting in a good crosslinking and well protected microenvironment for lipase. Immobilization may confer protection to the enzyme from
the denaturing effect of heat since it is well bound inside the immobilized matrix. In
addition, the immobilization matrix may also protect the enzyme from other factors
that would greatly influence enzyme activity and its kinetic behavior, such as the pH
of the solution and presence of certain inhibitors such as glycerol and alcohol.
69
4.4 PVA-Alginate Beads Stability
4.4.1 Chemical Stability Test
In this experiment, the beads untreated with acidic and alkali solutions were
used as a control i.e immersed in distilled water (pH 7). The remaining weight of
each bead treated with both hydrochloride acid (HCl) and at sodium hydroxide
(NaOH) at various pH values from 1.0 to 6.0 and 8.0 to 14.0 were recorded. The
remaining weight of the treated beads with HCl and NaOH were compared with that
of the control. Generally, under acidic condition within the pH range of 2 to 6, all the
beads displayed relatively lower weight loss percentage compared to the beads that
were treated with NaOH. The weight of the beads was stable throughout the period
of experiment and physically the beads were very stable within pH 2 to 7.
Considerable weight losses were recorded for beads that were stirred at the pH
of 8 to 11. The excess OH- ions that are present in the solution may act as a
chelating agent leading to the de-crosslinking of the beads. Obviously, there is a
probability for the beads to retain its weight in the solution with the pH ranging from
2 to 7 which is quite acidic in nature.
Treatment with sodium sulfate has positively affect the chemical stability of the
beads as beads that were not treated with sodium sulfate would rapidly and
completely dissolve in HCl and NaOH solutions rendering it impossible to record the
weight loss. By turning the beads virtually insoluble in both solutions, weight lost via
dissolution has been appreciably reduced.
70
The observed weight gain could be explained by the adsorption of chloride ions
(Cl-) that are present in the HCl solution during the chemical stability test. Part of the
uptake could be due to the ionic interaction between calcium ions (Ca2+) bound by
the polyguluronate units of alginate and the Cl- that are presented in HCl solution.
Therefore when the beads were immersed in a more concentrated HCl solution, more
Cl- would be adsorbed. This would cause an increase in the percentage of the
remaining weight of beads beyond its original weight as pH decreases, i.e. as Cl concentration increases. However, at the lowest pH value which is at pH equal to 1,
at which the concentration of HCl is 0.1 M, there is a notable decrease in the
percentage of remaining weight which seems to contradict the hypothesis that more
weight gain could be observed as pH decreases. At this particular point, the Ca2+ guluronate complex starts to disintegrate as it underwent proton-catalyzed hydrolysis
therefore decreasing the beads ability to uptake Cl- (Bajpai and Sharma, 2004).
4.4.2 Mechanical Stability Test
The PVA-alginate beads that were fabricated may be characterized as highly
elastic (Miller, 1959). Continuous stirring with constant RPM was conducted for
each run of experiment. Khoo and Ting (2001) recorded the difference in diameter
and weight of their PVA beads before and after stirring. But in this study, the method
is slightly different, only the weight was recorded and its recording was only after the
beads were dried in dessicator. The measurement of the beads diameter was opted
out because the diameter of the PVA-alginate beads will change after the beads were
immersed in water.
The percentage of weight remaining for each run in the mechanical stability test
was determined and it varied from 40.8 % to 91.2 %. Complete gelation inside the
71
beads that could be achieved within 24 h via treatment with saturated boric acid may
also play a role as incomplete gelation may cause the beads to become soluble. Also
adequate amount of borate ions to crosslink the high concentration of PVA
molecules would result in higher mechanical stability.
4.5 Sample analysis
The collected biodiesel were analyzed to determine several physical
properties and compare them to the ASTM D6751 to measure the quality of
biodiesel. A short description of these properties and their acceptable ranges are
included in parts 4.5.1 to 4.5.6 and at the end the amount of these properties that
were gained from experimental runs are shown in Table 1.
4.5.1 Kinematic viscosity
The kinematic viscosity is an important property to evaluate the quality of
biodiesel. The kinematic viscosity is a ratio of the inertial force to the viscous force.
This ratio is defined as follows:
𝜈=
𝜇
𝜌
(1)
The kinematic viscosity of liquid petroleum products is determined by
international standard. The feedstocks normally have kinematic viscosity in the range
72
of 32.20 to 48.47 mm2/s but the viscosity of methyl ester is around 10 times lower
than the oils. The required value for kinematic viscosity of biodiesel is 3.5-5 mm2/s.
4.5.2 Density
Even more than kinematic viscosity, density is another important property to
evaluate the quality of biodiesel. The density of biodiesel depends on the quantity of
methyl ester and remains methanol in product. As a result the type of feedstock is
effective on density of biodiesel. The required value for density of biodiesel is 860900 kg/m3.
4.5.3
Acid value
The content of free acids in the sample which effect on fuel aging was
measured by acid value. The amount of free acid in feedstock is very effective in
transesterification and it should not exceed from specific limitation (≤ 0.5).
73
4.5.4 Water content
Another important property that affect on the biodiesel quality is water
content. The existence of water in the biodiesel leads to corrosion problem in the
engine. It also reacts with glycerides to produce soap and glycerol.
4.5.5
Flash Point
The flash point of a volatile material is the lowest temperature at which it can
vaporize to form an ignitable mixture in air. Measuring a flash point requires an
ignition source. At the flash point, the vapor may cease to burn when the source of
ignition is removed. The flash point is not to be confused with the auto ignition
temperature, which does not require an ignition source, or the fire point, the
temperature at which the vapor continues to burn after being ignited. Neither the
flash point nor the fire point is dependent on the temperature of the ignition source,
which is much higher. Flash point refers to both flammable liquids and combustible
liquids. There are various standards for defining each term. Liquids with a flash point
less than 60.5 °C (140.9 °F) or 37.8 °C (100.0 °F) - depending upon the standard
being applied - are considered flammable, while liquids with a flash point above
those temperatures are considered combustible. The range of biodiesel flash point is
100-170 °C.
74
4.5.6
Pour and Cloud Points
The pour point of a liquid is the lowest temperature at which it becomes semi
solid and loses its flow characteristics. Also, the pour point can be defined as the
minimum temperature at which a liquid, particularly a lubricant, will flow.
The cloud point of a fluid is the temperature at which dissolved solids are no
longer completely soluble, precipitating as a second phase giving the fluid a cloudy
appearance. This term is relevant to several applications with different consequences.
In the petroleum industry, cloud point refers to the temperature below which wax in
diesel or bio-wax in biodiesels form a cloudy appearance. The presence of solidified
waxes thickens the oil and clogs fuel filters and injectors in engines. The wax also
accumulates on cold surfaces (e.g. pipeline or heat exchanger fouling) and forms an
emulsion with water. Therefore, cloud point indicates the tendency of the oil to plug
filters or small orifices at cold operating temperatures.
The acceptable range of pour and cloud points are -5 to 10°C and -3 to 15°C
respectively.
Table 4. 1. Biodiesel properties
Property
Biodiesel
Density (kg/m3)
876
Kinematic viscosity (at 40°C St(m2/s)×10-6
2.39
Acid value (mg KOH/g of oil)
0.63
Water content (%)
0.04
Flash point(°C)
163
Cloud point(°C)
4
Pour point(°C)
5
75
4.6
Experimental Results
The relationships between the four independent variables (reaction temperature,
time of reaction, methanol/oil ratio, and water content) and biodiesel yield were
studied. The experimental design and the FAME yield for each experimental run are
listed in Table 4.2 where the yields from the ANN and RSM models were also
compared with the experimental data.
76
Table 4.2. Experimental design matrix and experimental results of the response.
Run
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Temperature
(°C)
35.00
45.00
35.00
45.00
35.00
45.00
35.00
45.00
35.00
45.00
35.00
45.00
35.00
45.00
35.00
45.00
30.00
50.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
Reaction
time
(h)
5.00
5.00
13.00
13.00
5.00
5.00
13.00
13.00
5.00
5.00
13.00
13.00
5.00
5.00
13.00
13.00
9.00
9.00
1.00
17.00
9.00
9.00
9.00
9.00
9.00
9.00
9.00
9.00
9.00
9.00
Methanol/oil
ratio
(%)
4.00
4.00
4.00
4.00
6.00
6.00
6.00
6.00
4.00
4.00
4.00
4.00
6.00
6.00
6.00
6.00
5.00
5.00
5.00
5.00
3.00
7.00
5.00
5.00
5.00
5.00
5.00
5.00
5.00
5.00
Water
content
(%)
40.00
40.00
40.00
40.00
40.00
40.00
40.00
40.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
100.00
70.00
70.00
70.00
70.00
70.00
70.00
10.00
130.00
70.00
70.00
70.00
70.00
70.00
70.00
Exp.
Yield
(%)
77.50
79.00
84.00
85.00
76.00
77.00
82.00
84.00
79.00
79.00
83.00
84.00
79.00
77.00
83.00
84.00
80.00
81.00
75.00
87.10
82.00
82.00
81.00
81.00
84.00
85.00
84.50
85.00
84.00
84.80
MSE
R2
ANN
Yield
(%)
77.99
78.88
84.20
86.42
76.17
76.84
81.79
85.13
80.24
79.15
83.46
83.93
79.06
77.14
83.45
83.89
80.90
81.18
75.70
87.07
82.34
81.70
81.25
81.05
84.60
84.60
84.60
84.60
84.60
84.60
RSM
Yield
(%)
77.83
78.77
83.45
85.51
76.43
77.00
82.68
84.37
78.93
78.50
83.18
83.87
78.66
77.85
83.53
83.85
79.64
80.89
75.00
86.62
82.47
81.05
80.47
81.05
84.55
84.55
84.55
84.55
84.55
84.55
0.28
0.97
0.19
0.98
77
4.7
Regression model and statistical analysis
The response gained in Table 4.2 corresponded to the four independent variables
utilizing a polynomial equation such as Eq. (3) in chapter 3. To fit the acquired data
to Eq. (3)-chapter 3, least squares regression was employed. The best fit model in
terms of uncoded or actual and coded factors for FAME yield are as follows:
Uncoded fit model:
Yield  20.83  3.62 A  1.22B  6.36C  0.22 D  0.01AB  0.20 AC  2.3 10 3 AD 
3
3
3
0.40BC  2.86 10 BD  9.37 10 CD  0.04 A  0.06 B  0.70C  1.05 10 D
2
2
2
(2)
2
Coded fit model:
Yield  84.55  0.31A  2.90 B  0.35C  0.15D  0.28 AB  0.09 AC  0.34 AD
 0.16 BC  0.34 BD  0.28CD  1.07 A  0.93B  0.70C  0.95D
2
2
2
(3)
2
Table 4.3 contains the ANOVA assessments of the best fit model indicates
that the model is suitable for describing the experimental work. The parameters pvalue, F-value, lack of fit, and R2 were employed to measure the fitness of the
suggested model to the experimental data. The F-value at 51.88 indicated the
significance of the quadratic model. Furthermore, the impact of each term in the
model was evaluated, whereas the highly significant effect of some variables was
concluded based on the F-value. The linear terms for reaction temperature (A), time
(B), and methanol/oil ratio (C), also all the variables in quadratic form, and the
interaction terms for two factors (AD and BD) tabulated in Table 4.3 have large
effects on biodiesel yield because of high F-values and low p-values (<0.05).
78
Table 4.3. Analysis of variance (ANOVA) for model regression.
Source
Sum of
Squares
Model
A-Temp.
B-time
C-ratio
D-Water cont.
AB
AC
AD
281.4305
2.34375
202.4204
3.010417
0.510417
1.265625
0.140625
1.890625
BC
BD
CD
A^2
B^2
C^2
D^2
Residual
Lack of Fit
Pure Error
Cor Total
Mean
Square
F Value
p-value
Prob > F
14
1
1
1
1
1
1
1
20.10218
2.34375
202.4204
3.010417
0.510417
1.265625
0.140625
1.890625
51.87659
6.048387
522.3753
7.768817
1.317204
3.266129
0.362903
4.879032
< 0.0001 significant
0.0265
< 0.0001
0.0138
0.2691
0.0908
0.5559
0.0432
0.390625
1.890625
1
1
0.390625
1.890625
1.008065 0.3313
4.879032 0.0432
1.265625
31.51313
23.9467
13.32027
24.5917
5.8125
1
1
1
1
1
15
1.265625
31.51313
23.9467
13.32027
24.5917
0.3875
3.266129
81.32419
61.79793
34.37488
63.46244
4.7375
1.075
287.243
10
5
29
0.47375
0.215
df
0.0908
< 0.0001
< 0.0001
< 0.0001
< 0.0001
2.203488 0.1983
Not
significant
A p-value greater than 0.05 indicates that the corresponding model term is not
significant. As shown in Table 4.4, the two interaction terms, temperaturemethanol/oil ratio (AC) and time-methanol/oil ratio (BC), registered the highest pvalue; thus, they were considered insignificant terms and detached to improve the
current model and to produce an enhanced new model.
The determination coefficient and the regression equation were evaluated to
test the fit of the model (Mazaheri et al., 2010). The determination coefficient of the
reduced model decreased from 0.9798 to 0.9779 as a result of the reduction of
variables from the regression model (Yuan et al., 2008). On the other hand, the
adjusted determination coefficient increased from 0.9609 to 0.9623 after excluding
79
unnecessary terms (Fermoso et al., 2009). The F-value of the modified model
increased to 62.73 and its p-value still was less than 0.0001, which imply that the
model improved after deleting the AC and BC terms.
Table 4.4. ANOVA for Response Surface Reduced Quadratic Model
Source
Model
A-Temperature
B-time
C-ratio
D-Water
content
AB
AD
BD
CD
A^2
B^2
C^2
D^2
Residual
Lack of Fit
Pure Error
Cor Total
Sum of
Squares
280.8993
2.34375
202.4204
3.010417
df
12
1
1
1
Mean
Square
23.40827
2.34375
202.4204
3.010417
F Value
62.72955
6.280788
542.4468
8.067323
p-value
Prob > F
< 0.0001
0.0227
< 0.0001
0.0113
0.510417
1.265625
1.890625
1.890625
1.265625
31.51313
23.9467
13.32027
24.5917
6.34375
1
1
1
1
1
1
1
1
1
17
0.510417
1.265625
1.890625
1.890625
1.265625
31.51313
23.9467
13.32027
24.5917
0.373162
1.367816
3.391626
5.066502
5.066502
3.391626
84.44897
64.17243
35.69569
65.9009
0.2583
0.0830
0.0379
0.0379
0.0830
< 0.0001
< 0.0001
< 0.0001
< 0.0001
5.26875
1.075
287.243
12
5
29
0.439063 2.042151
0.215
0.2220
significant
not
significant
The final practical model after modification based on the analysis of ANOVA
data and statistical parameters is given by the following equations:
In terms of uncoded factors:
Yield  18.84  3.53 A  1.42 B  5.96C  0.22D  0.014 AB  2.30 10 3 AD 
3
3
3
2.86 10 BD  9.38 10 CD  0.04 A  0.06 B  0.70C  1.05 10 D
And in terms of coded factors:
2
2
2
2
(4)
80
Yield  84.55  0.31A  2.90 B  0.35C  0.15D  0.28 AB  0.34 AD  0.34 BD (5)
 0.28CD  1.07 A 2  0.93B 2  0.70C 2  0.95D 2
The p value <0.0001 means there is only a 0.01% probability that a model
with a similarly large F-value becomes the product of noise in the data. Moreover,
the lack of fit F-value was 2.04, denoting that the lack of fit was not significant in
relation to pure error (Wang et al., 2010).
To test the capability of the model, the residual distribution was analyzed by
detecting whether the residuals followed a normal distribution or not. Residuals are
the difference between actual and predicted values obeying a normal distribution as
long as the experimental error is random (Korbahti et al., 2008). The normalization
of residuals was primarily done according to their standard deviations or
―studentized‖ at first. Afterward, as shown in Figure 4.3 predicted studentized
residuals based on best fit normal distribution were plotted against experimental
studentized residuals. The straight line formed in Figure 4.3 proves that the
studentized residuals followed a normal distribution. Otherwise, an S-shape would
have been formed (Korbahti et al., 2008).
81
Normal Plot of Residuals
99
Normal % Probability
95
90
80
70
50
30
20
10
5
1
-2.14
-1.02
0.10
1.23
2.35
Internally Studentized Residuals
Figure 4.3. Normal % probability and studentized residual plot.
Studentized residuals are plotted against predicted biodiesel yield in Figure
4.4. In this figure, the plot is seen as randomly scattered, showing that the deviation
in the original observation was not because of response value (Korbahti et al., 2008).
The random scatter also implies that the suggested model is suitable to describe the
process.
82
Residuals vs. Predicted
Internally Studentized Residuals
3.00
1.50
2
0.00
2
-1.50
-3.00
74.98
77.88
80.77
83.67
86.56
Predicted
Figure 4.4. The studentized residuals and predicted response plot.
On the other hand, Figure 4.5 plots the predicted and actual FAME yields. In
the experimental designs, R2 was gained by calculating the variation amount around
the mean which was described by the related model (Myers and Montgomery, 2000).
R2 and R2adj values gained from the data in Figure 4.5 are very close, indicating that
significant terms are involved in the model; the high value of R2adj represents the high
significance of the model (Myers and Montgomery, 2000).
83
Predicted vs. Actual
87.00
2
2
Predicted
83.75
80.50
77.25
74.00
74.98
77.98
80.99
83.99
87.00
Actual
Figure 4.5. The actual and predicted plot
The outlier t-plot of the biodiesel production data is shown in Figure 4.6. This
plot shows the residual importance for each run to determine if any of the
experiments had particularly large residuals. According to Figure 4.6, no data can be
seen outside the intervals, proving the consistency of the model with all the data.
84
Residuals vs. Run
Internally Studentized Residuals
3.00
1.50
0.00
-1.50
-3.00
1
5
9
13
17
21
25
29
Run Num ber
Figure 4.6. The Outlier t plot.
4.8
Influence of reaction temperature and time
Statistical analysis of the experimental data showed that time (B) is an
effective variable in analyzing response. This variable has a large and positive
influence on FAME yield. Temperature (A) is also an effective variable, but it has
less impact because of low F-value and high p-value. Figure 4.7 demonstrates the
surface plot of response for the production of biodiesel. The biodiesel yield varied
with changes in reaction time and temperature under conditions that the methanol/oil
ratio was 5% and water content was 70% (w/w). The maximum biodiesel yield
obtained was 86.9% at 41.58 °C. The total reaction time had a positive effect on
FAME yield. As illustrated in Figure 4.7, the maximum FAME yield was achieved at
15.4 h. This conditions, compare to other works on biodiesel production from
85
jatropha oil, are reasonable. For instance Hawash et al. (2009), could produce
biodiesel from jatropha oil in supercritical condition by 100% yield in 4 min but in
320 °C, pressure of 8.4 MPa and methanol/oil ratio of 43:1 which this conditions
lead to very high price process of biodiesel production. As an another example,
Sriappareddy et al. (2008), produced biodiesel from jatropha oil using Rhizopus
oryzae on polyurethane foam particles by 80% conversion under ambient conditions
( T=30 °C and alcohol/oil ratio=3:1), but with longer reaction time (60 h).
Figure 4.7. The effect of reaction temperature and reaction time on FAME yield, for
water content of 70% and Methanol/oil Ratio of 5%.
86
4.9
Methanol/oil ratio and water content effects on FAME yield
The effect of methanol/oil ratio on FAME yield was positive. Based on the
stoichiometry of the transesterification reaction, at least three moles of alcohol are
required to produce three moles of ester for a reaction with one mole of oil. The
reversibility of the reactions along with the methanol/oil ratio results in more
products of methyl ester. The effect of methanol/oil ratio and moisture content on
biodiesel yield is shown in Figure 4.8. The reaction temperature and time were 41.58
°C and 15.41 h, respectively. The maximum FAME yield was 86.98% with a
methanol/oil ratio of 5.09%. The water content has a non-significant effect on the
yield of biodiesel production. Figure 4.8 illustrates that 86.98% yield was achieved
during the maximum response with the water content at 62.96.
Figure 4.8. The effect of methanol/oil ratio and water content on FAME yield for a
total time of 15.41 h and reaction temperature of 41.58°C.
87
4.10 Prediction with ANN
The weights and biases of the BP network are shown in Table 4.5. Biodiesel
yield under different ranges of input conditions can be predicted by this table. The
neural network (NN) described in Table 4.5 is limited to predicting the data inside
the region of the training samples. This limitation restricts the extrapolation to other
areas and could end to nonsensical outcomes (Antonio et al., 2006).
Table 4.5. Weights and biases of BP network
1
2
3
4
5
6
a
b
W1 a
X1
5.4469
-2.5956
3.1446
1.0327
-0.9082
-3.9774
X2
-0.1077
3.3872
4.1704
-6.6608
-1.2714
4.2756
X3
-0.1834
-2.5042
0.8810
-1.1976
0.0945
1.1401
X4
5.6632
10.8558
-1.7310
1.2791
0.2958
-2.6631
a
Bias
-2.7542
3.5282
-1.8694
-1.9580
-0.5990
2.1926
1
2
3
4
5
6
Biasb
W2 b
-0.2295
0.2752
0.1914
0.0915
-0.6393
0.2344
-0.4413
Weights and biases between the input layer and hidden layer;
Weights and biases between the hidden layer and output layer.
To optimize the parameters of the network, bias and weight values of hidden,
input, and output layer neurons were detected according to the maximum amount of
determination coefficient (R2) and by another important factor, namely, the mean
squared error (MSE), which refers to the lower value indicating better performance.
These two factors were calculated by the following equations (Antonio et al., 2006):
MSE 
2
1 n
(
y

y
)
 pi ai
n i 1
R2  1
(6)
 ( y pi  y ai )
n
2
i 1
n
2
 ( y pi  y m )
i 1
(7)
88
In these equations, n is the data set number, yai and ypi are the actual and the
predicted output values of the ith set, respectively, and ym is the actual output value
mean.
As depicted in Table 4.2, low MSE (0.28) and high R2 (0.97) prove the
agreement between the predicted and the actual values of the experimental data
(Antonio and Pilar., 2006). These data are also close to the related data from RSM.
However, slight differences signify that RSM performance was better than ANN in
the present work. Nevertheless, the results are generally similar and both models are
acceptable.
4.11 Process optimization
Numerical optimization was used to find the combination of all factors that
simultaneously fulfill the desired requirements. The upper and lower limits of each
variable (time and temperature of reaction, alcohol/oil molar ratio, and water content)
and its response were presented in surface plot. The maximum response was the
ultimate goal of this optimization. The optimal yield was obtained by employing GA
to optimize the input space of the ANN and RSM models (Antonio et al., 2006). The
maximum yields by RSM and ANN are listed in Table 4.6.
Table 4.6. Optimization using GA with RSM and ANN
Temperature Time
(°C)
(h)
42.13
15.73
RSM
42.52
16.81
ANN
Meth./oil molar
ratio (%)
4.67
4.38
Water content
% (w/w)
59.36
62.19
Yield
(%)
87.09
87.11
Error
(%)
0.01
0.01
89
CHAPTER 5
CONCLUSION AND RECOMMENDATION
5.1 Conclusion
Biocatalystic transesterification reaction was carried out with four important
variables (temperature, time, methanol/oil ratio and water content) to produce FAME
from Jatropha oil through the batch process. Initially, the lipase which utilized in this
research immobilized using an improved preparation technique of PVA-Alginate
matrix. Also an experimental work was successfully designed.
Subsequently, the process was modeled and optimized by employing RSM, ANN
and GA to catch the optimum parameters which leads to optimum yield of response
(biodiesel yield). Furthermore, the models proposed in RSM analyses were found to
be adequate and significant in predicting the FAME yield.
According to data from ANOVA table, time is an important variable in this
optimization and has a large and positive influence on FAME yield. After time,
temperature is the most effective variable and it has positive effect on the FAME
90
yield. The linear influence of this factor has significant negative effect on the FAME
yield, but its value is smaller than the positive quadratic effect. The ratio of
methanol/oil has a minor quadratic positive effect on the biodiesel yield. Also water
content has a large negative linear effect on the final response.
The optimization of four operating parameters led to a maximum FAME yield of
87.10% in the experiment. The maximum FAME yield from RSM and ANN were
86.62 and 87.01, respectively, under the reaction condition of 40 °C for 17 h, with
methanol/oil ratio of 5% and water content equal to 70. Both numerical and
experimental results fitted-well, indicating that the RSM and ANN models provided
good alternatives to cumbersome laboratory testing. The ANN and RSM-based GA
model provided optimum conditions for manufacturing biodiesel for a more efficient
process.
Results proved that immobilized Rhizopus oryzae lipase could be considered as
the suitable catalyst for methanolysis of oils with high amount of FFA under ambient
conditions. Compared to Novozym 435 which is the commercial biocatalyst,
Rhizopus oryzae shows faster performance in the same conditions.
5.2 Recommendation
Biocatalystic transesterification reaction is a relatively new technology in
biodiesel production. Hence, there are many aspects of the technology that need
to further develop in the near future to improve and industrialize the process.
Following are some recommendations for future works.
91
1) Currently, there are limited literatures on continuous biocatalystic reaction in
biodiesel production and most of the reported studies were conducted in batch
type reactor. Thus, it is beneficial to investigate the potential of continuous
process and it is vital for commercializing biodiesel manufacturing process
via biocatalysts such as enzymes.
2) Study on different types of feedstock and also several kind of enzyme can
help to find less cost materials and commercialize the process.
3) In term of catalyst preparation, the preparation of PVA to dissolve in water
would take around 1 hour. Larger amount of PVA would require a longer
time to completely dissolve in water. Thus, by some methods such as
employing microwave technique, the time can be reduced.
92
REFRENCES
Abigor, R. D., Uaudia, P. O., Foglia, T. A., Hass, M. J., Jones, K. C., Okpefa, E.,
Obibuzor J. U., Bafor M. E. (2000). Lipase-catalysed production of biodiesel
fuel from some Nigerian lauric oils. Biocheical Society Transaction 28, 979–
981.
Achten, W. M. J., Verchot, L., Franken, Y. J., Mathijs, E., Singh, V. P., Aerts, R.,
Muys, B., (2008). Jatropha bio-diesel production and use, Biomass and
Bioenergy 32, 1063–1084.
Ahmadi, M., Vahabzadeh, F., Bonakdarpour, B., Mofarrah, E., Mehranian, M.,
(2005). Application of the central composite design and response surface
methodology to the advanced treatment of olive oil processing wastewater
using Fenton’s peroxidation. Journal of Hazardous Materials 123, 187–195.
Angina, S. and Ram, P. (2000). Triglycerides-based diesel fuels. Renewable and
sustainable energy Review, 4, 111-133.
Ani, I., Nor Azimah, M. Z., Mohd, S. S., (2008). Immobilization of Baker’s yeast
invertase in PVA–alginate matrix using innovative immobilization technique.
Process Biochemistry. 43, 331–338.
Antonio, J. Y. and Pilar, D. M., (2006). A Neural Network approach to simulate
biodiesel production from waste olive oil. Energy & Fuels 20, 399-402.
Atadashi, I. M., Aroua, M. K., Abdul-Aziz, A., (2011). Biodiesel separation and
purification: A review. Renewable Energy 36, 437-443.
Bajpai, S. K. and Sharma, S. (2004). Investigation of swelling/degradation behavior
of alginate beads crosslinked with Ca2+ and Ba2+ ions. Reactive and
Functional Polymers 59, 129-140.
93
Bajaj, A., Lohan, P., Jha, P.N., Mehrotra, R., (2010). Biodiesel production through
lipase catalyzed transesterification: An overview. Journal of Molecular
Catalyst B: Enzyme 62, 9–14.
Bako, K. B., Kova, F. C. S., Gubicza, L., Hansco, J. K., (2002). Enzymatic biodiesel
production from sunflower oil by Candida antarctica lipase in a solvent free
system. Biocatalysis and Biotransformation 20, 437–439.
Ban, K., Kaieda, M., Matsumoto, T., Kondo, A., Fukuda, H., (2001). Whole-cell
biocatalyst for Biodiesel fuel production utilizing Rhizopus oryzae cells
immobilized within biomass support particles. Biochemical Engineering 8,
39–43.
Bas, D., Boyac, I.H., (2007). Modeling and optimization I: usability of response
surface methodology, Journal of Food Engineering 78, 836–845.
Basri, M., Rahman, R. N., Ebrahimpour, A., Salleh, A. B., Gunawan, E. R., Rahman,
M. B., (2007). Comparison of estimation capabilities of response surface
methodology (RSM) with artificial neural network (ANN) in lipase-catalyzed
synthesis of palm-based wax ester. BMC Biotechnology 30, 7-53.
Berriosa, M., Skeltonb, R. L., (2008). Comparison of purification methods for
biodiesel. Chemical Engineering Journal 144, 459–465.
Bezerra, M. A., Santelli, R. E., Oliveira, E. P., Villar, L. S., Escaleira, L. A., (2008).
Response surface methodology (RSM) as a tool for optimization in analytical
chemistry, Talanta 76, 965–977.
Bionas biofuels: the next generation sustainable energy. Properties of jatropha oil.
Available from, http://www.biofuelbionas.com/fuel.html 2010 .
Boucher, M. B., Weed, C., Leadbeater, N. E., Wilhite, B. A., Stuart, J. D., Parnas, R.
S., (2009). Pilot scale two phase continuous flow biodiesel production via
novel laminar flow reactor-separator. Energy Fuels. 23, 2750-2756.
94
Cayli, G. and Kusefoglu, S., (2007). Increased yield in biodiesel production from
used cooking oil by a two step process: comparison with one step process by
using TGA. Fuel processing Technology 98:118-122.
Chen, J. W., Wu, W. T., (2003). Regeneration of immobilized Candida antarctica
lipase for transesterification. Journal of Bioscience Bioengineering 95, 466–
469.
Chin, L. H., Hameed, B. H., Ahmad, A. L., (2009). Process optimization for
biodiesel production from waste cooking palm oil (Elaeis guineensis) using
response surface methodology. Energy & Fuels 23, 1040–1044.
Dalla-Vecchia, R., Sebrao, D., Nascimento, M. G. and Soldi, V. (2005).
Carboxymethylcellulose and poly (vinyl Alcohol) used as a film support for
lipase immobilization. Process Biochemistry 40, 2677-2682.
Danisman, T., Tana, S., Kacarb, Y., Ergenea, A., (2004). Covalent immobilization of
invertase on microporous pHEMA-GMA membrane, Food Chemistry 85,
461-466.
Demirbas, A., (2009). Political, economic and environmental impacts of biofuels: A
review, Applied Energy, 86, S108–S117.
D’Ippolito, S.A., Yori, J.C., Iturria, M.E., Pieck, C.L., Vera, C.R., (2007). Analysis
of a two step, noncatalytic, supercritical biodiesel production process with
heat recovery. Energy Fuels 21, 339–346.
Du, W., Xu, Y., Liu, D, Zeng, J., (2004). Comparative study on lipase catalyzed
transformation of soyabeen oil for biodiesel production with acyl acceptors.
Journal of Moecular Catalyst B: Enzyme 30, 125–129.
Edem,
D.
O., (2002).
Palm
oil:
biochemical,
physiological,
nutritional,
hematological, and toxicological aspects: a review, Kluwer Academic
Publishers 57 319–341.
95
Emregul, E., Sungur, S., Akbulut, U., (2005). Polyacrylamide- glatine carrier system
used for invertase immobilization. Food Chemistry 97(44),591-597.
Enweremadu, C. C., Mbarawa, M. M., (2009). Technical aspects of production and
analysis of biodiesel from used cooking oil—A review. Renewable and
Sustainable Energy Review 13, 2205–2224.
Fermoso, J., Gil, M. V., Arias, B., Plaza, M. G., Pevida, C., Pis, J. J., Rubiera,
F.,(2009). Application of response surface methodology to assess the
combined effect of operating variables on high-pressure coal gasification for
H2-rich gas production. International Journal of Hydrogen Energy 35, 1191–
1204.
Fukuda, H., Kondo, A.,( 2003). Patent No: US 6524839 B1.
Fukuda, H., Noda, H.,(2006). Patent No: US 6982155 B1.
Gerpen, J.V., (2005). Biodiesel processing and production. Fuel process Technology,
86, 1097-1107.
Gubitz, G. M., Mittelbach, M., Trabi, M., (1999). Exploitation of tropical oil seed
plant Jatropha curcas L. Bioresource Technology 67, 73–82.
Hama, S., Yamaji, H., Kaieda, M., Oda, M., Kondo, A., Fukuda, H., (2004). Effect of
fatty acid membrane composition on whole cell biocatalysts for biodiesel-fuel
production. Biochemical Engineering Journal 21, 155– 160.
Hama, S., Yamaji H., Fukumizu, T., Tamalampudi, S., Kondo, A., Miura, K.,
Fukuda, H., (2006). Lipase localization in Rhizopus oryzae cells immobilized
within biomass support particles for use as whole cell biocatalysts in biodiesel
fuel production. Journal of Bioscience and Bioengineering 101, 328– 333.
Hama, S., Yamaji, H., Fukumizu, T., Numata, T., Tamalampudi, S., Kondo, A.,
Nodac, H., Fukuda, H., (2007). Biodiesel fuel production in a packed-bed
reactor using lipase-producing Rhizopus oryzae cells immobilized within
biomass support particles. Biochemical Engineering Journal 34, 273–278.
96
Hanny, J. B., Shizuko, H., (2008) Biodiesel production from crude Jatropha curcas L.
seed oil with a high content of free fatty acids. Bioresource Technology 99
1716–1721
Hawash, S., Kamal, N., Zaher, F., Kenawi, O., El Dewani, G., (2009). Biodiesel fuel
from Jatropha oil via non-catalytic supercritical methanol transesterification.
Fuel, 88(3), 579-582.
Hossein, M., Keat, T. L., Subhash, B., Abdul, R. M. (2010). Subcritical water
liquefaction of oil palm fruit press fiber in the presence of sodium hydroxide:
An optimisation study using response surface methodology. Bioresource
Technology 101, 9335–9341.
Imai, K., Shiomi, T., Uchida, K. and Miya, M. (1985). Immobilization of enzyme
into poly (vinil alcohol) membrane. Biotechnology and Bioengineering 28,
1721-1726.
Iso, M., Chem, B., Eguchi, M., Kudo, T., Shrestha, S., (2001). Production of
Biodiesel from triglycerides and ethanol using immobilized enzyme. Journal
of Molecular Catalyst B: Enzyme 16, 53–58.
Jacobson, K., Gopinath, R., Meher, L. C., Dalai, A. K., (2008). Solid acid catalyzed
biodiesel production from waste cooking oil. Application of Catalyst, B. 85,
86-91.
Jain, S., Sharma, M.P., (2010)a. Prospects of biodiesel from Jatropha in India
Renewable and Sustainable Energy Review 14, 763–771.
Jain, S., Sharma, M. P., (2010)b. Kinetics of acid base catalyzed transesterification of
Jatropha curcas oil. Bioresource Technology 101, 7701–7706.
Jayed, M.H., Masjuki, H.H., Saidur, R., Kalam, M.A., Jahirul, M.I. (2009).
Environmental aspects and challenges of oilseed produced biodiesel in
Southeast Asia. Renewable and Sustainable Energy Review 13, 2452–2462.
97
Jianlong, W., Wei, W., (2009). Optimization of fermentative hydrogen production
process using genetic algorithm based on neural network and response
surface methodology. international journal of hydrogen energy 34 ,255–261.
Jitputti, J., Kitiyanan, B., Rangsunvigit, P., Bunyakiat, K., Attanatho, L. and
Jenvanitpanjakul, P. (2006). Transesterification of crude palm kenel oil and
crude coconut oil by different solid catalysts. Chemical Engineering Journal
116: 61-66.
Joon, C., Juan, A., Damayani, A. K. A., Ta, Y. W., Taufiq, Y. Y. H., (2011).
Biodiesel production from jatropha oil by catalytic and non-catalytic
approaches. Bioresource Technology 102, 452–460
Jones, N., Miller, JH., (1991). Jatropha curcas – a Multipurpose species for
problematic sites. Land Resources series 1: 40-43.
Joubert, P. H., Brown, J. M. M., Hay, I. T., Sebata, P. D. B. (1984). Acute poisoning
with Jatropha curcas (purging nut tree) in children. S. Afr. Med J., 65:729730.
Juan, J.C., (2011). Biodiesel production from jatropha oil by catalytic and noncatalytic approaches: An overview, Bioresource Technology Journal 102,
452–460.
Kaieda, M., Samukawa, T., Matsumoto, T., Ban, K., Kondo, A., Shimada, Y., Noda,
H., Nomoto, F., Ohtsuka, K., Izumoto, E. & Fukuda H. (1999). Biodiesel fuel
production from plant oil catalyzed by Rhizopus oryzae lipase in a watercontaining system without an organic solvent. Journal of Bioscience and
Bioengineering, 88, 6, 627-631.
Kai-Pei, Z., Jing-Qi, L., Ze-Lin, H., Zhen Y., (2010). Penicillium expansum lipasecatalyzed production of biodiesel in ionic liquids. Bioresource Technology.
S0960-8524(10)01860-2
98
Korbahti, B. K., Rauf, M. A., (2008). Response surface methodology (RSM) analysis
of photoinduced decoloration of Toludine Blue. Chemical Engineering
Journal; 136, 1–76.
Khoo, K. M. and Ting, Y. P. (2001). Biosorption of gold by immobilized fungal
biomass. Biochemical Engineering Journal 8:51-59.
Kulkarni, M. G., Dalai, A. K., (2006). Waste cooking oil – an economical source for
biodiesel: a review of Industrial Engineering and Chemical Resource 45,
2901–2913.
Knothe, G., (2005). Dependence of biodiesel fuel properties on the structure of fatty
acid alkyl esters. Fuel Process Technology 86, 1059–70.
Knothe, G. (2005). Viscosity of biodiesel in Knothe, G., J. V. Gerpen and J. Krahl.
The diodiesel Handbook (pg 81-82). Urbana, Illinois. AOCS Press.
Lee, H.V., Yunus, R., Juan, J.C. and Taufiq-Yap, Y.H. (2011). Process optimization
design for jatropha-based biodiesel production using response surface
methodology. Fuel Processing Technology 92 2420–2428.
Leuschner, F., (1966). German Patent NO. 1,227,855, German: German patent and
trademark office.
Machavaram, R., Prakash, C. J., Hifjur, R., (2009). Prediction of optimized
pretreatment process parameters for biodiesel production using ANN and GA.
Fuel. 88, 868–875.
Leung, D. Y. C., Wu, X., Leung, M. K. H., (2010). A review on biodiesel production
using catalyzed transesterification. Applied Energy 87, 1083–1095.
Li, L., Du, W., Liu, D., Wang, L., Li, Z., (2006). Lipase-catalyzed transesterification
of rapeseed oils for biodiesel production with a novel organic solvent as the
reaction medium. Journal of Molecular Catalyst B: Enzyme 43, 58–62.
99
Linko, Y. Y. , Lamsa, M., Wu, X., Uosukainen, E., Seppala, J., Linko, P. (1998).
Biodegradable products by lipase biocatalysis. Journal of Biotechnology 66,
41–50.
Ma, F.
and
Hanna, M.A (1999). Biodiesel production: a review. Bioresource
Technology, 70, 1-15.
Mampane, J. K., Jobert, P. H., Hay I. T. (1987). Jatropha curcas: use as traditional
Tswana medicine and its role as a cause of acute poisoning. Phytotherapy
Research 1, 50-51.
Matsumoto, T., Takahashi, S., Kaieda, M., Ueda, M., Tanaka, A., Fukuda, H.,
Kondo, A., (2001). Yeast whole-cell biocatalyst constructed by intracellular
overproduction of Rhizopus oryzae lipase is applicable to biodiesel fuel
production. Applied Microbiolgy and Biotechnolgy 57, 4.
Mazaheri, H., Keat, T. L., Subhash, B., Abdul Rahman, M., (2010). Subcritical water
liquefaction of oil palm fruit press fiber in the presence of sodium hydroxide:
An optimisation study using response surface methodology. Bioresource
Technology 101, 9335–9341.
Michael, S., Cao, F., Chen, Y., Zhai, F., Li, J., Wang, J., (2008). Biodiesel
production from high acid value waste frying oil catalyzed by superacid
heteropolyacid.
Biotechnology
and
Bioengineering
101:93-100.
PMID:18646228
Miller, G. L. (1959). Use of dinitrosalicylic acid reagent for determination of
reducing sugar. Analytical Chemistry 31, 426-428.
Mittelbach, M. (1990). Lipase catalyzed alcoholysis of sunflower oil. Journal of
American Oil Chemical Society 67, 168–170.
Modi, M.K., Reddy, J.R.C., Rao, B.V.S.K., Prasad, R.B.N., (2006). Lipase mediated
transformation of vegetable oils into biodiesel using propane-2-ol as acyl
acceptor. Biotechnology Letters 28, 637–640.
100
Myers, R. H., Montgomery, D. C. (2000). Response surface methodology: process
and product optimization using designed experiments. Second edition of John
Wiley & Sons, USA.
Nelson, L.A., Foglia, T.A., Marmer, W.N., (1996). Lipase-catalyzed production of
biodiesel. Journal of American Oil Chemical Society 73 (8), 1191–1195.
Nie, K., Xie, F., Wang, F., Tan, T., (2006). Lipase catalyzed methanolysis to produce
biodiesel: Optimization of the biodiesel production. Journal of Molecular
Catalysis B: Enzyme 43, 142–147.
Nor-Azimah, M. Z., Mohd, S. S., Ani, I., (2011). Development and modification of
PVA–alginate as a suitable immobilization matrix. Process Biochemistry 46,
2122–2129.
Noureddini, H., Gao, X., Philkana, R. S. (2005). Immobilized Pseudomonas cepacia
lipase for biodiesel fuel production from soybean oil. Bioresource
Technology 96, 769–777.
Ozer, A., Gurbuz, G., Calimli, A., Korbahti, B. K., (2009). Biosorption of copper (II)
ions
on
Enteromorpha
prolifera:
Application
of
response
surface
methodology (RSM). Chemical Engineering Journal 146(3), 377-387.
Palligarnai, T., Vasudevan, B. F.,(2010). Environmentally Sustainable Biofuels:
Advances in Biodiesel Research. Waste Biomass Valor, 1:47–63
Pillay, V., Sibbanda, W. and Danckwerts, M. P. (2005). Sequential design of a novel
PVA-based crosslinked ethylenic homoipolymer for extended drug delivery.
International Journal of Pharmaceutices. 301, 89-101.
Pinzi, S., Garcia, I. L., Lopez-Gimenez, F. J., Luque de Castro, M. D., Dorado, G.,
Dorado, M. P., (2009). The ideal vegetable oil-based biodiesel composition: a
review of social, economical and technical impacts. Energy Fuels 23, 2325–
2341.
PORIM, Determination of Saponification value, PORIM Test Methods, Ministry of
Primary Industries Malaysia, 1995, pp. 69–71.
101
PORIM, Determination of acidity, PORIM Test Methods, Ministry of Primary
Industries Malaysia, 1995, pp. 40–42.
Ranganathan, S. V., Narasimhan, S. L.,Muthukumar, K.(2008). An overview of
enzymatic production of biodiesel. Bioresource Technology 99, 3975–3981.
Rathore, V. and Madras, G. (2007). Synthesis of biodiesel from edible and nonedible oils in supercritical alcohols and enzymatic synthesis in supercritical
carbon dioxide. Fuel 86, 2650–2659.
Robles-Medina, A., Gonzales-Moreno, P.A., Esteban-Cerdan, L., Molina-Grima, E.,
(2009). Biocatalysis: towards ever greener biodiesel production. Biotechnolgy
Advance 27, 398–408.
Russbueldt, B. M. E., Hoelderich, W. F. (2009). New sulfonic acid ion-exchange
resins for the pre-esterification of different oils and fats with high content of
free fatty acids. Applied Catalysis A: General 362, 47–57.
Royon, D., Daz, M., Ellenrieder, G., Locatelli, S. (2007), Enzymatic production of
biodiesel from cotton seed oil using t-butanol as a solvent. Bioresource
Technology 98, 648–653.
Salamatinia, B., Mootabadi, H., Bhatia, S., Abdullah, A. Z., (2010). Optimization of
ultrasonic assisted heterogeneous biodiesel production from palm oil: a
response surface methodology approach, Fuel Processing Technology
91,441–448.
Samukawa, T., Kaieda, M., Matsumoto, T., Ban, K., Kondo, A., Shimada, Y., Noda,
H., Fukuda, H., (2000). Pretreatmenr of immobilized Candida antactica lipase
for biodiesel fuel production from plant oil. Journal of
Biosociety and
Bioengineers 90:180-183.
Sharma, Y. C., Singh, B., Upadhyay, S. N (2008). Advancements in development
and characterization of biodiesel a review, Fuel 87, 2355-2377.
Shimada, Y., Watanabe, Y., Samukawa, T., Sugibara, A., Noda, I.L., Fukuda, I.-I.,
Tominaga, Y., (1999). Conversion of vegetable oil to biodiesel using
102
immobilized Candida antarctica lipase. Journal of American Oil Chemical
Society 76, 789–793.
Shimada, Y., Watanabe, Y., Sugihara, A., Tominaga, Y., (2002). Enzymatic
alcoholysis for biodiesel fuel production and application of the reaction oil
processing. Journal of Molecular Catalyst B: Enzyme 17, 133–142.
Srivathsan, V. R., Srinivasan, L. N., Karuppan, M. (2008). An overview of
enzymatic production of biodiesel. Bioresource Technology 99, 3975–3981.
Sheedlo, M., A review of the processes of biodiesel production (2008). MMG 445
Basic Biotechnology ejournal, 4, 61–65.
Sriappareddy, T., Mahabubur, R. T., Shinji, H., Takao, N., Akihiko, K., Hideki, F.,
(2008). Enzymatic production of biodiesel from Jatropha oil: A comparative
study of immobilized-whole cell and commercial lipases as a biocatalyst.
Biochemical Engineering Journal 39, 185–189.
Szczesna-Antczak, M. and Galas, E. (2001). Bacillus subtilis cells immobilised in
PVA-cryogels. Biomolecular Engineering. 17:55-63.
Staubamann, R., Ncube, I., Gubitz, G. M., Steiner, W., Read, J. S. (1999). Esterase
and lipase activity in Jatropha curcas L. seeds. Journal of Biotechnology 75,
117-126.
Vicente, G., Martinez, M., Aracil, J. (2004). Integrated biodiesel production: a
comparison of different homogeneous catalysts systems. Bioresource
Technology 92, 297–305
Watanabe, Y., Shimada, Y., Sugihara, A., Tominaga, Y, (2002). Conversion of
degummed soybean oil to biodiesel fuel with immobilized Candida antarctica
lipase. Journal of Molecular Catalyst B Enzyme 17, 151–5.
Wang, J. L., Li, M. H., Lin, J. P., Wei, D.Z. (2010). Optimization of immobilization
for selective oxidation of benzyl alcohol by Gluconobacter oxydans using
response surface methodology. Bioresource Technology. 101(23), 8936-8941.
103
Xu, Y., Du, W., Zeng, J., Liu, D., (2004). Conversion of soyabeen oil to biodiesel
fuel using lipozyme TL IM in a solvent free medium. Biocatalyst and
Biotransformation 22, 45–48.
Yasin, Y., (2011). Biodiesel production from pomace oil by using lipase immobilized
onto olive pomace, Bioresource Technology 102, 3977–3980.
Yuan X, Liu J, Zeng G, Shi J, Tong J, Huang G.(2008) Optimization of conversion
of waste rapeseed oil with high FFA to biodiesel using response surface
methodology. Renewable Energy; 33; 1678–1684.
Zullaikah, S., Lai, C., C., Valui, S. R. and Ju, Y. H. (2005). A two step acid catalyzed
process for the production of biodiesel from rice bran oil. Biorisource
Technology 96:1889-1896.
104
APPENDIX A
CALCULATION OF ENZYME ACTIVITY
Biodiesel concentration produced after 15 hr of experiment= 860 g/lit
Molecular weight of jatropha oil= 870
One international unit of activity was defined as 0.1 µmol of transesterified jatropha
oil per minute under the assay conditions
Enzyme
activity 
860( g / L)
870( g / mol)
=0.9885 mol/L
=98.85×104 µ mol/L
=98.85×104/900 IU
=1098.3 IU
105
APPENDIX B
CALCULATION OF BIODIESEL YIELD
Weight of biodiesel = 1.411
Weight of jatropha oil = 1.835
FAME area = 1.13
Yield % 
( FAME area from GCMS  weight of product )
( weight of oil sample)
Yield % 
(1.13 1.414)
100
1.834
= 87.1 %
 100%
106
APPENDIX C
PUBLICATIONS
o Alireza Zarei, Nor Aishah Saidina Amin, Nor Azimah Mohd Zain, Amin
Talebian Kiakalieh, Iman Noshadi. Optimization of green production of
biodiesel from jatropha oil using Response Surface Methodology, Artificial
Neural Network and Genetic Algorithm. Bioresourse Technology, (under
review)
o Alireza Zarei , Nor Aishah Saidina Amin, Amin Talebian Kiakalaie and
Hamidreza Jalilian. Biodiesel production with immobilized enzyme and
effect of alcohol on the results. International Conference on Environment,
Energy and Biotechnology – ICEEB 2012.
o Alireza Zarei , Nor Aishah Saidina Amin, Hamidreza Jalilian and Amin
Talebian Kiakalaie. Short review on common methods of biodiesel
production. 3rd International Conference on Engineering and ICT
(ICEI2012).
o Amin Talebian Kiakalaieh, Nor Aishah Saidina Amin, Alireza Zarei, Iman
Noshadi. Transesterification of waste cooking oil by heteropoly acid (HPA)
catalyst: optimization and kinetic model. Applied Energy. (under review)
107
o Hamidreza Jaliliannosrati1, Alireza Zarei, and Amin Talebian Kiakalaieh. A
future view of renewable energy in Malaysia. 3rd International Conference on
Engineering and ICT (ICEI2012).
o Amin Talebian Kiakalaieh1, Nor Aishah Saidina Amin and Alireza Zarei.
Biodiesel production from high free fatty acid waste cooking oil by solid acid
catalyst.
International
Conference
on
Environment,
Energy
and
Biotechnology – ICEEB 2012.
o Hamidreza Jaliliannosrati, Amin Talebian Kiakalaieh and Alireza Zarei.
Comparison of Bioethanol and Biodiesel Feedstock with Futuristic Look at
Biofuel.
International
Conference
on
Environment,
Energy
and
Biotechnology – ICEEB 2012.
o Amin T. Kiakalaieh1, Nor A.S. Amin, Hamidreza Jaliliannosrati and Alireza
Zarei . The effect of heterogeneous catalysts in biodiesel production: A short
review. 3rd International Conference on Engineering and ICT (ICEI2012) .