Microbial Oil Production from Oil Palm Empty Fruit Bunch A thesis by publication submitted in fulfilment of the requirements for the degree of Doctor of Philosophy (PhD) Farah Binti Ahmad School of Chemical, Physics and Mechanical Engineering Science and Engineering Faculty Queensland University of Technology 2016
212
Embed
Microbial Oil Production from Oil Palm Empty Fruit Bunchiii frond) showed that oil palm biomass had the potential to increase the total palm oil production by 25%, at a cheaper feedstock
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Microbial Oil Production from Oil
Palm Empty Fruit Bunch
A thesis by publication submitted in fulfilment of the requirements for the degree of
Doctor of Philosophy (PhD)
Farah Binti Ahmad
School of Chemical, Physics and Mechanical Engineering
Science and Engineering Faculty
Queensland University of Technology
2016
i
Keywords
Accelerated Solvent Extraction
Bagasse
Biodiesel
Biofuel
Bioreactor
Empty fruit bunch
Fungi
Glucose
Glycerol
Lignocellulose
Lipid
Microalgae
Microbial oil
Multi-criteria analysis
Oil palm
Oleaginous microorganism
Sugarcane
Sustainability
Techno-economic
Xylose
Yeast
ii
Abstract
Oil palm empty fruit bunch (EFB) is one of the major solid wastes from palm
oil processing. However, EFB is not effectively reused and current practices for EFB
disposal may lead to environmental problems. EFB is a lignocellulosic biomass and
has the potential to be converted into oil through biochemical routes, where the oils
can be further used for biodiesel production. This research aimed to develop and
optimise a process for microbial oil production from EFB. The microbial oil
production process involves cultivation by oleaginous microorganisms on carbon
substrates. Oleaginous microorganisms (including microalgae, yeasts and fungi)
accumulate oils from carbon substrates under nitrogen-limiting conditions.
The research project identified six potential oleaginous microorganisms which
were Chlorella protothecoides and Chlorella zofingiensis (microalgae),
Cryptococcus albidus and Rhodotorula mucilaginosa (yeasts), and Aspergillus
oryzae and Mucor plumbeus (fungi). The microorganisms were cultivated on
glucose, xylose and glycerol to down-select the most suitable microorganisms for oil
production from lignocellulosic hydrolysates. The selection was performed through
multi-criteria analysis (MCA) approach using analytic hierarchy process (AHP) and
preference ranking organisation method for the enrichment of evaluations
(PROMETHEE) with graphical analysis for interactive aid (GAIA). Based on MCA,
the potential microorganisms were down-selected to the three highest ranking
microorganisms which were A. oryzae, M. plumbeus and R. mucilaginosa.
To evaluate microbial oil production from EFB, the three highest ranking
microorganisms were cultivated on EFB hydrolysates. EFB was first subjected to
dilute acid pretreatment followed by enzymatic hydrolysis of the solid residue. Two
types of feedstocks were used for the cultivation, which are EFB hydrolysates from
the pretreatment (detoxified by overliming) and enzymatic hydrolysis of solid
residue. The cultivation on EFB hydrolysates resulted in the highest oil
concentrations and oil yields by M. plumbeus. The fuel properties analysis showed
that the oils produced were suitable for biodiesel production. Techno-economic
evaluation of oil production from EFB and other oil palm biomasses (trunk and
iii
frond) showed that oil palm biomass had the potential to increase the total palm oil
production by 25%, at a cheaper feedstock cost.
To optimise oil production from EFB by M. plumbeus, response-surface
methodology was used to evaluate and optimised the parameters of cultivation (i.e.,
sugars concentration, yeast extract concentration, pH and spore concentration) based
on the oil concentration and oil yield. The optimum conditions for oil yield were
identified and applied to the cultivation on EFB hydrolysate in a 1 L bioreactor.
Cultivation in the bioreactor resulted in ~2 times higher oil yield in comparison to
shake-flask cultivation, likely as a result of improved mixing and oxygen-mass
transfer in the bioreactor.
Overall, this study demonstrated that EFB is a promising low-cost raw
material for oil production by M. plumbeus. The microbial oils from EFB can be
used for the production of biodiesel from non-food feedstock. The integration of
microbial oil production from oil palm biomass with existing palm oil processing
could enhance the profitability and sustainability of the palm oil industry.
Selecting oleaginous microorganisms for microbial oil production ............... 46
3.1 A multi-criteria analysis approach for ranking and selection of microorganisms for the production of oils for biodiesel production .................................................... 46
Evaluating microbial oil production from EFB ................................................ 79
4.1 Evaluation of oil production from oil palm empty fruit bunch by oleaginous microorganisms .......................................................................................................... 79
Optimising microbial oil production from EFB ............................................. 113
5.1 Improved microbial oil production from oil palm empty fruit bunch by Mucor plumbeus .................................................................................................................. 113
vi
5.1.3 Results and Discussion .......................................................................... 122
Appendices .................................................................................................................... I
Appendix 1: Optimising microbial oil extraction by Accelerated Solvent ExtractionIII
Appendix 2: Microbial oil production from sugarcane bagasse hydrolysates by oleaginous yeast and filamentous fungi ................................................................. XXII
Appendix 3: Time-course graph for DO, pH, agitation speed and aeration rate for bioreactor cultivation ....................................................................................... XXXVIII
vii
List of Figures
Figure 1-1 Overall research plan .................................................................................. 3
Figure 1-2 Overview of thesis ...................................................................................... 6
Figure 2-1 (a) Average oil yield per ha per year for selected oil crops ..................... 10
Figure 2-2 Oil palm tree, fresh fruit bunch and longitudinal section of fresh oil palm fruit ...................................................................................................... 11
Figure 2-3 EFB from palm oil mills ........................................................................... 12
Figure 2-4 General process for conversion of EFB into microbial oil ....................... 14
Figure 2-5 Lignocellulosic structure of plants. .......................................................... 15
Figure 2-6 A culture that contains filamentous fungi in pellet form. ......................... 28
Figure 3-1 Criteria hierarchy for the evaluation of microorganisms for microbial oil production. Cluster 1 criteria were evaluated based on cultivation results on glucose (G), xylose (X) and glycerol (L). .................. 56
Figure 3-2 (a) Biomass concentration, (b) oil content and (c) oil concentrations for growth of six microorganisms on glucose, xylose and glycerol. ........... 61
Figure 3-3 Consumption of (a) glucose, (b) xylose and (c) glycerol over 168 h of cultivation. ............................................................................................... 63
Figure 3-4 (a) PROMETHEE I partial ranking of alternatives and (b) PROMETHEE II complete ranking where RM denotes R. mucilaginosa, AO A. oryzae, MP M. plumbeus, CP C. protothecoides, CA C. albidus and CZ C. zofingiensis. ......................................................... 67
Figure 3-5 GAIA plane at (a) 100% zoom and (b) 400% zoom ................................ 69
Figure 3-6 GAIA Webs for top three alternatives from PROMETHEE which are (a) R. mucilaginosa, (b) A. oryzae and (c) M. plumbeus. Criterion C2-X is not visible due to overlapping by C6-X. ......................................... 73
Figure 4-1 Flow chart of hydrolysates preparation from EFB for microbial cultivation. ................................................................................................... 86
Figure 4-2 Sugars consumption of R. mucilaginosa, A. oryzae and M. plumbeus on EFB liquid (EFBLH) and enzymatic (EFBEH) hydrolysates. ................ 94
Figure 4-3(a) Oil contents (%, w/w) and (b) oil concentrations (g/L) of yeast R. mucilaginosa, and fungi A. oryzae and M. plumbeus cultivated on EFB liquid (EFBLH) and enzymatic (EFBEH) hydrolysates. ..................... 96
Figure 4-4 Process flow of proposed integration of microbial oil production from oil palm biomasses into the existing palm oil processes. .................. 105
Figure 5-1 (a) Mathematical models for oil concentration (Y1) and oil yield (Y2), and the parameters of analysis of variance (ANOVA) of each model. (b-c) Plots of predicted vs. actual values (experimental data). (d-e) Internally studentised residuals vs actual values. ............................. 126
viii
Figure 5-2 (a-e) Three-dimensional surface plots of binary interaction between different variables to the oil concentration. Sugar is sugar concentration, %YE is relative concentration of yeast extract, Spore is spore concentration and pH is initial pH. ................................................... 129
Figure 5-3 (a-d) Three-dimensional surface plots of binary interaction between different variables to the oil yield. Sugar is sugar concentration, %YE is relative concentration of yeast extract, Spore is spore concentration and pH is initial pH. ................................................................................... 130
Figure 5-4 The consumption of glucose (a) and xylose (b) for the experimental run .............................................................................................................. 133
Figure 5-5 (a) The consumption of glucose and xylose for bioreactor cultivation ................................................................................................... 138
Figure 5-6 Fungal morphology from the bioreactor cultivation .............................. 139
ix
List of Tables
Table 2-1 Various agro-industrial hydrolysates as the feedstock for oleaginous yeasts cultivation for oil production ............................................................ 25
Table 2-2 Various agro-industrial hydrolysates as the feedstock for oleaginous fungi cultivation for oil production .............................................................. 27
Table 3-1 Pairwise comparison matrix with respect to goal for criteria groups of Cluster 1 and Cluster 2. ........................................................................... 57
Table 3-2 Pairwise comparison matrix with respect to goal for criteria of Cluster 1. ...................................................................................................... 57
Table 3-3 Fatty acid compositions of oil extracted from six different microorganisms grown on various carbon substrates. ................................. 66
Table 3-4 Weight stability intervals for criteria with relative weight>5%. ............... 70
Table 4-1 Chemical compositions of raw and pretreated EFB. The composition of lignin was based on the composition of acid soluble and acid insoluble lignin. ............................................................................................ 91
Table 4-2 Sugars (glucose, xylose and arabinose), organic acids (formic acid, acetic acid and levulinic acid) and furans (5-hydroxymethylfurfural (HMF) and furfural) compositions of liquid (EFBLH) and enzymatic (solid residue) (EFBEH) hydrolysates of EFB. ........................................... 92
Table 4-3 Biomass concentrations, oil concentrations and oil yields of different oleaginous yeasts and filamentous fungi from batch fermentation ............. 98
Table 4-4 Fatty acids composition of microbial oils methyl ester of R. mucilaginosa, A. oryzae and M. plumbeus cultivated on EFB liquid (EFBLH) and solid residue enzymatic (EFBEH) hydrolysates, as well as fuel properties, ....................................................................................... 100
Table 4-5 (a) The summary of technical evaluation of microbial oil production from oil palm biomasses (EFB, trunk (OPT) and frond (OPF)) through the comparison of potential microbial oil yields per hectare to oil yield of crude palm oil. ....................................................................................... 102
Table 5-1 The coded and actual values of each variable and its levels for the experimental design ................................................................................... 120
Table 5-2 Concentrations of sugars, organic acids and 5-hydroxylmethylfurfural (HMF) in enzymatic hydrolysates (EHs). ........... 123
Table 5-3 Analysis of variance (ANOVA) for the response surface quadratic model of oil concentration (a) and oil yield (b) that had significant terms (Sugar concentration - X1, Relative concentration of yeast extract - X2, Spore concentration - X3, Initial pH - X4). .......................... 126
Table 5-4 The predicted responses from the simulation of mathematical models ... 127
Table 5-5 The concentration of ethanol accumulated at the end of the cultivation .................................................................................................. 135
x
Table 5-6 Results of different cultivation performed in this study .......................... 140
Table 5-7 The comparison of raw materials cost (RMC) (a), production cost (b) and cost of biodiesel (c) for the production of biodiesel from EFB and glucose ................................................................................................. 142
xi
List of Abbreviations
AHP Analytic hierarchy process
ANOVA Analysis of variance
ASE Accelerated solvent extraction
C/N Carbon-to-nitrogen
DO Dissolved oxygen
EFB Empty fruit bunch
EFBEH Empty fruit bunch enzymatic hydrolysate
EFBLH Empty fruit bunch liquid hydrolysate
EH Enzymatic hydrolysate
FFB Fresh fruit bunch
GAIA Graphical analysis for interactive aid
GC-MS Gas chromatography-mass spectrometry
HMF 5-hydroxymethyl furfural
HPLC High-performance liquid chromatography
MCA Multi-criteria analysis
MF Mesocarp fibre
OPF Oil palm frond
OPT Oil palm trunk
PDA Potato dextrose agar
PKS Palm kernel shell
PROMETHEE Preference ranking organization method for the enrichment of
evaluations
RSM Response surface methodology
YDP Yeast dextrose potato
xii
List of Publications
Peer-reviewed journal publications
1. Ahmad, F.B., Zhang, Z., Doherty, W.O.S., and O’Hara, I.M., A multi-criteria
analysis approach for ranking and selection of microorganisms for the
production of oils for biodiesel production. Bioresource Technology, 190
(2015), Pages 264-273.
2. Ahmad, F.B., Zhang, Z., Doherty, W.O.S., and O’Hara, I.M., Evaluation of
oil production from oil palm empty fruit bunch by oleaginous
microorganisms. Biofuels, Bioproducts and Biorefining, 10 (2016), Pages
production from sugarcane bagasse hydrolysates by oleaginous yeast and
filamentous fungi. Proceedings of the Australian Society of Sugar Cane
Technologists, 38 (2016).
Poster presentation
1. Ahmad, F., Zhang, Z., Doherty, W., and O’Hara, I.M., Microbial oil
production from palm oil empty fruit bunch hydrolysates. In 37th Symposium
on Biotechnology for Fuels and Chemicals, 27 - 30 April 2015, San Diego,
California.
xiii
Awards
1. Denis Foster Chemistry/Chemical Engineering award from 38th Conference
of Australian Society of Sugar Cane Technologists 2016 for best paper in
Chemistry/Chemical Engineering.
Miscellaneous
1. Cover Image, Volume 10, Issue 4 in Biofuels, Bioproducts and Biorefining,
10 (2016). Based on the Modeling and Analysis Evaluation of oil production
from oil palm empty fruit bunch by oleaginous microorganisms. Photo Credit:
Farah Binti Ahmad.
xiv
Statement of Original Authorship
The work contained in this thesis has not been previously submitted to meet
requirements for an award at this or any other higher education institution. To the
best of my knowledge and belief, the thesis contains no material previously
published or written by another person except where due reference is made.
Signature: QUT Verified Signature
Date: November 2016
xv
Acknowledgements
My deepest gratitude is to my supervisors Prof Ian O’Hara, Dr Jan Zhang and
Prof William Doherty, for their invaluable guidance and assistance throughout my
research project. I am grateful to have such dynamic supervisory team for this
research. I would like to express my gratitude to Malaysia Ministry of Higher
Education and International Islamic University of Malaysia (IIUM) for my
postgraduate scholarship.
I would like to thank Shane Russell and Dr Joel Herring from Central
Analytical Research Facility (CARF) (Institute for Future Environments), Wanda
Stolz and Dr Dani Tikel (Centre for Tropical Crops and Biocommodities (CTCB))
and Dr Chris Carvalho (School of Chemistry, Physics and Mechanical Engineering
(CPME)) for their assistance for analytical work undertaken in this research. I would
also like to thank the staff from CTCB labs, CARF labs and QUT Banyo Pilot Plant
Facility for their support while I was working in those facilities. I would like to
acknowledge A/Prof Junior Te’o from the School of Earth, Environmental and
Biological Sciences (EEBS) for the opportunity to perform bioreactor experiments,
as well as Vincent Chand (EEBS) for his support while working in Junior’s lab. I
would also like to thank Vitor Kawazoe for his assistance on oil extraction work
during his internship at CTCB. I thank Anthony Brinin (CTCB), and Ada Rosier and
Beric Nott from Faculty of Health for their assistance in microscopic work
undertaken in this study.
I would like to acknowledge CTCB and the school of CPME for the travel
funding support. I thank my fellow labmates from CARF labs and CTCB labs for
their support throughout my PhD journey. I would also like to thank Dr Christian
Long from Academic Language and Learning Services for his advice. I would also
like to acknowledge Prof Roger Hellens and A/Prof Robert Speight for feedback on
thesis.
Lastly, I am deeply grateful for the support, love and prayers from my parents,
family and friends, as well as the Sisters from QUT GP Musolla. All the praises and
thanks be to the Creator.
1
Introduction
1.1 Background
The palm oil industry is a major food and oleochemical industry with palm oil
being a key source of oils and fats. The scale of the palm oil industry, however, has
contributed to the generation of large amounts of solid and liquid wastes from palm
oil processing. Only a portion of the solid wastes, known as oil palm biomass, are
used for other applications, while poor management of oil palm biomass can result in
substantial environmental problems. Empty fruit bunch (EFB) is one of the major
solid wastes generated as a by-product of palm oil mills, and is currently not
effectively used.
EFB and other lignocellulosic biomass residues have the potential to be
converted to higher value products through microbial cultivation. The oils produced
can be used as feedstocks for the production of biodiesel, health products, food
products and chemicals. As palm oil is already extensively used in the oleochemical
industry, there is the opportunity for integrating oils produced from EFB in the
existing palm oil supply chain. The proposed integration could potentially enhance
the sustainability and profitability of palm oil production.
Carbohydrates (sugars) can be converted to oils through cultivation of
oleaginous microorganisms. Oleaginous microorganisms, including certain
microalgae, yeasts, fungi and bacteria, are capable of accumulating oil in
intracellular membranes in significant quantities, generally under stress cultivation
conditions. Fermentable sugars from biomass have the potential to be used as the
carbon substrates for oil production by oleaginous microorganisms. However, due to
the recalcitrant nature of biomass, processing is necessary to disrupt the complex
hemicellulose-lignin structure, so that cellulose is accessible for enzymatic
hydrolysis to break cellulose down to monomers (i.e., fermentable sugars).
Hydrolysates from both pretreatment and enzymatic hydrolysis of lignocellulosic
biomass could be used as the cultivation media for oleaginous microorganisms.
However, there is currently limited information on microbial oil production from
EFB and other lignocellulosic biomasses or agro-industrial wastes.
2
To develop an effective microbial production process, optimising oil yield is
critical as higher yields will have a significant impact on the economics of the
process. Strain selection is critical in order to identify potential microorganisms that
are able to grow on lignocellulosic hydrolysates, which may contain a variety of
sugar monomers and potential microbial growth inhibitors. Carbon-to-nitrogen (C/N)
ratio has been proposed as one of the key factors that may enhance yield as
oleaginous microorganisms typically accumulate oil under nitrogen-limiting
condition. Cultivation in bioreactor systems is also critical in order to develop an
understanding on the potential and challenges associated with scale-up.
1.2 Aims and scope of study
The overall objective of this project is to develop and optimise a process for the
production of microbial oil from EFB. The aims of this research are:
1. To develop a method for identifying high potential strains through cultivation on
pure sugar substrates and multi-criteria analysis.
2. To screen high ranking strains through cultivation on EFB hydrolysates and
select a prospective candidate for oil production from EFB hydrolysates.
3. To assess the potential viability of microbial oil production from EFB through
techno-economic evaluation.
4. To optimise cultivation conditions on EFB hydrolysates and explore the potential
for process scale up.
The aims of this study have been addressed through a research plan that
included three phases (Figure 1-1) which are
Phase 1: Selecting oleaginous microorganisms for microbial oil production
Phase 2: Evaluating microbial oil production from EFB
Phase 3: Optimising microbial oil production from EFB
3
1.3 Research contribution and significance
The research work undertaken in this study has resulted in the generation of
new knowledge. This includes the following original contributions.
Investigation of microbial oil production from EFB hydrolysates produced via
dilute acid pretreatment and enzymatic hydrolysis.
In this study, a comprehensive evaluation of microbial oil production from EFB
has been performed through assessment of cultivation on actual EFB
hydrolysates, assessment of fuel properties and assessment of the techno-
economics of microbial oil production from EFB and other oil palm biomasses.
Techno-economic evaluation of microbial oil production from oil palm biomass.
This study is critical as it evaluates potential viability of microbial oil production
process from EFB and other oil palm biomasses (trunk and frond).
Phase 1: Selecting oleaginous microorganisms for microbial oil production Selection of potential microorganisms from various groups (microalgae, yeasts
and fungi)
Cultivation of potential microorganisms on glucose, xylose and glycerol Multi-criteria assessment of selected microorganisms for oil production from
lignocellulosic hydrolysates Down-selection of potential microorganisms to 3 highest ranking
microorganisms for Phase 2
Phase 2: Evaluating microbial oil production from EFB Production of EFB hydrolysates through dilute acid pretreatment and
enzymatic hydrolysis of EFB
Cultivation of 3 highest ranking microorganisms on EFB hydrolysates and down-selection of prospective microorganisms for Phase 3
Fuel properties assessment of microbial oils from EFB Techno-economic evaluation of microbial oil production from EFB
Phase 3: Optimising microbial oil production from EFB Evaluation of the impact of cultivation parameters (carbon concentration,
nitrogen concentration, pH and inoculum size) on oil concentration and oil yield
Optimisation of cultivation condition on EFB hydrolysate Cultivation in bioreactor under optimised condition
Figure 1-1 Overall research plan
4
Microbial oil production from different groups (i.e., microalgae, yeasts and
fungi) of microorganisms on the same carbon substrates.
Optimisation of microbial oil production from EFB hydrolysate. This study is
significant as it evaluates the impacts of different conditions of cultivation on oil
yield and oil concentration.
Investigation of microbial cultivation on EFB hydrolysate in bioreactor system.
Microbial oil production from EFB by Rhodotorula mucilaginosa, Aspergillus
oryzae and Mucor plumbeus.
Multi-criteria analysis for ranking and selecting microorganisms for oil
production from lignocellulosic hydrolysates.
Microbial oil production from glycerol by Aspergillus oryzae and Mucor
plumbeus.
Cultivation of microalgae Chlorella protothecoides and Chlorella zofingiensis on
xylose.
Microbial oil production from sugarcane bagasse by Rhodotorula mucilaginosa,
Aspergillus oryzae and Mucor plumbeus. (See Appendix 2)
Optimising microbial oil extraction from Chlorella protothecoides using
Accelerated Solvent Extraction. (See Appendix 1)
This research is important as it contributes to the future profitability and
sustainability of the palm oil industry through (1) reducing environmental problems
associated with the management of palm oil processing wastes, (2) improving
sustainable practices in the palm oil industry, (3) increasing oil yield per ha of oil
palm cultivation and (4) adding value to a by-product of the palm oil processing
industry and therefore further contributing to higher economic returns to the palm oil
industry. The outcome of this research will not only benefit the palm oil industry, but
also other agro-industries that produce lignocellulosic residues.
1.4 Thesis outline
This thesis is a thesis by publication. Figure 1-2 shows an overview of the
thesis mapping the research aims by thesis chapter.
5
Chapter 1 (Introduction) describes the background to the research problem
that originates from the palm oil processing industry, as well as the potential solution
through microbial oil production. This chapter also outlines the aims of this research,
overview the thesis and identify the novelty and contribution to knowledge of the
research.
Chapter 2 provides a literature review of palm oil processing, waste
generation and potential solutions based on previous studies on value-addition to
palm oil biomass. This chapter reviews pretreatment and hydrolysis of
lignocellulosic biomass, which includes comparisons of different pretreatment
methods that have been applied to lignocellulosic biomass. This chapter also
discusses the challenges of using lignocellulosic hydrolysates from pretreatment for
microbial cultivation. Chapter 2 provides background information on microbial oil
production and a review of the cultivation of various groups of oleaginous
microorganisms including microalgae, yeasts and fungi. This chapter reviews various
applications of microbial oil for the production of biodiesel and oleochemical
industry applications. The chapter concludes with a discussion on the research gap.
Through the literature review, six oleaginous microorganisms were selected and
investigated in Phase 1 (Chapter 3). These microorganisms were Chlorella
protothecoides and Chlorella zofingiensis (microalgae), Cryptococcus albidus and
Rhodotorula mucilaginosa (yeasts), and Aspergillus oryzae and Mucor plumbeus
(fungi).
The results of Phase 1 are presented and discussed in Chapter 3. The six
selected microorganisms were grown on glucose, xylose and glycerol. Multi-criteria
analysis (MCA) using analytic hierarchy process (AHP) and preference ranking
organization method for the enrichment of evaluations (PROMETHEE) with
graphical analysis for interactive aid (GAIA), were used to rank and select the most
preferred microorganisms for oil production to be used for the production of
biodiesel. An MCA technique was developed based on the following criteria: oil
concentration, content, production rate and yield, substrate consumption rate, fatty
acids composition, biomass harvesting and nutrient costs. From this study, the MCA
identified A. oryzae, M. plumbeus and R. mucilaginosa as the highest ranking strains
for oil production from lignocellulosic hydrolysates.
6
Chapter 4 focuses on oil production by the three highest ranking strains on
lignocellulosic hydrolysates. The chapter presents the results and discussion of the
cultivation of A. oryzae, M. plumbeus and R. mucilaginosa on EFB hydrolysates.
This chapter discusses the use of both hydrolysates from dilute acid pretreatment and
enzymatic hydrolysis, as the carbon substrates, and the challenges of using
lignocellulosic hydrolysates for microbial oil production. The fuel properties of the
oils produced are also discussed in this chapter. The microorganisms are evaluated
for their capacity to grow on EFB hydrolysates based on oil productivity, tolerance to
Chapter 2: Literature review
Microbial oil production from oil palm empty fruit bunch
Chapter 4: Evaluating microbial oil production from EFB
2. Evaluation of oil production from oil palm empty fruit bunch. Published in Biofuels, Bioproduct and
Biorefining (2016)
Chapter 3: Selecting oleaginous microorganisms for microbial oil production
1. A multi-criteria analysis approach for ranking and selection of microorganisms for the production of oils for biodiesel production. Published in Bioresource Technology (2015)
Chapter 5: Optimising microbial oil production from EFB
3. Improved microbial oil production from oil palm empty fruit bunch by Mucor plumbeus. Accepted with modification for publication
in Fuel (2016)
Research aim 1: To develop a method for identifying high potential strains through cultivation on pure sugar substrates and multi-criteria analysis.
Research aim 2: To screen high ranking strains through cultivation on EFB hydrolysates and select a prospective candidate for oil production from EFB hydrolysates.
Research aim 3: To assess the potential viability of microbial oil production from EFB through techno-economic evaluation.
Research aim 4: To optimise cultivation conditions on EFB hydrolysates and explore the potential for process scale up.
Chapter 6: Conclusion and future work
Chapter 1: Introduction
Figure 1-2 Overview of thesis
7
inhibitors, sugar consumption profiles and fuel quality. From the cultivation on both
EFB hydrolysates, M. plumbeus recorded the highest oil concentration and oil yield.
The experimental data from the cultivation of M. plumbeus on EFB hydrolysates was
used for techno-economic evaluation of microbial oil production from oil palm
biomass. This chapter also elaborates and visualises the potential integration of
microbial oil from oil palm biomass within palm oil industrial process.
Chapter 5 focusses on optimising the cultivation of M. plumbeus on EFB
hydrolysate. This chapter also describes and compares the cultivation of M. plumbeus
on an EFB hydrolysate in bioreactor system with the cultivation in shake flasks. In
this chapter, the impact of different cultivation parameters on oil production from
EFB hydrolysates was investigated. The parameters selected for optimisation are
pH and inoculum size. In this chapter, the optimisation study was performed through
response-surface methodology varying these four cultivation parameters. The
optimised cultivation conditions were assessed based on oil concentration and oil
yield. This chapter discusses issues associated with ethanol accumulation produced
as a by-product of the microbial cultivation. This chapter also discusses the
morphology of the fungal system during the cultivation. The optimised cultivation
conditions were further used for the study of the cultivation on EFB hydrolysates in
bioreactor systems.
Chapter 6 (Conclusion) summarises the overall results of the research for the
selection of microorganisms for the microbial oil production from EFB hydrolysates.
This chapter also reviews the key findings of this research associated with the
research aims. This chapter also presents future work that can be developed from the
outcome of this research.
9
Literature Review
2.1 Overview of palm oil processing
2.1.1 Oil palm
Oil palm (Elaeis guineensis) is the world’s highest yielding oil crop, with
approximately eight times higher productivity than rapeseed and six times higher
than soybean (Figure 2-1(a)). Palm oil is extracted from the fruit mesocarp, which is
then fractionated into palm olein and stearin [1]. Palm olein is used in food
applications, especially as a feedstock for cooking oil, while palm stearin is mainly
used in non-food and oleochemicals production. Oil palm is widely grown in tropical
countries such as Malaysia and Indonesia with 47.6 million tonnes of palm oil
produced in 2013 (Figure 2-1(b)) [2]. Malaysia is the second largest producer of
palm oil in the world, and contributes 38% of the world’s production. Malaysia
accounts for about 10% of the global production of oils and fats, with around 4
million ha of land used for oil palm cultivation [3].
The demand for palm oil has been increasing every year, and since 2007, palm
oil has become the most widely consumed vegetable oil [4]. In 2012, 50.17 million
tonnes of palm oil was produced globally, a 5% increase from 2011 [5]. One of the
reasons for the escalating demand for palm oil is because it provides the cheapest
edible oil source (US$518/t) compared with other major vegetable oils such as
soybean (US$641/t) and rapeseed (US$784/t) [6]. Oil palm is a low energy input
crop, since it is a perennial crop and does not require annual sowing [7]. Oil palm
cultivation requires less energy input per tonne of oil produced compared to soybean
and rapeseed [3].
The increasing demand for palm oil is also being driven by its extensive
application in oleochemical industries. The fatty acids and alcohols produced from
these oils are essential raw materials for the production of surfactants [8]. The
application of these surfactants is mainly in the production of consumer products
such as laundry detergents, shampoos, soaps and cleaning products [8]. In 2005,
there were forty-eight oleochemical refineries in Malaysia, which holds a 25% share
of the global market for fatty alcohols and acids [9].
10
Indonesia
Malaysia
Thailand Nigeria Colombia
0
5
10
15
20
25
30
Pal
m o
il (
t/ye
ar)
Palm oil
Rapeseed oil
Soybean oil
Sunflower oil
0
0.5
1
1.5
2
2.5
3
3.5
Oil
yie
ld (
t/ha
/yea
r)
(a)
(b)
Figure 2-1 (a) Average oil yield per ha per year for selected oil crops from 2010-2012 [2]. (b) Top five producers of palm oil in the world with average
palm oil production from 2011-2013 [2].
Oil palm trees are replanted after an economic life of twenty-five years [10].
Harvesting of the palm fruits typically begins three years after planting, with a
maximum oil yield in the 12–14th year after which the yield continuously diminishes
until the end of the plantation’s economic life [11]. Palm fruits grow in bunches with
approximately 1000 to 3000 fruits per bunch [3]. Each bunch weighs approximately
10-15 kg with about 25% (w/w) of oil per bunch [12]. Palm oil is extracted from the
mesocarp of the palm fruit, which is the oily and fleshy outer layer of the fruit seed
or kernel shown in Figure 2-2. The palm kernel is also rich in oil, however, palm
kernel oil has different fatty acid compositions to palm oil. Palm oil mainly consists
of C16 and C18 fatty acids, while palm kernel oil comprises of C12 and C14 fatty
acids which are typically found in coconut oil [1]. Therefore, unlike palm oil, the
application of palm kernel oil is mainly in soap manufacturing [3].
11
Figure 2-2 Oil palm tree, fresh fruit bunch and longitudinal section of fresh oil palm fruit [12, 13]
Oil palm cultivation is rapidly expanding in Southeast Asia due to increasing
global consumer demand [14]. However, sustainability and environmental issues
have become one of the major concerns for palm oil expansion. As one of the
world’s leading producers of palm oil, Malaysia has adopted several measures for
enhancing sustainable palm oil production. The palm oil industry in Malaysia has
adapted Good Agricultural Practices (GAP) and Integrated Pest Management (IPG)
in order to ensure conservation of the environment and biodiversity [15]. In addition,
a no forest areas encroachment policy has been legislated in Malaysia, whereby
cultivation expansion can only occur on unused land or land converted from other
crops [15]. Zero burning was also legislated in Malaysia in 1989, as part of the drive
to more sustainable practices [16]. Through the implementation of these policies,
more than 50% of total land area in the country remains as rainforests [17].
Approximately 25% of oil palm plantation land area in the country has been
converted from land formerly used for rubber, coconut and cocoa plantations [7].
2.1.2 Palm oil processing: wastes generation and problems
Fresh fruit bunches (FFB) harvested from oil palm plantations are processed in
palm oil mills for oil extraction. FFB processing involves sterilisation, threshing and
stripping of fruits, digestion, and extraction of oil [16]. In 2011, there were
approximately 92.9 million tonnes of FFB harvested and processed in Malaysian
palm oil mills, which generated approximately 44 million tonnes of solid residues,
and 62 million tonnes of liquid waste known as palm oil mill effluent (POME) [18].
The solid residues from the palm oil extraction process comprises of, by weight,
around 53% of empty fruit bunch (EFB), 29% of shell and 18% of fibre [18]. EFB is
12
the fibrous mass left behind after the FFB stripping process (Figure 2-3). EFB
consists of a bundle of fibres with an average size of approximately 1 mm in length,
25 μm wide and 3 μm thick [19].
Figure 2-3 EFB from palm oil mills [20, 21]
The palm oil industry is one of the key biomass producers in Malaysia [18].
Aside from the abundance of residues generated by the extraction mills, there are
significant amounts of waste coming from oil palm plantations in Malaysia. For
instance, there are large amounts of oil palm frond (OPF) from the daily pruning of
the trees, and oil palm trunks (OPT) from replanting that require disposal [3]. These
wastes are typically shredded for in-situ composting [15]. Malaysian palm oil mills
have been utilising a portion of biomass wastes, particularly the palm kernel shell
(PKS) and mesocarp fibre (MF), as a source of electricity and steam generation [18].
However, the majority of EFB has not been optimally recycled for other applications
in the mills. Some millers opt to use EFB as mulch or/and fertiliser [22].
EFB is not preferred for burning or combustion as fresh EFB consists of about
60% water [23]. The high moisture content of EFB makes it unfavourable for
handling and transportation. EFB is usually left for decomposition at the mills or
plantations [18]. Another issue with the use of EFB as soil conditioner is that it can
attract oil palm pests [23]. Poor management of EFB decomposition can lead to
substantial methane emissions to the atmosphere [19]. Methane is one of the key
greenhouse gases which are known to be the major cause of global warming. In some
Southeast Asian countries, palm oil wastes are disposed of through open burning
which contributes air pollution [23]. As an alternative to disposal, these palm oil
wastes can instead be converted into higher value products and contribute to higher
FeSO4·7H2O and 0.1 g CaCl2 at pH 5.5 [21]. Glucose, xylose and glycerol (30 g/L)
were used as the carbon sources in the media supplemented with 4 g/L yeast extract.
Cultures were conducted in triplicate in 500 mL Erlenmeyer flasks containing 200
mL media placed on orbital shaking incubator (Ratek, Australia). Microalgal and
yeast strains were cultivated with 20% (v/v) inocula from their respective
precultivation medium (4 days), at an orbital rate of 180 rpm with temperature
maintained at 28°C. The cultivation for microalgae strains was carried out in the
dark. Fungal strains were cultivated with inoculum from 24 h of precultivation, at an
orbital rate of 160 rpm with temperature maintained at 30°C [22].
Microalgal and yeasts biomass were harvested by centrifugation at 6805 g for 7
min (Sorvall Biofuge Primo R, USA) [23]. Fungal biomass was harvested by vacuum
filtration (Whatman 54 filter paper). The harvested biomass samples were washed
three times (200 mL/wash) using Millipore water and freeze-dried to a constant
weight.
3.2.1.2 Oil extraction
Oil was extracted from the biomass by Accelerated Solvent Extraction (ASE)
technique using Dionex ASE 350 (Thermo Fisher Scientific Inc., USA). The samples
for extraction were prepared by mixing dry biomass (~0.1 g) with 0.4 g of
diatomaceous earth (Thermo Fisher Scientific, Inc., USA) and loaded into 5 mL
cells. The extraction conditions had been optimised and were as follows (see
Appendix 1): temperature, 130 ; static time, 5 min; rinse volume, 25% of cell
volume; purge time, 60 s; and using 4 static cycles. The solvent used was a mixture
of chloroform:methanol in a ratio of 2:1 (v/v) [24]. The extracted oil was collected in
pre-weighed collection bottles. The solvents were evaporated under a stream of
nitrogen. Unless otherwise specified, all results are reported on a dry weight (DW)
basis.
55
3.2.1.3 Oil analyses
Sugars and glycerol concentrations were analysed using high-performance
liquid chromatography (HPLC) by a Waters HPLC system equipped with a SP810
carbohydrate column (300 mm × 8.0 mm, Shodex, Japan) and a refractive index (RI)
detector (Waters 410, US). The column temperature was 85°C and the mobile phase
was water, with a flow rate of 0.5 mL/min [25].
For the determination of fatty acids composition, fatty acid methyl esters
(FAME) were prepared using the method described by Mulbry et al. (2009).FAME
analysis was performed by gas chromatography-mass spectrometry (GC-MS) by
Shimadzu GCMS-TQ8040 (Shimadzu Corporation, Japan) on a TG-WAXMS
column (30 m long × 0.32 mm I.D.× 1 µm film thickness; Thermo Fisher Scientific,
Inc., USA). The carrier gas was helium at a flow rate of 1.5 mL/min. A 10:1 split
injection was used. The injection temperature was set at 230°C, the MS ion source
temperature at 220°C and the MS interface temperature at 240°C. The GC-MS
method was carried out using the following temperature program: initial temperature
at 40°C, hold for 2 min, followed by 10°C/min ramp to 230°C and hold for 20 min.
Mass spectrometry was performed using Q3 scan with an m/z 20-650 scanning range.
Chromatograms and mass spectra were evaluated using the GCMSsolution software
(Shimadzu Corporation, Japan). The retention times and mass spectra were identified
using FAME mix (F.A.M.E. Mix, C8-C24; Sigma-Aldrich, Australia).
3.1.2.4 Multi-criteria analysis
Establishing criteria hierarchy
AHP and PROMETHEE-GAIA were used for MCA. PROMETHEE-GAIA
was implemented using Visual PROMETHEE 1.4 Academic Edition. The stated goal
of the MCA was to select the most suitable prospective microorganism(s) for oil
production from lignocellulosic hydrolysates’ model compounds, glucose and
xylose. The alternative solutions were selected to be the six microorganisms studied
which were C. protothecoides, C. zofingiensis, C. albidus, R. mucilaginosa, A.
oryzae and M. plumbeus. Figure 3-1 shows the criteria hierarchy that was established
from the key parameters reported in the Introduction section. The quantitative criteria
under Cluster 1 (C1 – C6) were evaluated based on the results of the experimental
56
study on microbial cultivation on glucose, xylose and glycerol substrates. The criteria
under Cluster 2 (C7 and C8) were evaluated qualitatively.
Figure 3-1 Criteria hierarchy for the evaluation of microorganisms for microbial oil production. Cluster 1 criteria were evaluated based on cultivation results on glucose
(G), xylose (X) and glycerol (L).
Establishing criteria weights
AHP techniques were used to determine the relative weightings of each
criterion [15, 19]. This was based on hierarchy, priority setting and logical
consistency [19, 26]. Relative priorities were given to each element through pairwise
The criteria assigned under Cluster 1, C1 - C6 were assessed with the priorities
of C1 > C2 > C3 > C4 > C5 > C6. The first two criteria, C1 (oil concentration) and
C2 (oil content) were given the highest priority as they reflect the key economic
advantage resulting from high concentration and yield of the desired product from
each carbon substrate. Substrate consumption rate (C3) reflects the potential
economic benefit of lower capital and operating costs from reduced fermenter
capacity. Oil yield (C4) reflects the efficiency with which the microorganism
converts the substrate (which is an operating cost) to product (which is a revenue).
Fatty acid profile (C5) evaluates the relative value of the oil for use in biodiesel
production, and is calculated as the percentage of saturated and mono-unsaturated
fatty acids in the oil produced. Oils with high levels of saturated and
monounsaturated fatty acids are desirable for biodiesel application [27].
58
Polyunsaturated fatty acids especially those with more than four double bonds are
less preferred for biodiesel production due to the low oxidative stability of the
biodiesel during storage [28]. However, microbial oils with high level of
polyunsaturated fatty acids, such as linoleic acid (C18:2n6) and linolenic acid
(C18:3n3) have the potential to be used in health products manufacturing [2]. Oil
productivity (C6) reflects average oil concentration per day of cultivation.
For the criteria belonging to Cluster 2 (C7 and C8), the alternatives studied
were categorised based on the groups of microorganisms as each alternative in the
same group were assumed to share similar characteristics. For evaluating qualitative
criteria, a 5-point scale was used (very good, good, average, bad, and very bad).
Fungi were classified as very good for C7 (Biomass harvesting cost) because fungal
strains generally grow in pellet form. Pellet form is preferable for harvesting as the
biomass can be harvested by simple sedimentation and filtration, whereas single cell
biomass requires centrifugation or finer filtration techniques. Harvesting by
sedimentation and filtration is a lower cost harvesting technique compared to
harvesting via centrifugation [5]. For criterion C8 (Nutrient cost), yeasts were given
the best ranking as yeast species generally require fewer nutrients in the media
compared to microalgae for oleaginous cultivation[10].
Ranking of alternatives
In PROMETHEE, the preference function converts the deviations between the
evaluation of two alternatives for each criterion into a preference degree ranging
from 0 to 1 [16]. The preference functions used in this study are V-shape functions
for quantitative criteria, and the usual function for qualitative criteria [17]. V-shape
function specifies values of preference threshold, p, which is the smallest deviation
that is considered as sufficient to generate a full preference [17]. The indifference
threshold, q, is the largest deviation that is considered negligible by the decision
maker and is equal to 0 in the V-shape function [17]. The values of p in this study
were determined using the built-in Preference Function Assistant in Visual
PROMETHEE.
GAIA was used to further analyse and visualise the outcomes of the analysis.
The following elements refer to results shown in the GAIA plane [15, 17]: (1) The
criteria are represented by axes. Axes are oriented in approximately the same
59
direction for criteria expressing similar preference and in opposite directions for
conflicting criteria. Axes are oriented orthogonally for unrelated criteria. (2)
Alternatives are represented by shapes. Alternatives with similar profiles are
positioned close to each other. Alternatives with better performance on a given
criterion are located in the direction of the corresponding criterion. (3) The weights
of criteria are represented by the pi vector on the decision axis. The orientation of
this axis shows which criteria are in accordance with PROMETHEE rankings and
which are not.
3.1.3 Results and discussion
3.1.3.1 Biomass concentrations and carbon substrate consumptions on glucose,
xylose and glycerol
Figure 3-2(a) shows the biomass concentrations of the six selected
microorganisms growing on glucose, xylose and glycerol substrates. The yeast strain,
R. mucilaginosa gave the highest biomass concentration of 16.8 (±1.8) g/L on
glucose, while the other microorganisms had similar biomass concentrations on
glucose ranging from 7.9 to 9.8 g/L. One possible reason for the high biomass
concentration of R. mucilaginosa is that the cultivation was not carried out in
complete darkness. Biomass production from other species of Rhodotorula, R.
glutinis was shown to increase when it was cultivated under light irradiation
conditions [29]. For cultivation on xylose, R. mucilaginosa, A. oryzae and M.
plumbeus all resulted in high biomass concentrations of 10.8 (±0.4) g/L, 10.0 (±0.4)
g/L and 9.3 (±0.7) g/L respectively. However, no significant biomass growth resulted
from C. protothecoides and C. zofingiensis cultivation when xylose was used as the
carbon source. These results are in agreement with a previous study that showed
Chlorella species (e.g., C. vulgaris and C. sorokiniana) were not able to assimilate
xylose heterotrophically [7]. Fungal strains M. plumbeus and A. oryzae also showed
the highest biomass concentrations on glycerol (10.2 ±0.4 g/L and 9.5 ±0.8 g/L
respectively). The results showed that both fungal strains, M. plumbeus and A.
oryzae, and both yeasts strains, R. mucilaginosa and C. albidus, were able to grow on
each of the three carbon sources studied. Interestingly, M. plumbeus and A. oryzae
showed relatively consistent biomass concentrations on glucose, xylose and glycerol
substrates.
60
61
(a)
(b)
(c)
Figure 3-2 (a) Biomass concentration, (b) oil content and (c) oil concentrations for growth of six microorganisms on glucose, xylose and glycerol.
0
2
4
6
8
10
12
14
16
18
20
Glucose Xylose GlycerolB
iom
ass
conc
entr
atio
n (g
/L)
C. protothecoides
C. zofingiensis
C. albidus
R. mucilaginosa
A. oryzae
M. plumbeus
0
5
10
15
20
25
30
35
40
Glucose Xylose Glycerol
Oil
con
tent
(%
, w/w
)
C. protothecoidesC. zofingiensisC. albidusR. mucilaginosaA. oryzaeM. plumbeus
0
0.5
1
1.5
2
2.5
3
3.5
4
Glucose Xylose Glycerol
Oil
con
cent
rati
on (
g/L
)
C. protothecoidesC. zofingiensisC. albidusR. mucilaginosaA. oryzaeM. plumbeus
62
The results of substrates consumption by the six microorganisms studied over
168 h of cultivation are shown in Figure 3-3. All six microorganisms were shown to
consume glucose more rapidly than xylose and glycerol. Generally, glucose is more
preferable than xylose as a fermentation substrate as assimilation of xylose requires
specific metabolic pathways [13]. Glucose was shown to be completely consumed by
fungal strains A. oryzae and M. plumbeus within only 48 h to 72 h of cultivation.
Yeast strain R. mucilaginosa and microalgae strain C. protothecoides consumed
glucose completely by the end of the cultivation period. Xylose was completely
consumed in the media by A. oryzae in 96 h, where as it took 144 h for M. plumbeus
and R. mucilaginosa to consume xylose completely. Extremely low consumption of
xylose was evident for either of the microalgae strains. All microorganisms
consumed glycerol at a slower rate than glucose and xylose. Consumption of glycerol
was again the fastest for the fungal species A. oryzae and M. plumbeus.
The two fungal strains, A. oryzae and M. plumbeus, demonstrated the highest
consumption rates on glucose, xylose and glycerol. It is known that upon depletion of
the carbon source, there exists the possibility of lipid turnover, in which storage
lipids are metabolised resulting in a reduction in lipid content [30]. In this study, the
lipid content was not monitored at each time point as the work focused on the
development of MCA method for screening and selection of optimal oil producing
microorganisms. It is noted, however, that the peak oil content for microorganisms
with rapid substrate consumption may have been higher than the results show.
63
(a)
(b)
(c)
Figure 3-3 Consumption of (a) glucose, (b) xylose and (c) glycerol over 168 h of cultivation.
0
20
40
60
80
100
0 24 48 72 96 120 144 168
Glu
cose
con
sum
ptio
n (%
, w
/w)
Cultivation time (h)
C. protothecoidesC. zofingiensisC. albidusR. mucilaginosaA. oryzaeM. plumbeus
0
20
40
60
80
100
0 24 48 72 96 120 144 168
Xyl
ose
cons
umpt
ion
(%,
w/w
)
Cultivation time (h)
C. protothecoidesC. zofingiensisC. albidusR. mucilaginosaA. oryzaeM. plumbeus
0
20
40
60
80
100
0 24 48 72 96 120 144 168
Gly
cero
l con
sum
ptio
n (%
, w
/w)
Cultivation time (h)
C. protothecoidesC. zofingiensisC. albidusR. mucilaginosaA. oryzaeM. plumbeus
64
3.1.3.2 Microbial oil production from different carbon substrates
Figure 3-2(b) shows the results of the oil contents of the strains on glucose,
xylose and glycerol substrates after 168 h cultivation. C. protothecoides cultivation
on glucose showed the highest oil content of 35.4% (w/w), followed by A. oryzae
(26.9%), M. plumbeus (26.2%), C. zofingiensis (24.7%), R. mucilaginosa (21.6%)
and C. albidus (19.5%). It has been demonstrated in previous studies that C.
protothecoides is an excellent oil producer on glucose with up to 58% oil content
obtained from batch cultivation in a 5 L bioreactor for 140 h [23]. The highest oil
content on xylose was achieved by M. plumbeus, which was 23.8%, followed by A.
oryzae, C. albidus and R. mucilaginosa (oil contents of 20.7%, 18.3% and 14.4%
respectively). As there was almost no growth of Chlorella strains on xylose, the oil
content was not measured. The highest oil contents on glycerol were achieved by M.
plumbeus, A. oryzae, and C. albidus, which were all around 26% (27.4%, 25.8% and
26.4% respectively). Lower oil contents on glycerol substrates were shown by R.
mucilaginosa, C. protothecoides and C. zofingiensis.
Figure 3-2(b) also shows that A. oryzae and M. plumbeus had consistent oil
contents with varying carbon sources. Although the two fungal strains had ~8-9%
lower final oil contents than C. protothecoides, these strains grew much faster and
are likely to result in comparable or higher oil productivity. Yeast strain R.
mucilaginosa produced the highest biomass concentration while still producing
similar oil contents to most of the other strains.
Figure 3-2(c) shows oil concentrations for the six microorganisms growing on
glucose, xylose and glycerol. The cultivation of R. mucilaginosa on glucose resulted
in the highest oil concentration of 3.61 (±0.16) g/L primarily as a result of the very
high biomass concentration compared to the other species. C. zofingiensis and the
two fungal strains had similar oil concentrations on glucose. M. plumbeus showed
the highest oil concentration on xylose and glycerol (2.21 ±0.04 g/L and 2.78 ±0.10
g/L respectively), followed by A. oryzae (2.07 ±0.05 g/L and 2.45 ±0.15 g/L
respectively). Figure 3-2(c) also shows the consistency in the oil concentrations
achieved by the two fungal strains across all three carbon sources compared to the
other species which tended to be more variable with varying carbon substrates.
65
3.1.3.3 Fatty acids profiles
The results of the fatty acid compositions of the six microorganisms growing
on glucose, xylose and glycerol are presented in Table 3-3. The major fatty acids
identified were palmitic (C16:0), stearic (C18:0), oleic (C18:1) and linoleic (C18:2).
Oleic acid was the predominant fatty acid in most cases which is in accordance with
previous studies [9]. Variations were observed for the cultivation of C. albidus on
glucose and xylose substrates, with linoleic acid as predominant fatty acid while
palmitic acid was the predominant fatty acid with C. zofingiensis on glycerol. The
reasons for high accumulation of palmitic acid by C. zofingiensis on glycerol are
unknown as this is the first study to cultivate C. zofingiensis on glycerol.
Nevertheless, this C. zofingiensis may have similar pathways to metabolise glycerol
to Chlorella saccharophila reported previously, which produced palmitic acid as
predominant fatty acid on glycerol but oleic acid on glucose [31, 32]. The
composition of oleic acid was decreasing and palmitic acid was increasing with
increasing ratio of glycerol mixed with glucose substrate [32].
3.1.3.4 Preference ranking
Based solely on the oil concentration results above, it could be concluded that,
of the microorganisms assessed, R. mucilaginosa and C. protothecoides were the
most prospective microorganisms for microbial oil production from glucose. On the
other hand, A. oryzae and M. plumbeus appeared to be the most prospective for oil
production from xylose. M. plumbeus had the lowest polyunsaturated fatty acid
content when grown on glucose and hence potentially produced better oil for
biodiesel production but had the highest polyunsaturated fatty acid when grown on
glycerol. Furthermore, these initial conclusions ignore the impact of other aspects
that impact on production cost including harvesting and nutrition costs. Therefore,
PROMETHEE-GAIA was used to systematically assess each alternative based on the
criteria shown above.
66
Table 3-3 Fatty acid compositions of oil extracted from six different microorganisms grown on various carbon substrates.
Microorganisms
Relative abundance of total fatty acids (%, w/w) SFAa MUFAa PUFAa C14:0 C16:0 C16:1 C18:0 C18:1 C18:2 C18:3 C20:0
Glucose medium C. protothecoides 6.3
(±2.0) 23.8 (±1.5)
3.8 (±0.7)
6.1 (±0.7)
42.8 (±2.6)
5.1 (±3.0)
6.8 (±2.4)
- 38.2 49.2 12.6
C. zofingiensis 7.2 (±1.1)
22.4 (±1.9)
7.8 (±2.0)
6.9 (±3.0)
42.2(±2.7)
6.5 (±9.8)
4.0 (±3.8)
- 37.6 51.5 10.8
C. albidus 1.5 (±0.6)
22.7 (±0.1)
1.7 (±0.4)
4.1 (±1.3)
34.1 (±2.5)
35.8 (±2.2)
- - 28.4 35.8 35.8
R. mucilaginosa 2.7 (±0.7)
18.9 (±3.7)
1.7 (±0.3)
6.9 (±2.2)
54.2 (±6.5)
12.6 (±8.4)
2.9 (±0.9)
- 28.6 55.9 15.5
A. oryzae 4.5 (±2.8)
25.5 (±4.4)
3.4 (±0.4)
15.6 (±6.0)
34.9 (±2.9)
9.8 (±8.3)
3.1 (±0.5)
1.1 (±0.7)
47.7 39.1 13.2
M. plumbeus 2.0 (±0.8)
28.8 (±0.8)
2.5 (±0.5)
22.1 (±1.7)
37.4 (±0.6)
2.8 (±2.1)
1.1 (±0.7)
1.6 (±0.3)
55.4 40.6 4.00
Xylose medium C. albidus - 29.5
(±2.0) - 13.4
(±1.8) 23.4 (±2.2)
33.7 (±0.3)
- - 42.9 23.4 33.7
R. mucilaginosa 1.8 (±0.2)
20.3 (±1.7)
1.11 (±0.3)
6.1 (±0.8)
49.2 (±3.1)
20.1 (±5.3)
1.3 (±1.2)
- 28. 3 50.4 21.4
A. oryzae 0.8 (±0.1)
20.5 (±1.9)
1.7 (±0.3)
16.4 (±0.7)
37.5 (±0.9)
21.1 (±3.2)
- 1.5 (±0.1)
39.4 39.4 21.2
M. plumbeus 1.3 (±0.6)
20.5 (±4.0)
1.8 (±0.6)
19.0 (±1.7)
33.8 (±3.1)
21.1 (±9.0)
1.0 (±0.8)
1.5 (±0.3)
42.2 35.7 22.1
Glycerol medium C. protothecoides 10.3
(±1.2) 26.6 (±0.2)
4.4 (±1.5)
6.5 (±0.5)
35.9 (±4.3)
7.0 (±9.4)
2.1 (±1.8)
- 46.8 43.4 9.9
C. zofingiensis 17.7 (±4.2)
56.4 (±3.2)
- 8.2 (±2.0)
10.4 (±6.4)
7.3 (±2.4)
- - 82.2 10.4 7.3
C. albidus 1.5 (±0.1)
24.4 (±1.1)
1.9 (±0.3)
5.5 (±2.2)
42.4 (±2.3)
24.3 (±0.6)
- - 31.4 44.3 24.3
R. mucilaginosa 4.6 (±1.2)
14.7 (±1.3)
1.6 (±0.2)
9.1 (±1.9)
47.6 (±3.6)
6.6 (±2.5)
15.7 (±3.2)
- 28.4 49.3 22.4
A. oryzae 0.9 (±0.5)
14.2 (±0.9)
1.9 (±0.5)
16.9 (±0.8)
34.4 (±1.1)
29.3 (±1.3)
0.5 (±0.0)
1.8 (±0.3)
33.9 36.4 29.8
M. plumbeus 0.5 (±0.7)
14.1 (±0.0)
2.8 (±1.1)
14.3 (±0.2)
30.9 (±2.4)
35.6 (±2.1)
0.5 (±0.2)
1.3 (±0.2)
30.2 33.7 36.1
a SFA means saturated fatty acids, MUFA means monounsaturated fatty acids and PUFA means polyunsaturated fatty acids.
Figure 3-4(a) shows the results of the PROMETHEE I partial rankings for the
six microorganisms studied. In PROMETHEE I, the presence of crossed tie lines
indicate that the alternatives are not comparable using this technique. For instance,
M. plumbeus is not comparable to A. oryzae because M. plumbeus obtained a higher
Phi-(negative preference flow), and a lower Phi+ (positive preference flow)
compared to A. oryzae.
67
(a)
(b)
Figure 3-4 (a) PROMETHEE I partial ranking of alternatives and (b) PROMETHEE II complete ranking where RM denotes R. mucilaginosa, AO A. oryzae, MP M.
plumbeus, CP C. protothecoides, CA C. albidus and CZ C. zofingiensis.
68
Figure 3-4(b) also shows the results of the PROMETHEE II complete rankings
for the six microorganisms studied. The only microorganisms that obtained positive
Phi scores were A. oryzae, M. plumbeus and R. mucilaginosa with the two fungal
species A. oryzae and M. plumbeus being the most preferred options with almost
equivalent Phi scores. As a result, based on the criteria selected and the experimental
results, these three microorganisms (A. oryzae, M. plumbeus and R. mucilaginosa)
were predicted to be more preferred for oil production from the lignocellulosic
hydrolysates model compounds, and glycerol than the other microorganisms.
The GAIA plane from the analysis is shown in Figure 3-5 and has a quality
level of 80.5% which is reliable as it is above 70% quality significance level. The pi
decision axis is aligned in the direction of the fungal strains A. oryzae and M.
plumbeus, which shows that these alternatives are preferred which is in agreement
with the PROMETHEE II ranking.
In the GAIA plane, the criteria vectors that lie in the same direction as the
decision vector reflect the influence that these criteria have on the decision. Figure
3-5 shows that the substrate consumption rate (C3) and fatty acid profiles (C5) for all
of the substrates express a positive preference on the decision.
Sensitivity analysis was carried out on the selected preference function and
criteria weights. By substituting the V-shape preference function with linear function
for all quantitative criteria without changing the preference threshold, (p), the
PROMETHEE II ranking remains the same. The sensitivity of the criteria weights to
the results are analysed based on weight stability intervals (Table 3-4). Weight
stability intervals are the limits where any variation in weight within the intervals
will not change the ranking of PROMETHEE II, given that there is no change to the
relative weights of other criteria [15]. Most of the criteria exhibited broad weight
stability intervals which show that the analysis is robust.
69
(a)
(b)
Figure 3-5 GAIA plane at (a) 100% zoom and (b) 400% zoom without the alternatives. The alternatives are denoted as RM for R. mucilaginosa,
AO for A. oryzae, MP for M. plumbeus, CP for C. protothecoides, CA for C. albidus and CZ for C. zofingiensis. The criteria are denoted as C1-G to C6-G for criteria of Group 1.1 (Glucose), C1-X to C6-X for criteria of Group 1.2 (Xylose) and C1-L to
C6-L for criteria of Group 1.3 (Glycerol). Some criteria are not visible due to overlapping such as C1-X by C1-L, C5-X and C3-X by C3-L, C5-G by C7, C2-X and
C6-X by C2-L.
70
Table 3-4 Weight stability intervals for criteria with relative weight>5%. Criteria Weight
(%) Weight stability intervals
C1-G Oil concentration on glucose 22.50 [2.70-28.02] C2-G Oil content on glucose 11.22 [0-22.15] C3-G Consumption rates on glucose 7.41 [3.40-71.58] C1-X Oil concentration on xylose 11.95 [2.51-34.65] C2-X Oil content on xylose 5.96 [0-19.23]
The three highest ranking alternatives, A. oryzae, M. plumbeus and R.
mucilaginosa were further analysed using GAIA Web to determine the influence of
individual criteria on the preference result (Figure 3-6). GAIA Web shows a
graphical representation of the unicriterion net flow scores for the selected
alternative. The criteria axes in GAIA Web are positioned with the same orientation
as in the GAIA plane, where criteria with similar preferences are located close to
each other. The GAIA Web shows the key criteria with the radial distance indicating
unicriterion net flows with -1 value at the centre of the web and +1 on the outer
circle.
Figure 3-6(a) shows that R. mucilaginosa performed strongly for the criteria of
oil concentration, oil yield and substrate consumption rate on glucose but the criteria
of oil content, fatty acid profile and oil productivity on glycerol were weak. Oil
concentration, oil content, and substrate consumption rate on xylose and glycerol
were all weak. On the other hand, A. oryzae shows very good preference for oil
concentration, oil content, and fatty acid profiles on xylose and glycerol and oil
content and fatty acid profile on glucose. In fact it is noted that A. oryzae showed
good preference results across most criteria with the exception of oil concentration
and oil yield on glucose, and productivity and oil yield on glycerol. The fungal strain
M. plumbeus showed very good preferences for most of the criteria on glucose,
xylose and glycerol with the exception of oil concentration and oil yield on glucose,
and productivity on glycerol.
The incomparability between M. plumbeus with A. oryzae in PROMETHEE I
can be assessed through the GAIA Webs. A comparison of the GAIA Webs between
these two species shows different strengths in preference between these fungal
strains for criteria such as fatty acid profiles but the incomparability is not highly
significant. The GAIA Webs confirmed the results obtained from the GAIA plane
71
reflecting that A. oryzae and M. plumbeus showed good preference for most of the
criteria specified.
72
(a)
(b)
73
(c)
Figure 3-6 GAIA Webs for top three alternatives from PROMETHEE which are (a) R. mucilaginosa, (b) A. oryzae and (c) M. plumbeus. Criterion C2-X is not visible due
to overlapping by C6-X.
The preferred alternatives for oil production for biodiesel production from
highest to lowest were established as follows: (1) A. oryzae; (2) M. plumbeus; (3) R.
mucilaginosa; (4) C. protothecoides; (5) C. albidus and (6) C. zofingiensis. The
microorganisms with positive Phi scores (A. oryzae, M. plumbeus and R.
mucilaginosa) were selected as the most prospective species and further analysed
using unicriterion net flow analysis in GAIA Webs. The variations in positive
preferences across these three microorganisms were confirmed by PROMETHEE I,
the GAIA plane and also the GAIA Webs. Therefore, fungal strains A. oryzae, M.
plumbeus and yeast strain R. mucilaginosa have potential for industrial oil
production for biodiesel applications.
The MCA proposed can be improved for ranking and selecting the best
microorganism for oil production from a specific type of hydrolysates, whereby the
priority for carbon substrates can be adjusted accordingly. MCA for oil production
from hydrolysates of liquid fraction of pretreated lignocellulosic materials may
74
include microorganisms’ tolerance to inhibitors such as furfural and 5-
hydroxymethylfurfural, as one of the criteria for ranking and selecting the best
microorganism.
3.1.4 Conclusion
In this study, a MCA approach was used to evaluate the performance of oil
production with different microorganisms. The MCA technique using AHP and
PROMETHEE-GAIA showed that the only microorganisms with positive Phi scores
were A. oryzae, M. plumbeus and R. mucilaginosa. Further GAIA analyses showed
that the fungal strains A. oryzae and M. plumbeus provided superior performance
across a wide range of criteria including growth on glucose and xylose substrates.
Overall, A. oryzae, M. plumbeus and R. mucilaginosa showed promise for biodiesel
production using the lignocellulose hydrolysates model compounds, glucose and
xylose.
3.1.5 Acknowledgements
The authors acknowledge Ministry of Education Malaysia for the postgraduate
scholarship of Farah B. Ahmad. The authors also thank the QUT Central Analytical
Research Facility for its support on sample analyses.
3.1.6 Reference
[1] Subramaniam, R., Dufreche, S., Zappi, M., and Bajpai, R., Microbial lipids from
renewable resources: production and characterization. Journal of Industrial
and furans (furfural and 5-hydroxymethylfurfural (HMF)) were analysed using the
same HPLC system equipped with a Aminex HPX-87H column (300 mm × 8.0 mm,
Bio-Rad, US) and the RI detector [23]. The column temperature was 65 °C and the
mobile phase was 5 mM H2SO4, with a flow rate of 0.6 mL/min.
For the determination of fatty acids composition, fatty acid methyl esters
(FAMEs) were prepared using oil derivatisation method as described by Mulbry et
al. [24]. FAME analysis was performed by gas chromatography-mass spectrometry
(GC-MS) by Shimadzu GCMS-TQ8040 (Shimadzu Corporation, Japan) on an Rtx®-
2330 column (60 m long × 0.25 mm I.D. × 0.2 µm film thickness; Restek, USA).
The carrier gas was helium at a flow rate of 1.5 mL/min. A 10:1 split injection was
used. The injection temperature was set at 250 °C, the MS ion source temperature at
220 °C and the MS interface temperature at 240 °C. The GC-MS method was carried
out using the following temperature program: initial temperature at 90 °C, hold for 2
min, followed by 7.5 °C/min ramp to 210 °C and 20 °C/min ramp to 240 °C, hold for
5 min. Mass spectrometry was performed using Q3 scan with an m/z 20-650
scanning range. Chromatograms and mass spectra were evaluated using the
GCMSsolution software (Shimadzu Corporation, Japan). The retention times and
mass spectra were identified using FAME mix (F.A.M.E. Mix, C8-C24; Sigma-
Aldrich, Australia).
Fuel properties were assessed using cetane number, kinematic viscosity at 40
°C, higher heating value and iodine value. Cetane numbers of FAMEs of microbial
oils were calculated based on the empirical equation proposed by Ramírez-Verduzco
et al. [25], where the cetane number of each fatty acid methyl ester was calculated
using the following equation
∅ 71. 8 0.302 20 (1)
89
From Equation (1), ∅ is the cetane number of the ith FAME, is the molecular
weight of the ith FAME and N is the number of double bonds in a given FAME. The
kinematic viscosity of biodiesel at 40 °C was calculated based on the following
equation
ln 12.503 2.496 ln 0.178 (2)
where is kinematic viscosity at 40 °C (mm2/s) of the ith FAME [25]. Iodine values
of FAME were calculated based on the determination of iodine values of individual
FAME by Krisnangkura [26].
Total nitrogen in EFB hydrolysate was analysed using TOC-VCSH (Shimadzu
Corporation, Japan) with TNM-1 (Total Nitrogen Unit) (Shimadzu Corporation,
Japan). Carbon to nitrogen (C/N) ratio (mass/mass) was calculated using the
following equation:
/ , .
. (3)
where the unit for concentrations of total carbon sources and nitrogen were in g/L,
0.4 was the mass fraction of carbon in the carbon sources (g/g) and 0.8 was mass
fraction of nitrogen in yeast extract (g/g). The oil yield (mg/g) was calculated by
dividing the oil concentration (mg/L) with the total glucose and xylose and acetic
acid consumed (g/L).
Technical and economic assessment of microbial oil production
Microbial cultivation efficiency was estimated based on the theoretical oil
yields of 320 mg/g for glucose and 340 mg/g for xylose [12, 27]. Therefore, % oil
conversion efficiency was calculated according to the following equation,
%
/
/ 100 (4)
where the total oil production from 1 tonne EFB was the summation of predicted oil
production from EFBLH and EFBEH by M. plumbeus. The theoretical oil production
from 1 tonne EFB was calculated using theoretical oil yields on glucose and xylose,
based on the total sugars contents in 1 tonne (dry weight) EFB from EFBLH and
EFBEH.
90
Assuming oil conversion efficiencies of other oil palm biomasses (OPT and
OPF) were the same as those for EFB, the predicted oil production from other oil
palm biomasses of per 1 tonne (dry weight) biomass was calculated as
%
1 10 (5)
Potential microbial oil production per ha (kg/ha) for EFB and other oil palm biomass
(OPT and OPF) was calculated using the following equation
/
1 /
/ (6)
For the economic assessment of large scale microbial oil production, the
relative feedstock cost (US$/kg oil) is calculated by dividing the selling price of
feedstock per tonne biomass (US$/t biomass) by the estimated quantity of microbial
oil produced per 1 tonne biomass (kg/t biomass), as
$/
$/
/ (7)
4.1.3 Results and discussion
4.1.3.1 Chemical composition of feedstock samples
Table 4-1 shows the biomass composition of raw EFB and the solid residue
following pretreatment. Raw EFB consists of 38.8% glucan, 22.4% xylan, 27.2%
lignin. The biomass compositions of EFB solid residues after pretreatment were
45.5% glucan, 6.1% xylan and 40.0% lignin (Table 4-1), with cellulose digestibility
at 43.2%. These results demonstrated that dilute acid pretreatment removed the
majority of hemicellulose (xylan) from the solid biomass as xylan dissolved in the
liquid fraction of hydrolysate [10]. In this study, hydrolysates of liquid fraction and
solid residue from pretreatment were used as substrates for microbial cultivation. The
compositional analyses of EFB hydrolysates are displayed in Table 4-2. EFBLH
91
mainly consists of xylose, whereas the major component of EFBEH is glucose. The
concentration of organic acids (i.e., formic acid, acetic acid and levulinic acid) and
furans (i.e., HMF and furfural) are shown in Table 4-2. Organic acids and furans are
the potential growth inhibitors to microbial cultivation, as reported in numerous
studies [7, 28-30]. HMF and furfural have been shown to reduce yield and inhibit
growth in ethanol production, as well as oil production [28-30]. It has been reported
that furfural concentrations above 1.0 g/L resulted in negative impacts on growth and
oil production of the fungal strain, Mortierella isabellina and the yeast strain
Cryptococcus curvatus [29, 30]. Organic acids or weak acids has been shown to
inhibit cell growth and reduce ethanol yield during the fermentation for ethanol
production [28]. However, the presence of formic, acetic and levulinic acids at a low
concentration in the fermentation for ethanol production has been shown to increase
the yield of ethanol [28]. There are, however, limited studies on the inhibitory effect
of organic acids for microbial oil production.
Table 4-1 Chemical compositions of raw and pretreated EFB. The composition of lignin was based on the composition of acid soluble and acid insoluble lignin. Sample Compositions (%, w/w)
Glucan Xylan Lignin Raw EFB a 38.8 ± 0.0 22.4 ± 0.0 27.2 ± 0.0 Pretreated EFB 45.5 ± 0.5 6.1 ± 0.1 40.0 ± 0.1 a consisted of 5.1% ash, 10.1% water extractives and 3.6% ethanol extractives
In this study, no growth was observed from the cultivation of non-detoxified
EFBLH, most likely due to the presence of high concentrations of furfural in the
hydrolysate. The detoxification step by overliming on the liquid fraction of
pretreatment process was shown to reduce the concentration HMF in the hydrolysate
by approximately two times, and the concentration of furfural was reduced by five
times. Furans present at a low concentration in EFBEH as the solid residue was
washed prior to enzymatic hydrolysis. The washing step is also important as it may
reduce the number of detoxification steps in the hydrolysates preparation process.
The loss of sugars in the detoxified EFBLH was observed at 21.4% for glucose and
7.4% for xylose. Yu et al. also showed 13-29% sugar loss in wheat straw
hydrolysate after detoxification using the overliming process [10]. The concentration
of acetic acid in EFBLH was higher than in the original non-detoxified EFBLH
92
possibly due to water loss from the hydrolysate during the overliming process. This
trend was also seen in the results of the study by Yu et al. where the acetic acid in the
hydrolysate increased from 4.0 g/L to 4.2 g/L after the detoxification process [10].
Table 4-2 Sugars (glucose, xylose and arabinose), organic acids (formic acid, acetic acid and levulinic acid) and furans (5-hydroxymethylfurfural (HMF) and furfural)
compositions of liquid (EFBLH) and enzymatic (solid residue) (EFBEH) hydrolysates of EFB.
Feedstock Non-detoxified EFBLH EFBLH EFBEH
Glucose (g/L) 0.42 0.33 17.42
Xylose (g/L) 5.28 4.89 2.91
Arabinose (g/L) 1.33 1.2 0.61
Formic acid (g/L) 1.07 0.82 0.5
Acetic acid (g/L) 7.06 7.89 2.61
Levulinic acid (g/L) 0.13 0.07 0
HMF (g/L) 0.21 0.08 0.05
Furfural (g/L) 2.89 0.56 0.24
4.1.3.2 Sugars consumption and microbial growth on EFB hydrolysates
In this study, R. mucilaginosa, A. oryzae and M. plumbeus were grown on
liquid hydrolysates and enzymatic hydrolysates resulting from the pretreatment of
EFB. The consumption of sugars by R. mucilaginosa, A. oryzae and M. plumbeus are
shown in Figure 4-2. Complete consumption of glucose was observed for almost all
strains on all hydrolysates. Figure 4-2 also shows lower consumption of xylose
compared to glucose in EFBLH. This is most likely due to the slower consumption
rate of xylose by microorganisms compared to the consumption rate of glucose, as
seen in Ahmad et al. [17]. The amount of xylose consumed in EFBEH was lower
than the xylose consumed in EFBLH. This is probably because xylose consumption
only commenced after glucose was almost depleted in the medium, as shown in
numerous studies [31]. For instance, xylose consumption by Trichosporon cutaneum
ACCC 20271 commenced at 120 h after glucose was almost completely consumed in
enzymatic hydrolysate of corncob residue [13]. Sequential sugars consumption is
common in the microbial cultivation of the mixture of glucose and xylose, due to the
catabolite repression mechanism by glucose or allosteric competition for sugar
93
transporters [32]. A number of studies have been conducted on the simultaneous
consumption of glucose and xylose, as co-consumption of sugars in lignocellulosic
hydrolysates is crucial for comprehensive utilisation for the conversion
lignocellulosic biomass to microbial oil [31]. For the cultivation on EFBEH of all
microorganisms, there were ~40% (w/w) of residual xylose remaining in the media at
the end of cultivation. In this study, M. plumbeus had the lowest consumption of
xylose from EFBLH in comparison to R. mucilaginosa and A. oryzae, possibly due to
its low xylose assimilation capacity. This trend was in agreement with the previous
study by Ahmad et al. [17]. Longer cultivation times may allow more complete
consumption of xylose by these microorganisms in glucose-rich media like EFBEH.
However, prolonged cultivation may not be economical depending on the oil
productivity from the residual sugars.
For cultivation on EFBLH, the biomass concentrations of R. mucilaginosa, A.
oryzae and M. plumbeus were at 5.81 (±0.32), 10.57 (±0.61) and 9.35 (±1.20) and
g/L respectively. The growth on EFBEH showed higher biomass concentrations from
fungal strains, A. oryzae and M. plumbeus at 12.0 and 12.6 g/L respectively, in
comparison to the biomass concentration of R. mucilaginosa at 11.3 g/L. It was also
noted that although fungi strains A. oryzae and M. plumbeus consumed less sugars
from EFBLH than R. mucilaginosa, the biomass concentrations of fungi strains,
however, were higher. The biomass concentrations of fungi strains on EFBLH were
higher even at lower sugars consumption possibly due to the consumption of acetic
acid by the microorganisms. A. oryzae showed acetic acid consumption of 97% and
M. plumbeus of 80% by the end of cultivation. Several studies showed that acetate
was effectively metabolised for oil production by oleaginous microorganisms such as
M. isabellina, C. curvatus and Y. lipolytica Po1g [30].
The biomass concentrations of the fungal strains obtained in this study from
EFBEH are comparable to the results of Ruan et al. on the cultivation of Mortierella
isabellina from enzymatic hydrolysate of corn stover with biomass production at
16.8 g/L [15]. The hydrolysate of corn stover was pretreated with dilute acid and
alkali pretreatments prior to enzymatic hydrolysis, and the hydrolysate consists of
22.2 g/L of glucose and 12 g/L of xylose [15]. Overall, this study shows that all three
selected strains have the capacity to grow on EFBLH and EFBEH and are able to
94
grow in EFB hydrolysates even with the presence of 0.08 g/L of HMF and 0.56 g/L
of furfural.
Figure 4-2 Sugars consumption of R. mucilaginosa, A. oryzae and M. plumbeus on EFB liquid (EFBLH) and enzymatic (EFBEH) hydrolysates.
Glucose consumption is represented by (R. mucilaginosa, RM), (A. oryzae, AO) and (M. plumbeus, MP). Xylose consumption is represented by (R.
mucilaginosa, RM), (A. oryzae, AO) and (M. plumbeus, MP).
4.1.3.3 Microbial oil production from EFB hydrolysates
The oil contents results are presented in Figure 4-3(a). The fungal strain M.
plumbeus showed the highest oil content on EFBLH at 19.8%. For the cultivation on
EFBEH, the highest oil contents recorded were by A. oryzae and M. plumbeus at
~37%. Both fungal strains A. oryzae and M. plumbeus showed higher oil
accumulation on EFBEH than EFBLH. One possible reason for this is that EFBEH
has higher C/N ratio (47.0) than EFBLH (10.2) since EFBEH contains a higher
carbon substrate concentration than EFBLH. The C/N ratio has been identified as the
most important factor affecting lipid accumulation by oleaginous microorganisms
[33]. The presence of HMF and furfural in EFBLH may also be a contributing factor
to lower oil accumulation. Even though numerous studies show that inhibitory
compounds have more damaging effects on growth than oil accumulation, Zhang et
al. argued that inhibitory compounds may have a negative impact on oil
RM RMAO AOMP MPRM
RM
AO
AO
MP
MP
0
10
20
30
40
50
60
70
80
90
100
EFBLH EFBEH
% c
onsu
mpt
ion
95
accumulation as well [7, 34]. This is because any negative effect preventing
microorganisms from reaching the non-growth phase where lipid accumulation
usually occurs, will be detrimental to lipid production as well [34]. Different types of
hydrolysates, however, did not affect oil accumulation by R. mucilaginosa, possibly
because of the lower capacity of R. mucilaginosa to accumulate oil. Ahmad et al.
showed that R. mucilaginosa resulted in a maximum oil content of ~22% on 30 g/L
glucose as well as on glycerol [17].
The oil contents of fungal strains A. oryzae and M. plumbeus obtained in this
study from the cultivation on EFBEH were much higher than the oil contents found
in the cultivation by Ahmad et al., using the pure sugar substrates at higher
concentrations which were 30 g/L glucose (at ~26% of oil) and 30 g/L xylose (~20-
23% of oil) [17]. Higher oil accumulation from cultivation on EFBEH is possibly
due to the lower yeast extract concentration used in this study compared to the study
by Ahmad et al. [17], and therefore resulted in a higher C/N ratio. Based on the total
nitrogen analysis, EFBEH consisted of 0.18 g/L nitrogen, in contrast to 0.36 g/L
nitrogen (based on the use of 4 g/L yeast extract) in the cultivation media in Ahmad
et al. [17]. Therefore, the C/N ratio of EFBEH was 47.0, whereas the C/N ratio of the
cultivation media in Ahmad et al. was 35.4.
Figure 4-3(b) shows oil concentrations from R. mucilaginosa, A. oryzae and M.
plumbeus cultivated on EFB hydrolysates. Fungal strains, M. plumbeus had recorded
the highest oil concentration growing on hydrolysates of EFB (1.85 ±0.33 and 4.69
±0.44 g/L on EFBLH and EFBEH respectively), followed by A. oryzae (1.40 ±0.60
g/L on EFBLH and 4.47 ±0.40 g/L on EFBEH). Microbial oil concentrations of
fungal strains M. plumbeus and A. oryzae on EFBLH obtained in this study compares
well to Tampitak et al. with oil concentration of 1.61 g/L by C. tropicalis on
hemicellulose hydrolysate of EFB [16]. The EFB hemicellulose hydrolysate was
prepared by alkaline and dilute acid pretreatment, was diluted to 20 g/L sugars,
followed by detoxification [16]. The result of oil concentration of M. plumbeus on
EFBEH is comparable to oil production by M. isabellina at 6.9 g/L from corn stover
hydrolysate by Ruan et al. [15].
96
(a)
(b)
Figure 4-3(a) Oil contents (%, w/w) and (b) oil concentrations (g/L) of yeast R. mucilaginosa, and fungi A. oryzae and M. plumbeus cultivated on EFB liquid
(EFBLH) and enzymatic (EFBEH) hydrolysates.
The oil production by M. plumbeus and A. oryzae on EFBEH in this study was
two times higher than C. tropicalis on residual pulp hydrolysate (2.73 g/L of oil) and
holocellulose hydrolysate (1.31 g/L of oil) of EFB, even though these two
hydrolysates and EFBEH have almost similar concentrations of sugar of 20 g/L [16].
The oil concentration of yeast R. mucilaginosa (2.17 g/L of oil) on EFBEH is similar
to the yeast strain used by Tampitak et al. (C. tropicalis) on EFB pulp hydrolysate.
The EFB pulp hydrolysate was pretreated with alkaline and dilute acid, followed by
acid hydrolysis, whereas the EFB holocellulose hydrolysate was prepared through
alkaline pretreatment and acid hydrolysis [16]. Both hydrolysates were then
subjected to dilution and detoxification [16]. Low oil production from EFB
holocellulose hydrolysate by Tampitak et al. in comparison to EFB pulp hydrolysate
0
5
10
15
20
25
30
35
40
45
EFBLH EFBEH
Oil
con
tent
(%
, w/w
)
R. mucilaginosa
A. oryzae
M. plumbeus
0
1
2
3
4
5
6
EFBLH EFBEH
Oil
conc
entr
atio
n (g
/L)
R. mucilaginosa
A. oryzae
M. plumbeus
97
was not discussed, even though both hydrolysates have similar glucose and overall
sugars concentrations. One possible reason for this is that there was a slightly higher
concentration of furfural (0.17 g/L) in EFB holocellulose hydrolysate compared to
EFB pulp hydrolysate (0.13 g/L furfural). Another reason for the variation in oil
production is possibly contributed by the concentration of organic acids in the
hydrolysates, which was not discussed. Overall, both fungal strains M. plumbeus and
A. oryzae showed good potential to produce oil from EFB hydrolysates, with oil
concentrations similar to the widely researched fungal strain M. isabellina.
Table 4-3 shows the results of oil concentrations and oil yields by a number of
oleaginous microorganisms from various agro-industrial wastes. Oil yield is an
important parameter for the cultivation as it measures the efficiency of
microorganisms to convert carbon substrates to oil. In this study, M. plumbeus
produced the highest oil concentrations and oil yields from both EFB hydrolysates in
comparison to A. oryzae and R. mucilaginosa. The cultivation of R. mucilaginosa
resulted in lower oil yields on EFBLH (64 mg/g) and EFBEH (93 mg/g). Lower
conversion efficiency of carbon substrates to oil in R. mucilaginosa was most likely
due to the assimilation of carbon substrates for the production of carotenoids, as
Rhodotorula yeasts are well-known carotenoids producers [35]. The oil yield of M.
plumbeus on EFBLH (185 mg/g) is slightly lower than EFBEH (231 mg/g), possibly
due to the presence of inhibitors in EFBLH. The oil yields of fungal strains on
EFBLH in this study are higher than the oil yield (80 mg/g) of C. tropicalis grown on
hemicellulose hydrolysate (liquid fraction of pretreatment) of EFB [16]. The oil yield
of M. plumbeus on EFBLH is comparable to the oil yield (172 mg/g) of
Trichosporon coremiiforme grown on detoxified corncob hydrolysate (2.9 g/L
glucose and 37.9 g/L xylose, 0.32 g/L HMF and 0.06 g/L furfural) [14], where both
hydrolysates contained a higher proportion of xylose than glucose. Table 4-3 also
shows that oil yields of M. plumbeus on lignocellulosic biomass are comparable to
those of other microorganisms, such as the more widely researched yeast strain C.
curvatus (140 mg/g on corncob liquid hydrolysate) and fungal strain M. isabellina
(147 mg/g on corn stover enzymatic hydrolysate) [10, 15].
98
Table 4-3 Biomass concentrations, oil concentrations and oil yields of different oleaginous yeasts and filamentous fungi from batch fermentation
of various hydrolysates of lignocellulosic agro-industrial wastes. Oil yields are expressed as mg oil produced per g sugars consumed, unless mentioned otherwise.
Feedstock Strains Biomass concentration (g/L)
Oil concentration (g/L)
Oil yield (mg/g)
Lignocellulosic biomass processing
Liquid hydrolysate from pretreatment process Corncob Rhodotorula
glutinis 15.1 5.5 130 * Dilute acid
pretreatment [5]
Corncob Trichosporon coremiiforme
20.4 7.7 172 Dilute acid pretreatment [14]
Wheat straw
Cryptococcus curvatus
17.2 5.8 140 a Dilute acid pretreatment [10]
Wheat straw
Mortierella isabellina
5.9 2.3 123 Dilute acid pretreatment [12]
EFB Candida tropicalis
6.4 6.7 80 Dilute alkaline and dilute acid pretreatment [16]
EFB R. mucilaginosa
5.8 0.8 64 a This study
EFB A. oryzae 10.6 1.4 110 a This study
EFB M. plumbeus 9.4 1.9 185 a This study
Solid residue hydrolysate Corn stover
Mortierella isabellina
18.7 6.9 147 Dilute alkaline and dilute acid pretreatment, then enzymatic hydrolysis of pretreatment slurries [15]
Corncob Trichosporon cutaneum
38.4 12.3 131 Enzymatic hydrolysis of residue, pretreatment was not specified [13]
EFB R. mucilaginosa
11.3 2.2 93 a This study
EFB A. oryzae 12 4.5 199 a This study
EFB M. plumbeus 12.6 4.7 205 a This study
* Data is not provided, complete reducing sugars consumption was assumed a Oil concentration (g/L) per reducing sugars consumed as well as acetic acid consumed
99
4.1.3.4 Fatty acid profiles and potential application of microbial oils produced
from EFB hydrolysates
Microbial oils have the potential to be utilized as alternative feedstocks for
biodiesel production, depending on the fatty acids profile of the microbial oil.
Biodiesel is a renewable fuel, typically made from vegetable oils and their
derivatives, especially methyl esters [36]. Microbial oils with similar fatty acids
compositions to vegetable oils have the potential to be used for biodiesel production.
The fatty acids profiles of transesterified microbial oils produced in this study
are shown in Table 4-4. The results showed that microbial oils have similar fatty
acids composition to vegetable oils, with the major fatty acids identified being
palmitic (C16:0), stearic (C18:0), oleic (C18:1) and linoleic acid (C18:2). For R.
mucilaginosa, oleic acid was the predominant fatty acid produced on EFBLH (52.9
%), which is in agreement with the results of Ahmad et al. in the cultivation of R.
mucilaginosa on glucose and xylose [17]. However, palmitic acid was the
predominant fatty acid of oil extracted from R. mucilaginosa grown on EFBEH (47.2
%). Linoleic acid was the predominant fatty acids for both A. oryzae and M.
plumbeus cultivated on EFBLH (36.9% and 39.4% respectively), followed closely by
oleic acid. For the growth of both A. oryzae and M. plumbeus on EFBEH, oleic acid
was identified as the predominant fatty acid at ~33 - 34%, which is in accordance
with the findings from Ahmad et al. on the cultivation of A. oryzae and M. plumbeus
on glucose and xylose [17].
Fuel properties of microbial oils can be further analysed in order to determine
the suitability of microbial oils to be utilised for the production of biodiesel. Among
the most important physical properties of biodiesel are cetane number, kinematic
viscosity at 40 °C, higher heating value and iodine value. Cetane number is a relative
measure of the ignition quality of fuels, whereby low cetane number fuels have the
tendency to increase gaseous and particulate exhaust emissions due to incomplete
combustion [25, 37]. Methyl esters of microbial oils in this study have estimated
cetane numbers to be greater than 59 (Table 4-4), which is above the required
minimum values of 47 and 51 according to the biodiesel specifications of ASTM
6751-08a and EN14214 respectively [25, 38].
100
The viscosity of biodiesel is important as high viscosity will cause poor
atomisation in the combustion chamber and subsequently lead to engine problems
like nozzle choking and engine deposits [36]. Kinematic viscosities at 40 °C of
microbial oils methyl esters were shown in Table 4-4 and the kinematic viscosity of
all microbial oils were within the limit of EN 14214 (3.5 to 5.0 mm2/s) [39]. The
viscosities of fatty acids ester that has been transesterified is typically lower than the
viscosities of fatty acids [36].
Table 4-4 Fatty acids composition of microbial oils methyl ester of R. mucilaginosa, A. oryzae and M. plumbeus cultivated on EFB liquid (EFBLH) and solid residue
enzymatic (EFBEH) hydrolysates, as well as fuel properties, (cetane number, kinematic viscosity, higher heating value and iodine number) of
transesterified microbial oils. Microorganisms Relative abundance of fatty acid
Polyunsaturated fatty acids, especially those with four or more double bonds,
are not preferable for biodiesel production due to its low oxidative stability [40].
However, biodiesel with higher amounts unsaturated fatty acids has a higher cold
filter plug point (CFPP) [38]. The limit of degree of unsaturation of oil suitable for
biodiesel production can be determined through its iodine value. Iodine value is
defined as the number of centigrams of iodine absorbed per gram samples, in which
fuels with higher iodine values have higher degrees of unsaturation [36]. The
estimated iodine values of each methyl ester (Table 4-4) are below the biodiesel
101
standard maximum limit of 115 (German biodiesel standard DIN V 51606) and 120
(EN14214) [36, 38]. Therefore, this study shows that microbial oils produced are
suitable as biodiesel fuels.
4.1.3.5 Techno-economic evaluation of microbial oil production from oil palm
biomasses
Technical assessment of microbial oil production from oil palm biomasses
In this study, microbial oil had been successfully produced from EFB, the
wastes of palm oil production, through the utilisation of hydrolysates from both
fractions of pretreatment, by yeast R. mucilaginosa and fungi A. oryzae and M.
plumbeus. As the cultivation on M. plumbeus resulted in the highest oil
concentrations and oil yields from all hydrolysates, the data of M. plumbeus was
further used for the techno-economic evaluation of microbial oil production from oil
palm biomass.
The prospect of using M. plumbeus for large-scale microbial oil production can
be evaluated by estimating the amount of oil that could be produced from oil palm
biomasses based on the oil yields [12]. The oil yields of M. plumbeus on EFBEH and
EFBLH in this study, based on consumed sugars, are 205 and 185 mg/g respectively.
The results of sugar yields from the study on optimised dilute acid pretreatment of
EFB by Jung et al. was used for estimating microbial oil production from EFB [41].
For microbial oil production from 1 t dry EFB (249.5 kg glucose and 23.9 kg xylose)
[41], 56 kg microbial oil can be produced from enzymatic hydrolysate of the solid
residue. On the other hand, liquid hydrolysate from 1 t dry EFB (68 kg glucose and
135 kg xylose) [41] can potentially yield 37.6 kg of microbial oil. The total oil
production from 1 t dry EFB is estimated to be 93.6 kg. The estimated microbial oil
produced from 1 t EFB is comparable to 103 kg of estimated microbial oil produced
per tonne wheat straw by fungal strain M. isabellina [12]. The theoretical production
of microbial oil from solid residue hydrolysate is estimated to be 155.6 kg oil per 1 t
dry EFB. Based on the microbial oil production by M. plumbeus from EFB in this
study, 60% oil of the theoretical microbial oil production (i.e., 60% oil conversion
efficiency (Equation 4)) was achieved. However, higher oil yields can likely be
obtained with the use of the bioreactor for cultivating M. plumbeus for large-scale
microbial oil production.
102
As the second largest producer of palm oil in the world, Malaysia produces a
substantial amount of EFB from palm oil mills. It is estimated that 1.38 t EFB (dry
weight) is produced per hectare from oil palm plantations [3]. Therefore, the
potential microbial oil production from palm oil EFB is estimated to be 129 kg/ha.
The quantity is approximately 3.4% of 3.84 t/ha average crude palm oil produced in
Malaysia in 2014 (Table 4-5(a)).
Table 4-5 (a) The summary of technical evaluation of microbial oil production from oil palm biomasses (EFB, trunk (OPT) and frond (OPF)) through the comparison of
potential microbial oil yields per hectare to oil yield of crude palm oil. Biomass availability was based on the Malaysian palm oil sector. (b) The summary
of economic evaluation of oil production from oil palm biomasses through the comparison of the relative feedstock cost of oil production from oil palm biomasses and the feedstock cost of crude palm oil production from fresh fruit bunch (FFB).
(a) Oil palm biomass/ Palm oil
Biomass availability (t/ha)
Sugar content per 1 t biomass (kg/t)
Oil yield per hectare (t/ha)
EFB 1.4 476 0.14
OPT (bagasse) 4.2 435 0.35
OPF (basal) 15.3 164 0.48
Crude palm oil - - 3.84
(b) Feedstock for oil production
Selling price (US$/t)
Oil yield per 1 t biomass (kg/t)
Feedstock cost (US$/kg oil)
EFB 15.80 96.3 0.16
OPT (bagasse) 15.00 83.5 0.17
OPF (basal) 25.00 31.5 0.79
FFB 123.00 206.1 0.60
The total oil production can be increased even further through the use of other
oil palm biomasses, such as oil palm frond (OPF), oil palm trunk (OPT), palm kernel
shell (PKS) and mesocarp fibre (MF), for microbial oil production. It is estimated
that there was 87 million t oil palm solid wastes (dry weight) generated in Malaysia
and these wastes are expected to increase to 100 million t by 2020 [3]. All of these
oil palm biomasses identified are lignocellulosic, and potentially can be used as the
feedstock for microbial oil production. From palm oil processing, MF and PKS were
103
produced annually at around 1.42 t/ha and 0.85 t/ha respectively [3]. From oil palm
plantations, there are approximately 14 t/ha OPT and 46 t/ha pruned OPF generated
in Malaysia [3]. OPF and OPT have been tested previously as the carbon substrates
for ethanol fermentation [42, 43]. Therefore, the opportunity exists for utilising the
fermentable sugars from OPF and OPT for microbial oil cultivation. The use of these
biomasses as the additional feedstocks will enhance the total oil yield from a
microbial oil production plant, as well as improve the economics of oil production.
Therefore, in this study, the prospect of microbial oil production from OPF and OPT
were also included in the techno-economic evaluation. However, there was no
sufficient information to perform similar evaluation on MF and PKS. The technical
evaluation for OPF and OPT was completed based on Equation 5 and Equation 6.
OPT can be utilised for microbial oil production from the use of OPT bagasse,
which is the remains of OPT that is squeezed for sugar juice production
(approximately 30% of wet felled OPT) [43]. It is estimated that 4.2 t/ha of dried
OPT bagasse is produced annually in Malaysia. The glucose yield of 435 kg/t OPT
bagasse can be obtained from the enzymatic hydrolysis of cellulose and starch of
OPT bagasse [43]. On the basis of 60% cultivation efficiency, the potential microbial
oil production is estimated to be 0.35 t/ha. On the other hand, OPF also has the
potential to be used for oil production as it contains 164 kg glucose per 1 t OPF (dry
weight) [42]. Assuming that the top two-thirds of OPF with leaflets was re-used as
fertiliser and only the basal (lower third) of OPF was utilised as feedstock of oil
production, there is 15.3 t/ha of OPF available for oil production. Therefore, OPF is
estimated to yield microbial oil up to 0.48 t/ha. Table 4-5(a) showed that oil palm
biomasses have the potential to produce approximately 0.97 t oil per hectare, which
has the potential to increase total oil production by 25% from the same plantation
area.
Economic assessment of microbial oil production from oil palm biomasses
The relative feedstock cost (US$/t) (Equation 7) was used to economically
evaluate the feasibility of microbial oil production from oil palm biomasses. The
feedstock cost of oil production from EFB is estimated based on the selling price of
EFB of $15.80/t [44], where the feedstock cost is $0.17/kg oil. Taking the selling
price of OPF at $25/t from Zahari et al. (based on transportation cost, harvesting and
104
collection cost and pre-processing cost) [45], the feedstock cost of oil production
from OPF is $0.79/kg oil. The selling price of OPT bagasse was estimated to be $15/t
(based on transportation and pre-processing cost from Zahari et al.). Therefore the
feedstock cost of oil production from OPT is $0.18/kg oil. On the other hand, based
on the average 2014 yield of FFB at 18.63 t/ha (selling price of $123/kg) and the
yield of crude palm oil at 3.84 t/ha [46], the feedstock cost of crude oil production
from FFB is estimated to be at $0.60/kg oil. To conclude, the feedstock cost of oil
production from the current palm oil sector can be lowered through the use of
alternative feedstocks for oil production such as oil palm biomasses, especially EFB
and OPT for oil production (Table 4-5(b)).
The feasibility of increasing oil production in the palm oil industry from EFB and other oil palm biomasses
From the technical assessment discussed in preceding subsections, there is
great potential to increase the oil production in the palm oil industry through the
utilisation of processing wastes from the industry. In addition, the economic
assessment demonstrated that the integration of oil palm biomasses with existing
palm oil processing, can potentially provide cheaper feedstock cost for oil
production. The proposed process integration of microbial oil production from oil
palm biomasses is illustrated in Figure 4-4, where the microbial oil production is
incorporated into the existing palm oil industry from plantation harvesting and oil
extraction to products manufacturing.
105
Figure 4-4 Process flow of proposed integration of microbial oil production from oil palm biomasses into the existing palm oil processes.
The flow diagram of palm oil production processes was drawn based on life cycle assessment on crude palm oil production in Malaysia [47] (OPF, oil palm frond;
OPT, oil palm trunk; FFB, fresh fruit bunch).
106
Microbial oil produced from oil palm biomasses can potentially be utilised to
supplement existing palm oil production, either for biodiesel production from non-
food feedstock, as well as manufacturing oleochemicals and food products depending
on the fatty acids composition of microbial oils. Microbial oils with high
compositions of monounsaturated and saturated fatty acids are suitable for biodiesel
production, whereas microbial oils that are rich in palmitic and stearic acid are
compatible to crude palm oil and can be integrated with the palm oil refinery for food
and oleochemicals production. The proposed integration can improve the
sustainability of palm oil processing, where the lignocellulosic by-products from
palm oil processes are recycled back to the industry. The proposed integration also
showed that glycerol, the by-product of transesterification process for biodiesel
production, can be re-utilised for microbial oil production which is significant for the
biorefinery approach. Ahmad et al. demonstrated that M. plumbeus could grow and
produce oil from glycerol [17]. In addition, the proposed integration can enhance the
profitability of palm oil processing, with the potential increase of oil production in
the palm oil industry from cheaper feedstocks, through microbial production from
the lignocellulosic by-products of palm oil processes.
4.1.4 Conclusion
Overall, this study demonstrated the biochemical conversion of pretreated oil
palm EFB to oil through microbial cultivation using R. mucilaginosa, A. oryzae and
M. plumbeus. The hydrolysates from both fractions of pretreatment (liquid fraction
and solid residue) were utilised for the microbial cultivation for maximising the total
microbial oil production from EFB. M. plumbeus showed the highest oil
concentrations and the highest oil yields on both hydrolysates of EFBLH and
EFBEH. The prospect of increasing oil production in the palm oil industry from EFB
and other oil palm biomasses (OPT and OPF) was evaluated technically in terms of
the oil production of M. plumbeus per hectare, and was further assessed
economically through the relative feedstock cost for oil production. The assessments
show that microbial oil production from oil palm biomasses have the potential to
increase existing oil production in the palm oil industry by up to 25% with lower
feedstock cost for oil production. The integration of microbial oil production from oil
107
palm biomasses into the palm oil industry can enhance the sustainability and
profitability of the industry.
4.1.5 Acknowledgements
The authors acknowledge the Ministry of Higher Education Malaysia for the
postgraduate scholarship of Farah B. Ahmad. The authors also thank the QUT
Central Analytical Research Facility for its support on sample analyses, as well as
Vitor Takashi Kawazoe for oil extraction and derivatisation process.
4.1.6 Reference
[1] FAOSTAT. Food and Agriculture Organization of The United Nations Statistics
Division. Browse Data 2013; Available from: http://faostat3.fao.org/home/E.
[2] Mohammed, M.A.A., et al., Hydrogen rich gas from oil palm biomass as a
potential source of renewable energy in Malaysia. Renewable & Sustainable Energy
Reviews, 2011; 15: 1258-1270.
[3] AIM, National Biomass Strategy 2020: New wealth creation for Malaysia’s
biomass industry Version 2.0 ed. 2013, Malaysia: Agensi Inovasi Malaysia.
[4] Chiew, Y.L. and Shimada, S., Current state and environmental impact
assessment for utilizing oil palm empty fruit bunches for fuel, fiber and fertilizer – A
case study of Malaysia. Biomass and Bioenergy, 2013; 51: 109-124.
[5] Liu, Y., Wang, Y., Liu, H., and Zhang, J.a., Enhanced lipid production with
undetoxified corncob hydrolysate by Rhodotorula glutinis using a high cell density
Figure 5-1 (a) Mathematical models for oil concentration (Y1) and oil yield (Y2), and the parameters of analysis of variance (ANOVA) of each model. (b-c) Plots of
predicted vs. actual values (experimental data). (d-e) Internally studentised residuals vs actual values.
Table 5-3 Analysis of variance (ANOVA) for the response surface quadratic model of oil concentration (a) and oil yield (b) that had significant terms (Sugar
would increase the oil concentration. Higher sugar concentration leads to a higher
C/N ratio of the cultivation media, where higher a C/N ratio enhances oil
accumulation by oleaginous microorganisms [6]. The synergy of sugar concentration
with yeast extract concentration (X1X2) was significant to oil yield (p-value =
0.0419) (Figure 5-2(b)). From the surface plots of oil yield, lowering the yeast
extract concentration supplemented to the media would lead to higher oil yield at any
concentration of sugars (Figure 5-3(a)).
The yeast extract concentration was shown to be significant to both oil
concentration and oil yield (p-value = 0.0343 and p-value = 0.0007 respectively).
Lowering yeast extract concentration supplied to the cultivation media promotes the
oil production. This study revealed that fungi could grow even without yeast extract
for optimum oil production. The binary interaction of yeast extract and spore
concentration (X2X3) was significant to the oil concentration and oil yield (p-value =
0.0197 and p-value = 0.0031 respectively) (Table 5-3). The surface plots (Figure 5-
2(d) and Figure 5-3(c)) showed that at the maximum yeast extract concentration,
increasing spores inoculum concentration decreases the oil production. This is
possibly because higher yeast extract concentration provides surplus-nitrogen
condition, where this condition was known to stimulate cell proliferation [18] and
subsequently fungal biomass formation. Therefore, a higher spore inoculation to
surplus-nitrogen medium may have caused a rapid formation of biomass in the
culture, which could lead to poor oxygen-mass transfer and subsequently low oil
concentration (Figure 5-2(a, d and e)).
From the response surface analysis, pH (X4) was shown to be significant to oil
concentration (p-value = 0.0013), as well as binary interaction of pH and sugar
concentration (X1X4) (p-value = 0.0002) (Table 5-3(a)). Based on the surface plots
of the interaction between pH and sugar concentration for both oil concentration
(Figure 5-2(c)) and oil yield (Figure 5-3(b)), it can be concluded that the impact of
pH on oil concentration or oil yield varied according to the level of sugar
concentration. From Figure 5-3(b), at the minimal sugar concentration (30 g/L), a
higher oil yield could be achieved by decreasing the value of pH. However, a total
129
opposite correlation applied at maximum sugar concentration of 100 g/L, where
decreasing pH values would decrease the oil yield.
(a) (b)
(c) (d)
(e)
Figure 5-2 (a-e) Three-dimensional surface plots of binary interaction between different variables to the oil concentration. Sugar is sugar concentration, %YE is
relative concentration of yeast extract, Spore is spore concentration and pH is initial pH.
130
(a) (b)
(c) (d)
Figure 5-3 (a-d) Three-dimensional surface plots of binary interaction between different variables to the oil yield. Sugar is sugar concentration, %YE is relative concentration of yeast extract, Spore is spore concentration and pH is initial pH.
The correlation between pH and sugar concentration might be influenced by
acetic acid, three times more than that in EH30. The pKa value of acetic acid is 4.75,
which means that at pH 4.75, the concentration of the un-dissociated and dissociated
forms of acetic acid in the cultivation media were equal [19]. Acetic acid in un-
dissociated form was found to be inhibitory to microbial growth [20]. When the
medium was adjusted to pH 7.0, there would be 99% of acetic acid dissociated into
acetate anions [20], which reduces the inhibitory effect of acetic acid on
microorganisms. Therefore, pH 7.0 was more favourable for oil production on
EH100.
131
The surface analysis showed that the spore concentration (X3) did not have
significant influence on the oil concentration and oil yield (p-value = 0.1390 and p-
value = 0.1477 respectively).
5.1.3.3 Substrates consumption, microbial oil fatty acids composition and other
metabolites accumulation
The impact of different cultivation parameters on microbial growth was
analysed through the sugar consumption behaviour of M. plumbeus on EFB
enzymatic hydrolysates (EHs) (Figure 5-4). From the experimental run performed on
EH30 and EH100 (Figure 5-4(a and e)), the cultivations with the maximum yeast
extract and spore concentration resulted in complete consumption of glucose within
48 h (EH30) and 74 h (EH100), regardless of the initial pH of the media. The
cultivation with yeast extract concentration had a faster consumption rate of sugars in
comparison to the cultivation with no yeast extract supplementation for the
cultivation on EH30 and EH65 (Figure 5-4(a and c)).
The trend of sugar consumption was shown to be correlated to the
concentration of yeast extract (nitrogen concentration). The cultivation with
maximum yeast extract concentration had the fastest consumption of glucose and
xylose. As nutrient (e.g., nitrogen) depletion was shown to negatively impact cell
proliferation and carbon substrates uptake rates [21], the opposite may promote
carbon substrates consumption. Even though the cultivation with maximal yeast
extract supplementation could potentially reduce the number of days of cultivation
due to fast sugar consumption, the oil yields were low in comparison to the
cultivation with no additional yeast extract (Figure 5-3(a, b and d)).
The overall rate of glucose consumption of all cultivation runs was higher
than the consumption rate for xylose. In this study, for all cultivations, M. plumbeus
exhibited sequential sugar assimilation where xylose consumption began after the
majority of glucose was consumed in the media. The ratio of the amount of glucose
to xylose in the media was 12:1. The sequential sugar assimilation was common for
media that contained a higher proportion of glucose than xylose, where the
assimilation pattern could be due to catabolite repression by glucose or allosteric
competition for sugar transporters [22].
132
(a)
(b)
(c)
0
5
10
15
20
25
30
35
0 24 48 72 96 120 144 168
Glu
cose
con
cen
trat
ion
(g/
L)
Time (hour)
min YE, minspore, pH5max YE, minspore, pH5min YE, maxspore, pH5max YE, maxspore, pH5min YE, minspore, pH6max YE, minspore, pH7min YE, maxspore, pH7max YE, maxspore, pH7mean YE, meanspore, pH6
0
0.5
1
1.5
2
2.5
3
0 24 48 72 96 120 144 168
Xyl
ose
con
cen
trat
ion
(g/L
)
Time (h)
min YE, minspore, pH5max YE, minspore, pH5min YE, maxspore, pH5max YE, maxspore, pH5mean YE, meanspore, pH6min YE, minspore, pH6max YE, minspore, pH7min YE, maxspore, pH7max YE, maxspore, pH7
Figure 5-4 The consumption of glucose (a) and xylose (b) for the experimental run on EH30. The consumption of glucose (c) and xylose (d) for the experimental run on
EH65. The consumption of glucose (e) and xylose (d) for the experimental run on EH100.
min YE, minspore, pH5max YE, minspore, pH5min YE, maxspore, pH5max YE, maxspore, pH5min YE, minspore, pH7min YE, maxspore, pH7max YE, maxspore, pH7mean YE, meanspore, pH6
0
1
2
3
4
5
6
7
8
9
10
0 24 48 72 96 120 144 168
Xyl
ose
con
cen
trat
ion
(g/L
)
Time (h)
min YE, minspore, pH5max YE, minspore, pH5min YE, maxspore, pH5max YE, maxspore, pH5min YE, minspore, pH7min YE, maxspore, pH7max YE, maxspore, pH7mean YE, meanspore, pH6
134
The analysis of fatty acids composition showed that the microbial oils consist
of palmitic (C16:0), stearic (C18:0), oleic (C18:1) and linoleic (C18:2) acid, which
was similar to fatty acid compositions reported for the cultivation of other oleaginous
microorganisms [11, 14, 23].
Ethanol was detected in some of the cultivation runs (Table 5-5). Ethanol
accumulation in the cultivation with higher sugar concentrations showed that M.
plumbeus did not possess strict aerobic metabolism. Some Mucor species (e.g.,
Mucor indicus) were utilised for ethanol production from acid hydrolysates of rice
straw and spruce forest residues [24]. In this study, ethanol accumulation was
prevalent in the cultivation with high sugar concentration (~100 g/L). High carbon
substrates loading caused a rapid growth of biomass that could lead to an increase in
the viscosity of the culture and further reduce the efficiency of oxygen-mass transfer
inside the fungal biomass. This phenomenon could result in carbon assimilation
under limiting-oxygen conditions which then led to the production of ethanol.
Ethanol was accumulated most likely at the expense of carbon-to-oil conversion
efficiency, where cultivation with ethanol accumulation of more than 20 g/L
obtained low oil yields (5 mg/g for Run #10 and 36 mg/g for Run #11) (Table 5-5).
The cultivation on EH100 also resulted in higher ethanol yields than oil yields (Table
5-5). Even though ethanol is a valuable co-product, the ethanol yields were too low
in comparison to the theoretical ethanol yield (514 mg/g glucose), where the process
of ethanol recovery could lead to unfavourable economics for the overall cost of
production.
135
Table 5-5 The concentration of ethanol accumulated at the end of the cultivation and ethanol yields per sugars consumed for various enzymatic hydrolysates (EH30,
EH at 30 g/L sugars; EH65, EH at 65 g/L sugars; EH100, EH at 100 g/L sugars). Hydrolysate Run Condition Ethanol
concentra-tion (g/L)
Ethanol yield (mg/g)
Oil yield (mg/g)
EH30 2 Min YE, min spore, pH5 2.99 150 56
23 Max YE, min spore, pH5 0.86 34 53
12 Min YE, max spore, pH5 1.41 62 84
3 Max YE, max spore, pH5 0 0 50
17 Min YE, min spore, pH7 2 67 36
15 Max YE, min spore, pH7 0 0 50
19 Min YE, mean spore, pH7
1.02 35 53
16 Max YE, max spore, pH7 0 0 37
18 Mean YE, mean spore, pH6
1.36 44 42
EH65 1 Min YE, mean spore, pH6
6.66 132 39
26 Mean YE, min spore, pH6
7.88 139 51
9 Mean YE, max spore, pH6
1.92 35 47
8 Mean YE, mean spore, pH7
6.65 115 40
Centre points
Mean YE, mean spore, pH6
8.33 169 36
EH100 21 Min YE, min spore, pH5 11.23 237 59
11 Max YE, min spore, pH5 27.66 268 36
4 Min YE, max spore, pH5 5.28 73 58
10 Max YE, max spore, pH5 25.84 250 5
5 Min YE, min spore, pH7 12.55 131 78
14 Min YE, max spore, pH7 13.83 135 77
22 Max YE, max spore, pH7 6.17 62 41
6 Mean YE, mean spore, pH6
13.04 129 47
136
5.1.3.4 Validating optimisation parameters for oil production from EFB
hydrolysates (EH) by M. plumbeus
The main criteria for finding optimum parameters for oil production are high
oil yield and high oil concentration. The cultivation with maximum sugars (100 g/L)
at pH 7.0 with no additional yeast extract was predicted to give the highest oil
concentrations (7.6 g/L), with predicted oil yield at ~77 mg/g (Table 5-4, Run #14).
The cultivation with minimum sugar concentration and maximum spore
concentration at pH 5.0 without additional yeast extract was predicted to result in the
highest oil yield at 84 mg/g (Table 5-4, Run #12). However, the cultivation run at the
maximum sugar concentration led to the accumulation of more than 5 g/L ethanol
with ethanol yields of more than 50 mg per g sugars consumed (Table 5-5). The
cultivation at 100 g/L also resulted in rapid formation of biomass in high volume
which was a potential challenge for cultivation in the bioreactor. Therefore, the
parameters for optimisation were selected at 30 g/L sugar concentration, pH 5.0 with
spore concentration at 6.3 log spores number/mL medium, without additional yeast
extract. The optimised parameters were predicted to result in an oil concentration of
2.60 g/L and 84 mg/g of oil yield.
The validation experiment resulted in the production of 11.6 g/L biomass and
2.67 g/L oil (oil content of 23.1%), with an oil yield of 94 mg/g. The relative
amounts of sugars consumed from EH after seven days cultivation were 86% of
glucose and 38% of xylose, respectively. The analysis on the fatty acids of microbial
oil from the optimised cultivation revealed that the microbial oil consists of 18.9%
palmitic, 31.0% stearic, 34.7% oleic and 15.5% linoleic acids, which is similar to
fatty acid compositions of M. plumbeus grown in EFB hydrolysates in our previous
study [11]. The predominant fatty acids found in the microbial oil were also similar
to the fatty acid compositions of vegetable oils that has potential for biodiesel
production [25].
137
The cultivation of the control, performed on EH at optimised sugar
concentration (30 g/L), resulted in the production of 13.2 g/L biomass and 2.13 g/L
oil with 16.2% oil content and 52 mg/g oil yield. The resulting oil concentration and
oil yield from the control were lower than those of the optimised cultivation.
However, both glucose and xylose in the cultivation medium of control were
consumed completely by the end of cultivation. The cultivation medium of control
contains a higher nitrogen content and lower C/N ratio than the optimised culture.
The higher nitrogen content may have contributed to a better consumption of the
carbon sources in the comparison study, as sufficient nitrogen supply promotes cell
proliferation. Even with high sugar consumption rates in the comparative study, the
efficiency of carbon substrates conversion into oil was lower than the optimised
cultivation. It is noted that higher sugar consumption rates did not necessarily lead to
a higher oil yield, which provided further evidence on the impact of nitrogen
concentration on oil production as discussed in Section 5.1.3.3.
5.1.3.5 Microbial oil production in bioreactor
The cultivation of M. plumbeus on EH was subsequently performed in a
bioreactor, based on the optimised conditions, to investigate the impact of the
bioreactor system on microbial oil production as well as the prospective of process
scale-up. Figure 5-5(a) shows gradual consumption of sugars throughout the
cultivation. The pH increased slightly at the beginning of the cultivation, but dropped
gradually after 24 h, and reached pH 5.0 at the end of the cultivation. The DO level
was reduced sharply within the first 24 hours as glucose was consumed rapidly at
this phase of the cultivation (Figure 5-5(b)). Due to DO control in the bioreactor
system, the DO level was maintained at 20% by increasing the agitation speed from
its initial speed at 200 rpm (minimum setpoint) to 335 rpm (maximum setpoint) by
41 h. The agitation speed was reduced afterward to the original speed at 200 rpm,
corresponding to the raising DO level, and remained at 200 rpm until the end of
cultivation.
For the first two days of the cultivation, fungal biomass only grew in pellets.
The transition in fungal morphology in the bioreactor became apparent after three
days of cultivation with the appearance of dispersed mycelia mixed with fungal
pellets (Figure 5-6). The transition in morphology, from pellet-only biomass to
138
mixture of pellet and dispersed mycelia, was most likely attributed to the high
agitation speed. Filamentous fungus Mortierella isabellina in dispersed mycelial
form was reported to obtain a higher oil yield and oil content in comparison to fungi
in pellet form [5]. This is because mycelia aggregates possibly had better oxygen and
nutrient intake due to lower biomass density and smaller radius [5]. One of the
challenges of scaling up fungal cultivation is the tendency of fungal biomass in
bioreactor to aggregates and grow on the wall (Figure 5-6(d)) and the sensor probes
of bioreactor [24].
(a)
(b)
Figure 5-5 (a) The consumption of glucose and xylose for bioreactor cultivation on actual enzymatic hydrolysate (EH) of EFB, in comparison to the shake flask
validation run. (b) The trends of dissolved oxygen (DO) and pH level throughout the cultivation, as well as percentage of sugar consumption for the bioreactor cultivation.
In comparison to the shake-flask cultivation, the growth of M. plumbeus on EH
in a bioreactor showed an improvement in the oil concentration and the oil yield
(Table 5-6). Even at a similar C/N ratio, the cultivation in the bioreactor resulted in
0
5
10
15
20
25
30
35
40
0 24 48 72 96 120 144 168
Con
cen
trat
ion
(g/
L)
Time (h)
GlucoseXyloseGlucose in validation runXylose in validation run
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
0
20
40
60
80
100
120
0 24 48 72 96 120 144 168
pH
Rel
ativ
e C
once
ntr
atio
n (
%)
Time (h)
% SugarconsumedDO
pH
Cascade control applied from beginning
139
much higher oil concentration and oil yield. The enhanced oil production was
possibly attributed to enhanced oxygen supply because of better agitation and
aeration in the culture relative to the cultivation in the shake flask. The transition in
fungal morphology was one of the possible factors of the increased oil production.
The sugar consumption of the cultivation in the bioreactor was similar to the shake-
flask cultivation under a similar C/N ratio, as discussed previously, that the nitrogen
content of media might play role in sugar consumption.
Figure 5-6 Fungal
morphology from the bioreactor cultivation on actual enzymatic hydrolysate (EH). (a, c and e) Microscopic images
(fluorescence) at magnification of 100x. (b and f) Microscopic images (bright field) at magnification of 200x. (d) Bioreactor at 41 h of cultivation.
Table 5-6 showed the comparison between the bioreactor cultivation of M.
plumbeus on EH to other bioreactor cultivation of oleaginous yeasts and fungi on
various lignocellulosic hydrolysates. The oil concentration (5.3 g/L) and oil yield
(168 mg/g) from the bioreactor cultivation on EH were comparable to the results of
bioreactor cultivations from other studies such as oil production from corn stover by
0 h
(a) (b)
41 h
(c) (d)
94 h
(e) (f)
140
M. isabellina (6.9 g/L oil) and from rice straw by Rhodotorula glutinis (6.9 g/L oil)
[23, 26]. The results of oil concentration produced from corn stover by Rhodotorula
graminis was much higher than the oil concertation obtained in this study, possibly
due to higher sugar concentration in corn stover hydrolysate [27]. However, the oil
yield in this study was higher than that obtained from the cultivation on corn stover
hydrolysate (89 mg/g). The study of cultivation of Rhodotorula glutinis on rice straw
hydrolysates in an airlift bioreactor by Yen et al. obtained higher oil concentration
than the present study at 6.9 g/L, possibly due to the higher C/N ratio as the oil
concentration was similar to the results of the cultivation M. isabellina at C/N ratio
of 91 [23]. Even though the oil concentration was higher in Yen et al., the oil yield
was comparable to this study.
The outcome of the bioreactor cultivation study demonstrated the potential of
scaling-up of M. plumbeus for large-scale microbial oil production from EFB
hydrolysates without the addition of a nitrogen source such as yeast extract. The
improved yield of microbial oil converted from palm oil processing wastes (i.e.,
EFB) through bioreactor cultivation, without an additional nitrogen source could
potentially improve the economics of large-scale microbial oil production. Therefore,
the microbial oil production process from EFB was promising for commercialisation
of biodiesel production.
Table 5-6 Results of different cultivation performed in this study and the comparison with other batch cultivation of oleaginous yeasts and fungi from
the literature. Feedstock (hydro-lysate)
Reactor type
Strains Glucose (g/L)
Xylose (g/L)
C/N ratio
Oil (g/L)
Oil yield (mg/g)
Refe-rence
EFB residue
Shake-flask
Mucor plumbeus
31.2 3.7 62 2.7 94 This study
EFB residue
Stirred-tank bioreactor
Mucor plumbeus
33.6 2.8 65 5.3 168 This study
Corn stover residue
Stirred-tank bioreactor
Mortierella isabellina
28.6 16.1 91 6.9 147 [23]
Corncob stover
Stirred-tank bioreactor
Rhodotorula graminis
126.0 87.1 n/n 16.3 89 a [27]
Rice straw Airlift bioreactor
Rhodotorula glutinis
23.9 6.1 n/a 6.9 170 [26]
n/a Not available a Data was not provided
141
5.1.3.6 Biodiesel production from EFB
The fatty acids composition of oil from the bioreactor cultivation on EFB
hydrolysate were 15.6% palmitic, 15.9% stearic, 21.1% oleic and 47.4% linoleic
acid, which was different to the oil produced from the shake-flask cultivation. The
higher composition of unsaturated fatty acids in the bioreactor cultivation could be
due to higher oxygen availability in the bioreactor that led to the formation of
unsaturated fatty acids through the aerobic desaturase/elongase pathway [28].
The fatty acids compositions were further used to evaluate the fuel properties
of microbial oil via empirical calculation. The fuel properties were analysed based on
the assessment of fatty acid methyl ester (FAME) as described in Ahmad et al. [29].
The results showed that the oil has a cetane number of 57.11, iodine value of 99.95
and kinematic viscosity of 4.34 mm2/s which is within the limit set by the European
standard for FAME of EN 14214 (>51 for cetane number, <120 for iodine value and
3.5 to 5.0 mm2/s for kinematic viscosity). Therefore, microbial oil from EFB was a
promising source for good quality biodiesel production.
The production cost of biodiesel production from EFB was calculated based on
Koutinas et al. for biodiesel production from pure glucose through microbial
cultivation, oil extraction and transesterification (Table 5-7) [30]. For the production
of 10,000 t biodiesel from EFB, based on the oil yield of 0.168 g/g sugars and the
conversion rate of oil to biodiesel of 90%, it was estimated that 66138 t of sugars
from EFB was required. The cost of sugars production from EFB was estimated to be
at $256/t based on the selling price of diluted sugars from lignocellulosic biomass
[14, 31].
142
Table 5-7 The comparison of raw materials cost (RMC) (a), production cost (b) and cost of biodiesel (c) for the production of biodiesel from EFB and glucose based on Koutinas et al. for the production of 10,000 t/year biodiesel [30].
(a) Process Raw material Quantity
(t/y)
Unit cost
($/t)
Total cost
($ million/y)
Cultivation RMC on
pure glucose
Glucose 42081 400 16.833
Yeast extract 4370 800 3.495
Total 20.328
Cultivation RMC on
EFB
EFB sugar 66138 256 16.931
Yeast extract 0 0 0
Total 16.931
Extraction & transesterification RMC (hexane, methanol, NaOH and HCl) 0.905
RMC for glucose a 21.233
RMC for EFB a 17.836
(b) Cost item Total cost
($ million/y)
Fixed capital investment (FCI) 73.65
Operating labour cost (OLC) 1.005
Utility cost (UC) 7.563
Production cost for glucose bc 58.895
Production cost for EFB bc 53.604
143
(c) Feedstock Cost of biodiesel
($/t)
Glucose 5900
EFB 5360
Theoretical EFB (at theoretical oil yield) d 4483
a RMC = Cultivation RMC + Extraction & transesterification RMC
1. they meet the criteria for authorship in that they have participated in the
conception, execution, or interpretation, of at least that part of the publication in their
field of expertise;
2. they take public responsibility for their part of the publication, except for the
responsible author who accepts overall responsibility for the publication;
3. there are no other authors of the publication according to these criteria;
4. potential conflicts of interest have been disclosed to (a) granting bodies, (b) the
editor or publisher of journals or other publications, and (c) the head of the
responsible academic unit, and
5. they agree to the use of the publication in the student’s thesis and its publication
on the Australasian Research Online database consistent with any limitations set by
publisher requirements.
In the case of this article:
Optimising extraction of microalgal oil using Accelerated Solvent Extraction by
response surface methodology. Accepted with modification for publication in Journal
of Engineering Science and Technology.
Contributor Statement of contribution
Farah B. Ahmad The author contributed to initial
experimental design; conducted
experiment, analysis and data
interpretation; and wrote the first draft of
manuscript and subsequent revisions of
the manuscripts.
Signature
Date
08/06/2016
Zhanying Zhang This author provided valuable assistance
in data interpretation and reviewed the
VI
manuscript.
William O. S. Doherty This author contributed to data
interpretation and provided valuable
input in reviewing the manuscript.
Ian M. O’Hara This author supervised overall
experimental design, analysis, data
interpretation, and edited the manuscript
draft.
Principal Supervisor Confirmation
I have sighted email or other correspondence from all Co-authors confirming their
certifying authorship.
Name
Ian O’Hara
Signature
Date
Nomenclatures v/v Volume per volume w/w Weight per weight (g/g) X1 ASE static cycles X2 ASE static time (min) X3 ASE operating temperature ( ) Y Response variable of quadratic model xi Input variable of quadratic model xj Input variable of quadratic model k Number of factors of quadratic model Greek Symbols β0 Intercept of quadratic model βi Linear coefficient of quadratic model βij Quadratic coefficient of quadratic model
VII
βii Linear-by-linear interaction between xi and xj regression coefficients of quadratic model
ε Random error of quadratic model Abbreviations ASE Accelerated Solvent Extraction CCD Central composite design DE Diatomaceous earth FAME Fatty acid methyl ester PLE Pressurised liquid extraction RSM Response Surface Methodology
1 Introduction
Lipid or microbial oil shows great promise for second generation biodiesel
production. Microbial oil is an alternative to the conventional feedstock used for
biodiesel production, which are plant oils such as soybean oil (edible oil) or Jatropha
oil (inedible oil). It is more advantageous to use microbial oil than plant oils for
biodiesel production due to several factors such as less labour intensive to cultivate,
has short life cycle, is easy to scale up and has no seasonal and climate requirement
[1]. Oleaginous microorganisms are microorganisms (eg., microalgae, yeasts and
fungi) that are able to accumulate more than 20% lipids within the cells [2].
Oleaginous microalgae, yeasts and fungi have been reported to be able to produce oil
from the cultivation on various carbon sources, including industrial and agricultural
wastes [3].
The extraction process is a crucial step in harvesting microalgae from the
culture for oil production in order to ensure maximum yield of desired product from
microalgae biomass. Solvent extraction is the conventional technique for the
extraction of oil or lipid from biomass. The three most commonly used solvent
extraction techniques for extracting oil from microbial biomass are the Bligh-Dyer,
Folch and Soxhlet extraction techniques [4-6]. However, these extraction methods
are usually multi-step procedures and use large amount of solvents for extended
periods of time [7, 8]. For instance, Soxhlet extraction requires longer extraction time
(8 h) [4], due to slow diffusion and desorption of desired extracts from the sample
matrix to the extraction solvents [9]. Folch technique involves two-steps extraction
method, which is extraction followed by purification using water [6].
Pressurised liquid extraction (PLE), or known as Accelerated Solvent
Extraction (ASE) is an alternative solvent extraction technique which involves
VIII
extraction at elevated temperatures under high pressures. ASE is an automated
technique consists of stainless-steel cells that hold the samples for liquid extraction at
the set up temperature [8]. Temperature, pressure and solvent delivery were
electronically controlled by heaters and pumps [8]. In ASE, the high pressure can be
used in order to keep the solvents in liquid state at temperatures above their boiling
points [10]. The application of higher pressures improves extraction efficiency as the
pressure helps diffuse the solvents into desired extracts trapped within the matrix
pores of the biomass [10]. Therefore, the mechanical pretreatment may not be
necessary prior to the extraction by ASE, unlike the conventional extraction
techniques.
In addition, ASE is more advantageous than conventional extraction
techniques as it allows higher number of samples loading and uses less amount of
solvent with shorter extraction time [5-7]. It has been reported that extraction of
persistent organic pollutants (POPs) from soils and sediments using PLE required
only 20 min of extraction time and 10 times less solvent than Soxhlet extraction [8].
ASE also has health and safety benefits, as it reduces the potential for contact with
chemical solvents. Studies showed that the extraction using ASE resulted either in
increased or comparable amounts of oil in comparison to conventional extraction
techniques [11, 12]. A study on oil extraction from algae biomass showed that a
higher amount of oil was obtained using ASE technique than the Folch method,
where the solvents used for both methods were chloroform/methanol (2:1, v/v) [11].
Higher total fatty acids yields were achieved from the extraction of cereal, egg yolk
and chicken breast muscle samples using ASE than a modified Folch method with
the use of isopropanol/hexane (2:3, v/v) for both methods [12].
Previous studies have shown that solvent types and temperature affected oil
yields in ASE [11-13]. The extraction of oil from the algae biomass showed higher
fatty acid yields were obtained with the use of chloroform/methanol (2:1, v/v)
compared to the use of isopropanol/hexane (2:3, v/v) and hexane [11]. The
combination of non-polar and polar solvent (such as chloroform/methanol) was
shown to be more effective for extracting neutral lipid (i. e., microbial oil) from
microbial biomass, in comparison to the use of non-polar solvent (such as hexane)
alone [14]. Another extraction study on dry microalgae biomass using ASE reported
on the effect of temperature to the oil yield where the study demonstrated that
IX
slightly higher amounts of total fatty acid methyl esters (FAMEs) were obtained at
120 °C compared to 110 °C, and significantly higher FAMEs at 120 °C compared to
temperatures below 100 °C [15]. Despite these reports, there is no systematic study
on optimisation of ASE from microbial biomass that correlates the important
extraction parameters such as temperature, the number of process cycles and the
process time to the oil yields.
The aim of this study was to optimise the extraction of microbial oil from
microalgae Chlorella protothecoides using ASE technique by response surface
methodology (RSM). The parameters for determining optimum oil yield were the
number of static cycles, static time (min) and temperature (°C), with oil yield (%,
w/w) as the response parameter. The optimisation study was conducted through the
experimental design, experimental run using microalgal biomass and experimental
data analysis for the development of a mathematical model.
2 Experimental Procedures
2.1 Microbial biomass preparation
Chlorella protothecoides ATCC® 30581 (ATCC, USA) was used in this study.
Microalgae was maintained in a growth chamber with light intensity from 38 - 47
μmol/m2/s at 25 °C under a 14 hour light/10 hour dark cycle. Microalgae were
subcultured in modified Medium 847 as described in the Product Information Sheet
for ATCC® 30581™. The cultivation conditions were similar to inoculum
preparation and microalgae cultivation described previously [16], with 10% (v/v)
inoculum was used. Microbial biomass was harvested by centrifugation at 6805 X g
for 7 min followed by freeze-drying [16].
2.2 Extraction of oil from microalgal biomass using ASE
Dionex ASE 350 (Thermo Fisher Scientific Inc., USA) was used for extraction
of oil from microalgal biomass. The biomass samples were prepared by mixing dry
microalgal biomass (0.25 g) with 4 g of ASE Prep DE (diatomaceous earth) (Thermo
Fisher Scientific Inc., USA) before being loaded into 33 mL cells [11]. The detailed
extraction process in Dionex ASE 350 is illustrated in Figure 1. A static extraction in
the cell commences after solvent filling followed by cell heating, up until before the
X
cell is rinsed with fresh solvents. Static time is the period where static extraction
occurs, and static cycle is the number of times where static extraction occurs.
The extraction conditions were as follows for the control run and RSM
experimental run: rinse volume, 50% of cell volume; purge time, 60 s; with varying
static cycles, temperature and static time using chloroform/methanol (2:1, v/v) as the
extraction solvents. The control run was performed based on the optimised condition
reported in previous study (ASE with 4 static cycles, static time of 120 and
temperature of 5 min) [11]. The extracted oil was collected in pre-weighed collection
bottles. Solvent was later evaporated under a stream of nitrogen. Oil yield (%, w/w)
was calculated as follows,
Oilyield %, /
x100% (1)
2.3 Design of experiment by response surface methodology (RSM)
A response surface methodology (RSM) with face-centred central composite
design (CCD) was applied for designing the experiments of optimising oil yield from
the extraction of microalgae biomass by ASE. The parameters (independent
variables) selected for optimising the oil yield by ASE are static cycles, static time
and temperature. Design of experiments, mathematical modelling and optimisation of
process parameters were performed using the Design Expert 7 Trial version software
package (Stat-Ease Inc., USA). The independent parameters used in this study were
static cycles (X1), static time (min) (X2) and temperature (°C) (X3). The response
factor (dependant variable) for optimisation was oil yield (%) (Y). The coded levels
for parameters, -1 and 1, indicate the limits of each factor, where the actual values of
each factor and its levels for this experimental design are shown in Table 1. The
range of the factor was based on the preliminary study performed previously [17]. A
total of 12 experimental runs were conducted in random with 4 factorial points, 6
axial points and a centre point (in duplicate for experimental error calculation).
XI
Fig. 1 Flow diagram of Dionex ASE 350 extraction process. X1 is the number of
static cycles, X2 is static time (min) and X3 is operating temperature (°C). Cell is the
stainless steel sample holder where the extraction process occurs.
Table 1 Coded and actual values of the parameters in the experimental design
Factor Notation Units Coded levels of parameters -1 0 1
Static cycles X1 1 4 6
X1
Sample preparation: Mix biomass with DE
Cell is loaded into the oven
Cell is filled with solvent and heated for a fixed time to ensures thermal equilibrium of samples
Cell heating and static extraction (at set up X2 and X3)
Cell is rinsed with fresh solvent
Remaining solvent is purged with N2 gas
Residual pressure is released from the cell
Cell is unloaded from the oven
Solvent + Extract (Oil)
XII
Static time X2 min 2 6 10 Temperature X3 °C 100 130 160
2.4 Statistical analysis and modelling
From the experiments that have been performed based on the design by RSM,
the oil yield (response variable, Y) was fitted by a quadratic model to correlate the
response variable to the independent variables. The experimental data obtained were
calculated and analysed through an empirical second-order polynomial function:
∑ ∑ ∑ ∑ ε (2)
where Y is the predicted response; β0 the intercept, βi the linear coefficient, βij the
quadratic coefficient, βii is the linear-by-linear interaction between xi and xj
regression coefficients, xi, xj are input variables that influence the response variable
Y, k is the number of factors and ε is the random error [18]. Analysis of variance
(ANOVA) was evaluated through statistical analysis of the model. The statistical
significance of the model terms was assessed using P-value approach.
3 RESULTS AND DISCUSSION
3.1. Mathematical modelling of the experimental data
The second-order model was employed for approximating the relationship
between the oil yield and the independent variables, as shown below
B Value Standard deviation 1.03 Mean 23.15 Coefficient of variance (%) 4.44 PRESS 8.14 R2 0.9970 Adjusted R2 0.9836 Predicted R2 0.9885
The mathematical model was further evaluated by plotting the predicted oil
yield against the actual oil yield as shown in Figure 2(A). The plot demonstrated a
good agreement of the predicted oil yield to the actual oil yield. The model was also
evaluated through the plot of residuals versus fitted values (Figure 2(B)). The plot
showed that the residuals are structureless and do not display any obvious pattern
[18]. The undesirable pattern in the plot can appear in the form of megaphone or
outward-opening funnel due to the increase of variance, as the magnitude of the
predicted values increases [18]. Therefore, this analysis demonstrated that the
mathematical model is correct and the assumptions are satisfied [18]. The model is
reliable for predicting the extraction of oil microalgae biomass using ASE technique.
XV
Fig. 2. (A) Predicted versus actual plot of oil yield. (B) Internally studentised
residuals versus predicted plot of oil yield.
3.2 Evaluating the interaction among factors that influences extraction of oil
from microbial biomass
From the ANOVA results, it shows that the interaction between static cycles
and temperature is significant (P-value < 0.0500). While the interactions between
static time and static cycles, and static time and temperature are not significant. Low
significant effect of static time to the extraction in this study was in agreement with
the study of extraction of oil from a wet microalgae that showed no considerable
different on total FAMEs at 5 min, 10 min and 15 min [15]. Similar profiles were
also observed on fatty acids extraction from cereal lipids study as increasing static
time from 5 min to 10 min and 15 min had no effect on lipids’ fatty acids yield [12].
A significant impact of temperature on the extraction is possibly because higher
temperatures can improves solubility and mass transfer of liquid solvents into
samples matrix [21]. Higher temperature can also enhance the disruption of the
strong solute-matrix interactions, due to Van Der Waals forces, hydrogen bonding
and dipole attractions between the solutes molecules and actives sites on the matrix
[10]. Static cycle showed significant effect to the oil yield possibly due to the
addition of fresh solvent (i.e., methanol and chloroform) for each static extraction.
The addition of fresh solvent may increase the concentration gradient between the
solution inside the cell and the surface of sample matrix. The increase of
A
B
XVI
concentration gradient promotes faster mass transfer rate according to Fick’s first law
of diffusion [10].
A significant impact of temperature on the extraction is possibly because
higher temperatures can improves solubility and mass transfer of liquid solvents into
samples matrix [10]. Higher temperature can also enhance the disruption of the
strong solute-matrix interactions, due to Van Der Waals forces, hydrogen bonding
and dipole attractions between the solutes molecules and actives sites on the matrix
[10]. In addition, increasing temperature reduces viscosity of liquid solvents which
facilitates diffusion into matrix particles [10].
Static cycle showed significant effect to the oil yield possibly due to the
addition of fresh solvent for each static extraction that may increase the
concentration gradient between the solution inside the cell and the surface of sample
matrix [10]. This is because higher concentration gradient promotes faster mass
transfer rate according to Fick’s first law of diffusion [10].
3.3 Optimising the parameters of the extraction of oil from microbial biomass
The response surface analysis through 3-dimensional response surface plot
(Figure 3) was performed for optimising the yield of oil from microbial biomass
using ASE. Fig. 3(A) shows that at 130 and 6 min static time, the amount of oil
extracted increased with increasing static cycles, up until approximately 4-5 static
cycles. This result showed that oil could be extracted from even after 4 static cycles,
which is in accordance to other study on oil extraction from macroalgae biomass
(Rhizoclonium hieroglyphicum) using ASE [11]. From Fig. 3(B), oil yield
significantly increased with increasing temperature. The impact of static cycle on the
oil yield was more apparent at lower temperature than higher temperature. Fig. 3(C)
shows that at 4 static cycles for all extraction temperatures, the oil yields increased
with increasing static time up until an optimum time at approximately 6 min. The
extraction at 4 static cycles in Fig. 3(C) demonstrated that at a very high extraction
temperature such as at 160 °C, the oil yield was gradually decreasing when the static
time was more than 6 min. This is because that there is the possibility for the
samples to degrade during prolonged extraction process at a very high temperature.
High extraction temperature has been suspected to cause thermal degradation of
XVII
lipids [13]. Even though there was a small decrease in oil yield at lower temperature
(100 °C) for the extraction at more than 6 min, the drop in oil yield was too low.
Fig. 3 Dimensional surface plots of binary interaction between different variables to
the oil yield: (A) static cycles and static time at 130 °C, (B) static cycles and
temperature at 6 min and (C) static time and temperature at 4 static cycles. The
figures were generated from Design-Expert software.
A
B
C
XVIII
The parameters of optimisation of extraction of oil by ASE were determined
based the numerical optimisation according to the criteria showed in Table 4.
Numerical optimisation was performed based on an objective function called
desirability [20]. At desirability (0 to 1) of 1 from the numerical optimisation (Figure
4), the parameters of optimum oil yield selected were 4 static cycles, 6 min and 130
for the extraction using ASE, with the maximum oil yield of 34.9% (w/w).
Table 4 Criteria for numerical optimisation of maximum oil yield
Criteria Goal Lower limit Upper limitStatic cycles In range 1 6 Static time (min) In range 2 10 Temperature (°C) In range 100 160 Oil yield Maximise 1.59 34.89
In this study, the control run on microalgal biomass resulted in an oil yield of
26.0% (w/w), which is lower than the maximum oil yield obtained in this study.
Therefore, the extraction using optimised conditions in this study showed 1.34 fold
increases in oil yield from the control run. The optimised ASE conditions in this
study demonstrated an improvement in extraction technique for quantifying the
amount of oil in the biomass.
Fig. 4 3-Dimensional surface plot of the binary interaction at 160 °C between static
time and static cycle to the desirability value. The figure was generated from Design-
Expert software.
XIX
Conclusions
The RSM was utilised for optimising the oil yield from the extraction on
microalgal biomass using ASE technique. The mathematical model developed from
the response surface analysis was reliable to predict the oil yield. The results showed
a good agreement of the predicted oil yield to the actual oil yield from the
experimental run. Based on the surface response analysis, the optimised ASE
conditions determined were 4 static cycles, static time of 6 min and temperature of
160 , with the maximum oil yield of 34.9% (w/w). The optimised oil yield also
resulted in significant improvement of oil yield in comparison to the oil yield from
the control run, with 1.34 fold increases in oil yield.
Acknowledgements
The authors acknowledge Ministry of Education Malaysia for the postgraduate
scholarship of the first author. The authors also acknowledge the QUT Central
Analytical Research Facility for its support for sample analyses.
References
[1] Liang, M.-H. and Jiang, J.-G., Advancing oleaginous microorganisms to produce
lipid via metabolic engineering technology. Progress in Lipid Research, 2013; 52:
395-408.
[2] Meng, X., et al., Biodiesel production from oleaginous microorganisms.
Renewable Energy, 2009; 34: 1-5.
[3] Huang, C., et al., Single cell oil production from low-cost substrates: The
possibility and potential of its industrialization. Biotechnology Advances, 2013; 31:
129-139.
[4] Araujo, G.S., et al., Extraction of lipids from microalgae by ultrasound
application: Prospection of the optimal extraction method. Ultrasonics
Sonochemistry, 2013; 20: 95-98.
[5] Bligh, E.G. and Dyer, W.J., A rapid method of total lipid extraction and
purification. Canadian Journal of Biochemistry and Physiology, 1959; 37: 911-917.
[6] Folch, J., Lees, M., and Stanley, G.H.S., A simple method for the isolation and
purification of total lipides from animal tissues. Journal of Biological Chemistry,
1957; 226: 497-509.
XX
[7] Poerschmann, J. and Carlson, R., New fractionation scheme for lipid classes
based on “in-cell fractionation” using sequential pressurized liquid extraction.
Journal of Chromatography A, 2006; 1127: 18-25.
[8] Ramos, L., Kristenson, E.M., and Brinkman, U.A.T., Current use of pressurised
liquid extraction and subcritical water extraction in environmental analysis. J. of
Chromatogr. A, 2002; 975: 3-29.
[9] Björklund, E., Nilsson, T., and Bøwadt, S., Pressurised liquid extraction of
persistent organic pollutants in environmental analysis. Tr. in Anal. Chem., 2000;
19: 434-445.
[10] Richter, B.E., et al., Accelerated solvent extraction: a technique for sample
cultivation was performed in triplicate with 30 mL working volume in 150 mL
Erlenmeyer flask at 28 °C on an OM15 orbital shaking incubator (Ratek, Australia)
for 7 days. The biomass was harvested and freeze-dried to constant weight [9].
2.4 Oil extraction
Oil was extracted from the biomass by Accelerated Solvent Extraction (ASE)
technique using Dionex ASE 350 (Thermo Fisher Scientific Inc., USA) [9]. The
extraction conditions were as follows: temperature, 130 ; static time, 7 min; rinse
volume, 25% of cell volume; and using two static cycles, using chloroform/methanol
in a ratio of 2:1 (v/v). The results are reported on a dry weight (DW) basis unless
otherwise specified.
2.5 Sugars, organic acid, furans and oil analyses
Sugar, organic acid and furan concentrations were measured using high-
performance liquid chromatography (HPLC) as described in previous studies [9, 15].
For the determination of fatty acids composition, fatty acid methyl esters (FAME)
were prepared and analysed using gas chromatography-mass spectrometry (GC-MS)
based on Ahmad et al. (2015). The following GC-MS method was used: injection
temperature at 250 °C, initial temperature at 90 °C, hold for 2 min, followed by 7.5
°C/min ramp to 210 °C and 20 °C/min ramp to 240 °C, hold for 5 min.
Results and discussion
3.1 Composition of sugarcane bagasse hydrolysates
In this study, hydrolysates from both streams of the dilute acid pretreatment of
bagasse (Figure 1) were used as the cultivation media for oil production. The liquid
fraction of dilute acid pretreatment typically contains readily fermentable soluble
sugars. The solid residue of pretreated biomass primarily consists of cellulose and
therefore, requires enzymatic hydrolysis for breaking down cellulose and other
complex sugars to fermentable sugars. The fermentable sugars (i.e. glucose and
xylose) from hydrolysates of pretreated bagasse (Table 1) were utilised as carbon
substrates for microbial oil production.
The detoxification step was necessary for bagasse hydrolysates in this study as
it contains considerable concentrations of potential growth inhibitors (e.g. furfural
and HMF). Furfural had been reported to inhibit the growth of oleaginous yeast
XXIX
Cryptococcus curvatus at the concentration of 1.0 g/L [16]. HMF however, has been
shown to have a less inhibiting impact on growth [5, 16, 17]. No growth was
observed in this study from the cultivation on non-detoxified bagasse hydrolysates.
In this study, the detoxification technique of overliming was applied and the
detoxification had reduced the concentrations of potential inhibitors. It is often
recommended to include a washing step on pretreated solid residue prior to
enzymatic hydrolysis, in order to avoid the presence of potential inhibitors in the
enzymatic hydrolysate. In addition, the washing step may also reduce detoxification
requirements.
Table 1 Selected sugars (glucose and xylose), acetic acid, HMF and furfural compositions of sugarcane bagasse hydrolysates.
Pretreatment stream
Feedstock Glucose (g/L)
Xylose (g/L)
Acetic acid (g/L)
HMF (g/L)
Furfural (g/L)
Liquid fraction
Non-detoxified SCBLH
7.76 8.32 5.85 0.91 3.90
SCBLH 6.11 6.00 6.60 0.04 0
Solid residue Non-detoxified SCBEH
29.96 3.46 2.36 0.38 3.39
SCBEH 31.34 1.21 2.53 0 0.17
3.2 Biomass concentration and sugars consumption from the cultivation on
bagasse hydrolysates
The oleaginous microorganisms used in this study for oil production from
bagasse hydrolysates were Rhodotorula mucilaginosa, Aspergillus oryzae and Mucor
plumbeus. These microorganisms were the highest ranking microorganisms from a
multi-criteria analysis by Ahmad et al. (2015) that had been conducted to select
preferred oleaginous microorganisms for oil production [9].
All microorganisms used in this study were able to grow on bagasse
hydrolysates. As shown in Figure 2, fungi M. plumbeus and A. oryzae had the highest
biomass concentrations on SCBLH at 9.8 and 9.2 g/L respectively, followed by R.
mucilaginosa. The results of the biomass concentrations on SCBLH compare well to
XXX
the growth of Yarrowia lipolytica Po1g on the detoxified liquid hydrolysate of
bagasse (13.51 g/L of xylose and 3.93 g/L of glucose) with biomass concentration of
11.42 g/L [6].
For the cultivation on SCBEH, M. plumbeus had the highest biomass
concentration at 19.9 g/L followed closely by A. oryzae Figure 2. The higher biomass
concentrations from the growth on SCBEH than SCBLH correspond to higher sugars
concentration of the SCBEH. The biomass concentrations of the fungi obtained from
SCBEH are comparable to the cultivation of fungus Mortierella isabellina with
biomass concentrations at 16.8 g/L on enzymatic hydrolysate of corn stover (22.2
g/L of glucose and 12 g/L of xylose) [18].
Figure 2 Microbial biomass concentrations (g/L) from bagasse hydrolysates.
The cultivation on SCBLH showed poorer consumptions of glucose by
microorganisms in comparison to the cultivation on SCBEH (Figure 3), possibly due
to the presence of HMF in the cultivation media. HMF may have caused a slower
growth rate and subsequently low consumption of carbon substrates. This is because
microorganisms that metabolise HMF may have a longer lag phase for growth [5]. In
the cultivation on SCBEH, the consumption of xylose in is much lower than the
consumption of glucose (Figure 3). Lower consumptions on xylose by
microorganisms were possibly because xylose consumption began only after there
was almost no glucose left in the medium, as reported in numerous studies [19].
0
5
10
15
20
25
SCBLH SCBEH
Bio
mas
s co
ncen
trat
ion
(g/L
) R. mucilaginosa
A. oryzae
M.plumbeus
XXXI
Figure 3 Sugars consumption (glucose and xylose) by microorganisms on bagasse hydrolysates.
3.3 Microbial oils production from bagasse hydrolysates
The oil content results (Figure 4(a)) showed that M. plumbeus had the highest
oil contents on both SCBLH (14.9%) and SCBEH (23.8%). The oil content of A.
oryzae is 9.0% on SCBLH and 19.6% on SCBEH. For R. mucilaginosa, the oil
contents on both hydrolysates are very similar (~11%).
The oil concentrations (Figure 4(b)) showed that M. plumbeus had the highest
oil concentrations on SCBLH and SCBEH at 1.6 g/L and 4.7 g/L respectively,
followed by A. oryzae. The oil concentration of M. plumbeus on SCBEH compares
well to oil concentration of 6.68 g/L by Y. lipolytica Po1g from bagasse hydrolysate.
Overall, this study showed that the use of M. plumbeus resulted in the best biomass
and oil production growing on both SCBLH and SCBEH in comparison to A. oryzae
and R. mucilaginosa.
0
10
20
30
40
50
60
70
80
90
100
Glucose Xylose Glucose Xylose
% s
ugar
s co
nsum
ptio
n (w
/w)
R. mucilaginosa
A. oryzae
M. plumbeus
SCBLH SCBEH
XXXII
(a)
(b)
Figure 4 (a) Oil content (%, w/w) and (b) oil concentration (g/L) of microbial biomass from the cultivation on bagasse hydrolysates.
3.4 Evaluation of potential biodiesel application of microbial oils
Microbial oils with fatty acid compositions similar to vegetable oils have the
potential to be used as feedstock for biodiesel production. In this study, the majority
of fatty acids identified were palmitic (C16:0), stearic (C18:0), oleic (C18:1) and
linoleic acid (C18:2) (Table 2), which is similar to the fatty acid composition of
vegetable oils. For R. mucilaginosa, oleic acid was the predominant fatty acid
growing on both hydrolysates, which is analogous to the results of the cultivation of
R. mucilaginosa on glucose and xylose [9]. Linoleic acid was found to be the
predominant fatty acid for both fungi strains on SCBLH, and oleic acid was the
predominant fatty acid on SCBEH.
Fuel properties of microbial oils were evaluated using cetane number and
iodine value (Table 2). Cetane number is an indicator to the ignition quality of fuels
0
5
10
15
20
25
30
SCBLH SCBEH
Oil
con
tent
(%
, w/w
)
R. mucilaginosa
A. oryzae
M.plumbeus
0
1
2
3
4
5
6
SCBLH SCBEH
Oil
con
cnet
ratio
n (g
/L)
R. mucilaginosa
A. oryzae
M.plumbeus
XXXIII
[20]. Iodine value measures the degree of saturation of fuels. The fatty acid methyl
ester (FAME) of microbial oils in this study were within the limit of the European
biodiesel specification since the European Standard for FAME (EN14214) specifies
cetane number of biodiesel has to be above 51 and iodine number below 120 [21].
Therefore, this study shows that the microbial oils produced are suitable for second
generation biodiesel production from bagasse.
Table 2 Fatty acids composition of microbial oils methyl esters from the cultivation on bagasse hydrolysates, as well as cetane number [20] and iodine number [22] of
the transesterified microbial oils. Feedstock Microorganisms Relative abundance of fatty acid
3.5 Prospects for large-scale microbial oil production from bagasse hydrolysates
by M. plumbeus
M. plumbeus has great potential for oil production from bagasse as it had the
highest biomass and oil concentration on both bagasse hydrolysates. Large-scale
microbial oil production from bagasse can be assessed through the yields of oil. The
oil yields (mg oil per g carbon substrates consumed including glucose, xylose and
acetic acid) of M. plumbeus were 167 mg/g on SBCLH and 142 mg/g on SCBEH.
The oil production was calculated by multiplying the oil yields with the sugar yields
from 1 t (DW) bagasse. A study on optimised dilute acid pretreatment and enzymatic
hydrolysis of bagasse reported that the sugars yield from 1 t bagasse (DW) contained
29 kg of glucose and 198 kg of xylose from liquid hydrolysate, and 360 kg of
glucose and 15 kg of xylose from enzymatic hydrolysate [23]. By combining the oil
production from both hydrolysates, it is estimated that up to 91 kg of microbial oil
XXXIV
can be produced from 1 t dry bagasse by M. plumbeus. The utilisation of both
hydrolysates maximises the total oil production from bagasse. The total oil
production from bagasse by M. plumbeus is comparable to the oil production by M.
isabellina with 103 kg oil from 1 t (DW) of wheat straw [14]. Table 3(a) shows the
comparison of the feedstock cost for oil production from bagasse and pure glucose,
where bagasse provides nine times lower feedstock cost compared to the use of pure
glucose. The oil yield and the price of pure glucose were based on previous studies
[9, 24].
Table 3 (a) The comparison of estimated feedstock cost for oil production from bagasse and pure glucose. (b) The estimated yield of biodiesel per 1 t (dry weight)
bagasse based on the oil yields of M. plumbeus (a) Microbial oil
feedstock Price (US$/t feedstock)
Oil yield (kg/t feedstock)
Feedstock cost (US$/kg oil)
Sugarcane bagasse 50 91 0.55 Pure glucose 400 80 5.00 (b) Yield (L/t dry bagasse) Biodiesel from SCBLH
(FAME density of 874.2 g/L) 37
Biodiesel from SCBEH (FAME density of 860.5 g/L)
56
Biodiesel from bagasse 93
Biodiesel yields in Table 3 were calculated based on 91% (w/w) microbial oil
conversion [14, 25] and estimated FAME densities [20]. From the 10 million wet t of
bagasse (with 50% moisture content) generated annually from sugar mills in
Australia, a maximum of around 465 million L biodiesel could potentially be
produced if all of the bagasse was used for microbial oil production. Based on diesel
consumption of 5200 million L in Queensland [26], this would be equivalent to about
9% of Queensland’s diesel consumption. A greater proportion is potentially able to
be produced through the use of trash and future production of high fibre sugarcanes.
While wide scale application of microbial oils from bagasse for biodiesel production
would only replace a small fraction of total Queensland diesel consumption, local
and regional opportunities may exist in some areas such as for replacement of diesel
use in sugarcane harvesting and transport.
XXXV
Conclusion
Rhodotorula mucilaginosa, Aspergillus oryzae and Mucor plumbeus showed
the capacity to produce oil from hydrolysates of bagasse. The fungus M. plumbeus
showed the highest oil concentrations on both hydrolysates, therefore, the oil yields
of M. plumbeus were further used to estimate large-scale microbial oil production
from bagasse. From the estimated total oil production, this study shows the potential
to reduce the feedstock cost for microbial oil production through the use of bagasse.
Therefore, bagasse shows a great promise as the feedstock for biodiesel production
that may supplement diesel use for local consumption. The utilisation of both
bagasse hydrolysates (SCBLH and SCBEH) may create sustainable and profitable
microbial oil production.
Acknowledgements
The authors acknowledge Ministry of Education Malaysia for the postgraduate
scholarship of the first author. The authors also acknowledge the QUT Central
Analytical Research Facility for its support on sample analyses, as well as Vitor
Takashi Kawazoe for oil extraction and derivatisation process.
References
[1] Liang, M.-H. and Jiang, J.-G., Advancing oleaginous microorganisms to produce
lipid via metabolic engineering technology. Progress in Lipid Research, 2013; 52:
395-408.
[2] Ratledge, C., Fatty acid biosynthesis in microorganisms being used for Single