BIOCONVERSION OF SUGARCANE BAGASSE AND SOYBEAN HULLS FOR THE PRODUCTION OF A GENERIC MICROBIAL FEEDSTOCK A thesis submitted to The University of Manchester for the degree of Doctor of Philosophy in the Faculty of Engineering and Physical Sciences 2015 CHEN-WEI CHANG Satake Centre for Grain Process Engineering School of Chemical Engineering and Analytical Science The University of Manchester, UK
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
BIOCONVERSION OF SUGARCANE BAGASSE AND
SOYBEAN HULLS FOR THE PRODUCTION OF A
GENERIC MICROBIAL FEEDSTOCK
A thesis submitted to The University of Manchester
for the degree of Doctor of Philosophy
in the Faculty of Engineering and Physical Sciences
2015
CHEN-WEI CHANG
Satake Centre for Grain Process Engineering
School of Chemical Engineering and Analytical Science
The University of Manchester, UK
Table of contents
i
Table of contents
List of tables ................................................................................................... vii
List of figures ................................................................................................... ix
List of abbreviations and acronyms ................................................................ xix
Abstract ......................................................................................................... xx
Declaration ................................................................................................... xxi
Figure 8.1 Material balance for the SSF-based process showing the lignocellulose
(600 g of sugarcane bagasse and 400 g of soybean hull) to ethanol conversion
yield. Calculations were based on the average ethanol yield (0.31 g/g total
reducing sugar consumed) for fermentation using hydrolysate. The ratio of SSF
solids to water was 4% (w/w) in the simultaneous hydrolysis and fungal autolysis
to produce a suitable generic microbial feedstock for ethanol fermentation. .. 199
List of abbreviations and acronym
xix
List of abbreviations and acronyms
μm micrometre(s)
µL microliter(s)
aw water activity
C/N ratio carbon to nitrogen ratio
CER carbon dioxide evolution rate
DNS 3,5-dinitrosalicylic acid
DTG derivative thermogravimetric analysis
FAN free amino nitrogen
FPU filter paper unit
g gram(s)
h hour(s)
IP inorganic phosphorus
kg kilogram(s)
L litre(s)
LAP laboratory analytical procedure
mg milligram(s)
mL millilitre(s)
mm millimetre(s)
mm2 square millimetre(s)
mm3 cubic millimetre(s)
OUR oxygen uptake rate
rpm round per minute
RQ respiratory quotient
SB sugarcane bagasse
SCGPE Satake Centre for Grain Process Engineering
SEM scanning electron microscope
SH soybean hull
SmF submerged fermentation
List of abbreviations and acronym
xx
SSF solid-state fermentation
U enzyme activity unit
USB universal serial bus
USD United states dollar(s)
v/v volume by volume
w/w weight by weight
X magnification factor
YD yeast extract solution supplemented with glucose
YDX yeast extract solution supplemented with glucose and xylose
Abstract The University of Manchester
xx
Bioconversion of sugarcane bagasse and soybean hulls for the production of a generic microbial feedstock
Abstract Lignocellulose, mostly from agricultural and forestry resources, is a potential renewable material for sustainable development of biorefineries. From previous studies, reducing sugar production through biological pretreatment involves two steps: solid-state fermentation (SSF) for delignification, followed by enzymatic hydrolysis by adding celluloytic enzymes (cellulase and xylanase etc.). In the process described in this thesis, the necessary enzymes are produced in-situ and the hydrolysis proceeds directly after the solid-state fermentation. Enzyme hydrolysis releases free amino nitrogen (FAN), reducing sugar and many other potential nutrients from the fermented materials. This method additionally avoids the need for removal of inhibitors compared with conventional chemical pretreatment processes. A range of solid-state fermentations were carried out to investigate the effect of washing procedure, particle size and nitrogen supplement on Trichoderma longibrachiatum growth. From these preliminary studies it was concluded that nitrogen supplementation is a crucial factor to improve significantly the fungi growth and production of feedstock using sugarcane bagasse as raw material. In order to evaluate the influence of environmental humidity on petri dish experiments, moist environments were investigated, with over 75% relative humidity to limit water evaporation from solid-state fermentation. The results showed that moist environments gave approximately 1.85 times the reducing sugar yield than dry environments. The process of simultaneous enzymatic hydrolysis of substrates and fungal autolysis were also studied. The degree of hydrolysis was affected by initial fermented solid to liquid ratio, temperature and pH range. The optimal conditions for subsequent hydrolysis of fermented solids were determined. The optimal solid to liquid ratio, 4% (w/w), temperature 50°C and pH 7 were established. The highest final reducing sugar, 8.9 g/L and FAN, 560 mg/L, were measured after 48 h. The fungal autolysis was identified by image analysis as well as by the consumption of nutrient and the release of free amino nitrogen and phosphorous. Solid state fermentation in a multi-layer tray bioreactor and a packed-bed bioreactor were also developed, with moist air supply for oxygen provision and heat removal. Fermented solids in the multi-layer bioreactor led to the highest subsequent hydrolysis yield on reducing sugar, FAN and Inorganic Phosphorous (IP), 222.85 mg/g, 11.56 mg/g and 19.9 mg/g, respectively. These series of fermentation experiments illustrate the feasibility for the application of consolidated bioprocessing, through simultaneous pretreatment and enzyme production as a more economic and environment-friendly process compared with those reported for chemical pretreatment followed by commercial enzyme process. A growth kinetic model regarding both growth and respiration is also proposed. Ethanol production was studied using the generic feedstock produced from sugarcane bagasse and soybean hulls. Total ethanol yield reached 0.31 mg g-1 (61.4% of theoretical yield) after 30 h of submerged fermentation. The result of subsequent fermentation has already shown the potential of the generic microbial feedstock to be used to produce varied products depending on the microorganism utilised. Chen-Wei Chang PhD Thesis June 2015
Declaration
xxi
Declaration
No portion of the work referred to in the thesis has been submitted in support of
an application for another degree or qualification of this or any other university or
other institute of learning.
Copyright statement
xxii
Copyright statement
i. The author of this thesis (including any appendices and/or schedules to this
thesis) owns certain copyright or related rights in it (the “Copyright”) and s/he
has given The University of Manchester certain rights to use such Copyright,
including for administrative purposes.
ii. Copies of this thesis, either in full or in extracts and whether in hard or electronic
copy, may be made only in accordance with the Copyright, Designs and Patents
Act 1988 (as amended) and regulations issued under it or, where appropriate, in
accordance with licensing agreements which the University has from time to
time. This page must form part of any such copies made.
iii. The ownership of certain Copyright, patents, designs, trade marks and other
intellectual property (the “Intellectual Property”) and any reproductions of
copyright works in the thesis, for example graphs and tables (“Reproductions”),
which may be described in this thesis, may not be owned by the author and may
be owned by third parties. Such Intellectual Property and Reproductions cannot
and must not be made available for use without the prior written permission of
the owner(s) of the relevant Intellectual Property and/or Reproductions.
iv. Further information on the conditions under which disclosure, publication and
commercialisation of this thesis, the Copyright and any Intellectual Property
and/or Reproductions described in it may take place is available in the University
IP Policy(see http://documents.manchester.ac.uk/DocuInfo.aspx?DocID=487), in
any relevant Thesis restriction declarations deposited in the University Library,
The University Library’s regulations (see
http://www.manchester.ac.uk/library/aboutus/ regulations) and in the
Figure 2.12 Isotherms of solids used in solid-state fermentation (- - -) Desorption isotherm of autoclaved wheat grains at 35°C; (-) Isotherm of corn at 20°C (lower curve), 35°C (middle curve) and 50°C (upper curve) Adapted from (Calçada, 1998; Mitchell et al., 2006; Nagel et al., 2001a)
2.5.4.5 Effect of pH
Filamentous fungi have good growth over a wide range of pH (2 - 9) with an
optimal range of 3.8 to 6.0 (Gowthaman et al., 2001). For solid substrate
cultivation, pH control during the fermentation is extremely difficult as free water
is absent and cannot easily monitored by any instruments (Manpreet et al., 2005;
Saucedo-Castaneda et al., 1992). Also, it is not practical to add acid or alkali to the
medium to adjust the pH during the fermentation due to the lack of mixing and
possibility of changing moisture content. To minimise pH variation, buffers can be
used, for instance, urea as a nitrogen source and Mandel's medium as mineral salts
source, relieving the difficulties in controlling and maintaining pH during
fermentation (Saucedo-Castaneda et al., 1992).
2.5.4.6 Effect of aeration
Aeration is a significant parameter in SSF because of oxygen consumption in the
aerobic bioprocess as well as heat and mass transport in a heterogeneous
Chapter 2 Literature review
35
environment (Pandey and Larroche, 2008). Because there is no free water and
steady mixing in SSF process, oxygen transfer is limited to the liquid film of the
solid surface and stagnant. Although all concentrations (oxygen and temperature)
are uniform at the beginning of fermentation, the oxygen and temperature
gradient will develop as the fermentation progresses due to mass transfer
resistances (Gowthaman et al., 1993). In order to perform efficient solid-state
fermentation, these gradients must be minimised. Aeration could provide four
beneficial functions for SSF, namely (1) to maintain aerobic conditions, (2) to
remove CO2 from solid bed, (3) to manage the temperature of bed and (4) to
maintain the moisture level of bed (Krishna, 2005; Raimbault, 1998). It must be
emphasized, however, unsaturated air supply to solid state fermentation could
lead to intensive evaporation happened, causing water loss of fermenting solids
and inhibit fungal growth.
2.5.5 Reactor
A solid-state fermentation bioreactor is designed to provide the desired
environment for the fermentation, where biological reactions carry out converting
the raw materials into products. According to Mitchell et al., (2000) there are four
main types of reactor on the basis of how they are mixed and aerated (Figure 2.13).
This classification covers the most common bioreactor types used in solid state
fermentation. These are the tray bioreactor, the packed-bed bioreactor, the
rotating drum bioreactor, and the mechanically mixed bioreactor. These are
briefly explained in the following sections.
Tray bioreactor
A tray bioreactor typically comprises a large number of individual flat trays,
stacked one above the other with a gap. Aeration air with control of humidity and
temperature is forced into the chamber and circulates around the surfaces of trays.
Individual trays are made of wood, metal or bamboo. They have shallow depth (2
to 4 cm) and perforated bottoms to increase the accessibility to oxygen and heat
dissipation (Gowthaman et al., 2001; Krishna, 2005; Mitchell et al., 2006).
Chapter 2 Literature review
36
Figure 2.13 Basic design features of the diverse solid-state fermentation bioreactors. Adapted from (Mitchell et al., 2006)
Packed-bed bioreactor
A Packed-bed bioreactor in which the solid is generally packed into cylindrical
column and air is blown in the base plate from the bottom to provide oxygen and
cool the bed (Mitchell et al., 2000). For commercial implementation of solid state
fermentation, packed-bed bioreactors have the advantage of simple features,
larger capacity and better aeration control than tray bioreactor. Moreover, It is
easy to connect a trickle bed extractor to column reactor for product recovery
after the fermentation (Gowthaman et al., 2001; Krishna, 2005). However, it has
been reported that temperature and steep gas occur in this type of bioreactor,
even though the saturated air was supplied to alleviate this problems. Thus, Lu et
al. (1998) proposed the multi-layer packed-bed reactor to improve the mass
transfer and heat dissipation and reduce the risk of channelling happened in the
traditional packed bed bioreactor.
Rotating drum bioreactor
A rotating drum bioreactor in which the bed is continuously mixed or mixed
periodically with a specific frequency (Mitchell et al., 2006). They typically are
Chapter 2 Literature review
37
horizontal or inclined cylinders which allow the mixing by rotation of the partially
filled solid bed. The drum is rotated slowly to gently mix the substrate around 1-
15 rpm. The tumbling reaction could reduce the heterogeneity of solid substrate
and heat accumulation during fermentation (Gowthaman et al., 2001). However,
the solid particles tend to agglomerate over a period of time in this type of
bioreactor and also the rotational action could damage shear sensitive mycelium
of fungal cells (Gowthaman et al., 2001).
Chapter 2 Literature review
38
2.5.6 Mathematical models of SSF
In solid-state fermentation, macroscale and microscale sub-models can be
thought in the development of mathematical models. The heat and mass transfer
across the bed as a whole and between the headspace and the bioreactor wall are
described in the macroscale sub-model (Figure 2.14). In contrast, the microscale
sub-model describes the growth kinetics of the microorganism and how these
kinetics depend on various mechanisms that occur at the scale of individual
substrates (Viccini et al., 2001).
Figure 2.14 The marcoscale and microscale process happened within an SSF bioreactor (Viccini et al., 2001).
Because of the complicated transport phenomenon within SSF bioreactors, the
majority of bioreactor models only have simple empirical kinetic sub-models
(Mitchell et al., 2003). Different significant gradients (gas, moisture and
Chapter 2 Literature review
39
temperature concentration) happened within a packed-bed bioreactor can be
measured during fermentation. Therefore, the balance equations in SSF
bioreactors are required to use partial differential equations, which need more
computational effort than ordinary differential equations to represent the
balances of systems. Thus, once the system could provide efficient mixing and
good heat dissipation at the marcoscale, it could be considerably simplifying the
balance equations and allowing the assumption that whole substrates are
subjected to identical external conditions (Mitchell et al., 2004).
The microscale phenomena included in the growth kinetic model comprise: the
growth rate of the microorganism depends on the environmental conditions such
as pH, nutrient concentrations, temperature, and water activity; intra-particle
processes such as hydrolysis of polymers and the diffusion of enzymes and
hydrolysis products (Viccini et al., 2001). Further discussion on the kinetic of fungal
growth is included in Chapter 7, where different techniques based on dry weight,
glucosamine and metabolic measurements are analysed and studied. The balance
equations of packed-bed bioreactor will be also investigated.
2.5.7 Application of SSF
A solid culture fermentation has been widely used especially in East Asia for
traditional food fermentation, ethanol production by Koji process, mould-ripened
cheese, and composting process (Kim et al., 1985). Nowadays, numerous products
such as enzymes, organic acids, biogas, compost, antibiotics, surfactants,
fermented foods are produced by SSF techniques (Manpreet et al., 2005; Robinson
et al., 2001).
Since in this research, SSF would be conducted using Trichoderma longibrachiatum,
only the SSF application involving Trichoderma species are reported. The potential
of Trichoderma species as lignocellulose-degrading microorganisms was
recognized in the early 1960s (Selby and Maitland, 1967). Trichoderma species are
frequently found growing in moist soil and on other substrates, such as wood and
Chapter 2 Literature review
40
bark. In addition, they are well known for its ability to produce certain extracellular
enzymes and are already used for production of a variety of extracellular enzymes
and metabolites, such as cellulase, hemicellulase, beta-glucosidase, proteins at an
industrial scale.
Nowadays, filamentous fungi are the major producer for commercial cellulases
production. Cellulolytic fungi belonging to the genera Trichoderma (T. viride,
T. longibrachiatum, T. reesei) have been considered the most powerful and
productive destroyers of crystalline cellulose (Gusakov, 2011; Merino and Cherry,
2007). Commercial cellulase preparations based on several Trichoderma species
are produced on an industrial scale by many companies worldwide (Table 2.4). The
demand for cellulases is consistently on the rise because of its various applications
such as textile detergent, bioenergy and paper industries. However, commercial
enzymes are still predominantly produced by submerged fermentations instead of
solid state fermentations due to several challenges (Waites et al., 2009).
Table 2.4 Commercial cellulases produced by companies and their sources
(Singhania et al., 2010)
Enzyme samples Supplier Source
Cellubrix (Celluclast) Novozymes, Denmark T. longibrachiatum and A. niger
Novozymes 188 Novozymes A. niger
Cellulase 2000L Rhodia-Danisco (Vinay, France) T. longibrachiatum/T. reesei
Rohament CL Rohm-AB Enzymes (Rajamaki, Finland) T. longibrachiatum/T. reesei
Viscostar 150L Dyadic (Jupiter, USA) T. longibrachiatum/T. reesei
Multifect CL Genencor Intl. (S. San Francisco, CA) T. reesei
Bio-feed beta L Novozymes T. longibrachiatum/T. reesei
Energex L Novozymes T. longibrachiatum/T. reesei
Ultraflo L Novozymes T. longibrachiatum/T. reesei
Viscozyme L Novozymes T. longibrachiatum/T. reesei
Cellulyve 50L Lyven (Colombelles, France) T. longibrachiatum/T. reesei
GC 440 Genencor-Danisco (Rochester, USA) T. longibrachiatum/T. reesei
GC 880 Genencor T. longibrachiatum/T. reesei
Spezyme CP Genencor T. longibrachiatum/T. reesei
GC 220 Genencor T. longibrachiatum/T. Reesei
Accelerase®1500 Genencor T. Reesei
Cellulase AP30K Amano Enzyme A. niger
Cellulase TRL Solvay Enzymes (Elkhart, IN) T. reesei/T. Longibrachiatum
Econase CE Alko-EDC (New York, NY) T. reesei/T. Longibrachiatum
Cellulase TAP106 Amano Enzyme (Troy, VA) T. viride
Biocellulase TRI Quest Intl. (Sarasota, FL) T. reesei/T. Longibrachiatum
Biocellulase A Quest Intl. A. niger
Ultra-low microbial (ULM) Iogen (Ottawa, Canada) T. reesei/T. Longibrachiatum
Chapter 2 Literature review
41
Several industrial important enzymes produced under SSF using Trichoderma
species have been extensively studied. For example, (Brijwani et al., 2010) co-
cultivated T.reesei and A. oryzae to produce cellulase enzymes in the SSF using
soybean hull and bran as a nutrient medium. Numerous studies cultivated T.
longibrachiatum to produce cellulase and xylanase in SSF using wheat straw,
wheat bran, and sugarcane straw as the solid substrate (Azin et al., 2007; Guerra
et al., 2006; Kovacs et al., 2004; Ridder et al., 1999). As can be seen above, fungi
in the genus Trichoderma have an ability to produce cellulolytic enzymes on a
broad spectrum of substrates, especially in lignocellulosic materials. This feature
leads to an extension of the study aiming to utilize agricultural residues to produce
on-site enzymes and lignocellulose deconstruction simultaneously which is
reported in this thesis.
Enzymatic hydrolysis of lignocellulose A wide range of agricultural wastes include protein, carbohydrates, and fibre in
significant amounts, making it an ideal substrate to produce value-added products.
However, many microorganism prefer to use monosaccharide (glucose, xylose etc.)
instead of polysaccharide. Therefore, hydrolysis of these natural polymer
(cellulose, hemicellulose) to produce sugars after pretreatment process is
important and critical step for biorefinery.
Today, the major strategy used for lignocellulosic ethanol production comprises
three main steps: (1) pretreatment process (2) enzymatic hydrolysis and (3)
ethanol fermentation. The enzymatic hydrolysis contributes significantly to the
production cost of cellulosic ethanol, and hence the improvement in the enzyme
digestibility efficiency is a prerequisite. As shown in Figure 2.15, a lot of nutrient
materials for growth of microorganisms could be used once these long chain
polymer be degraded. To utilise the lignocellulosic materials completely, different
enzymes need to be required to hydrolyse for specific polysaccharides and
oligosaccharides. However, high cost of commercial enzyme cause obstacle to use
diverse enzyme to obtain abundant nutrient solution. The key method to solve the
Chapter 2 Literature review
42
problem is to produce in-situ enzyme complex from natural materials using fungi
instead of using commercial enzyme.
Figure 2.15 Flow diagram of polysaccharides and the hydrolytic enzymes used to cleave them into shorter polymers and simple sugars (Schuster and Chinn, 2013)
The use of insoluble and highly heterogeneous substrate poses several challenges
in conversion to fermentable sugars. For example, hydrolysis at high solids
concentrations can lead to the problem of product inhibition and insufficient
mixing, which results in lower performance of enzyme digestibility. The presence
of lignin, which shields the cellulose chain and interferes or adsorbs the enzyme
complex, causes non-productive binding. In addition, the enzyme activities may be
decreased as time goes on due to denaturation. There may also be other, as
unidentified, reasons for decreased conversion (Kristensen et al., 2009; Puri et al.,
2013).
At the end of enzyme hydrolysis, high fermentable sugar concentrations are
preferable for the subsequent fermentation. This will increase the product titres
and facilitate the efficiency of downstream processing, particular in product
recovery. However, the heterogeneity and structure of lignocellulosic materials,
Chapter 2 Literature review
43
high viscosity of substrates, is an important factor to limit efficient mixing using
high solid substrates. The viscosity of slurries increases dramatically over a certain
level (typically 20% of solids ratio), and appear the linearity of the solids effect
with increasing loading. One alternative strategy to overcome is the fed-batch
substrate addition, which reduced the initial viscosity to allow well mixing
(Puri et al., 2013).
Chapter 2 Literature review
44
Microbial feedstock production It is widely accepted that the finite resource we currently use as feedstock for the
production of chemicals and energy will have been consumed and chemical
processing will be obliged to turn to bioprocessing gradually to achieve sustainable
development. However, whatever other factors influence the choice, the real
driving force for the adoption of new process is the relative production cost
associated with obtaining materials and processing compared with traditional
process. Also, the cost of microbial medium is one of the major obstruction facing
the successful transition from hydrocarbon-based to sugar-based chemical and
fuel production (Koutinas et al., 2004; Lynd et al., 1999; Webb and Wang, 1997).
2.7.1 Reducing sugars production from lignocellulose
The most common source of carbon for fermentation is fermentable sugar (usually
as glucose, xylose or sucrose). For fermentation medium preparation, hydrolysing
lignocellulose to fermentable monosaccharides efficiently is a major task. Overall,
typical sugar production from lignocellulose includes several steps: biomass
pretreatment, detoxification of liquid fraction and cellulose hydrolysis
(Figure 2.16). As mentioned before in section 2.4, the typical principle of
pretreatment is to hydrolyse hemicellulose, lignin into solution and remains
cellulose unaltered in the solid residue and can be further processed to produce
glucose. Acid or steam pretreatment conducted at a high temperature is one of
the most effective ways due to high glucose yield after enzyme hydrolysis. The
severe conditions, however, lead to microbial inhibitors generation from
monosaccharides such as furfural and hydroxymethyfurfural (HMF) (Cardona et al.,
2010). Also, alkaline pretreatment can induces the inhibitors produced, the costly
detoxification processes are necessary to remove these inhibitory byproducts (Bak
Figure 2.18 Overview of fungi autolysis (White et al., 2002)
Chapter 2 Literature review
51
Little is known about the effects of reactive oxygen in the filamentous fungi. Thus,
investigations about the relationship between oxidative stress and fungal autolysis
could be relevant to submerged fungal bioprocesses. This significance is due to the
supply of high dissolved oxygen concentrations for most fungal growths at the
industrial scale production (White et al., 2002). In 2010, Wang et al. showed that
fermented materials were transferred to another vessel for further hydrolysis at
the end of solid-state fermentation. Biological reaction in the process of hydrolysis
might include not only enzymatic hydrolysis of materials but also the autolysis of
fungi cells. The oxygen limited environment could lead the autolysis of strict
aerobic fungi in a submerged environment (Wang et al., 2010).
Chapter 3 Objective and research programme
52
CHAPTER 3
Objectives and research programme
Introduction In the previous chapter it was pointed out that lignocellulosic residues do not
currently compete with food demands and can therefore provide a low cost
feedstock for production of commodity chemicals and fuels, offering economic
and environmental advantages. The aim of the project described in this thesis was
to demonstrate the feasibility of utilising lignocellulosic residues to produce value-
added products through sequential bioprocessing. As part of this goal, the project
was focused on the development of a novel strategy for the production of generic
fermentation feedstocks based on solid state fermentation (SSF) bioprocessing
using sugarcane bagasse and soybean hull.
The following section 3.2 is basically a description of the research idea based on
the original research plan, following an analysis of the lignocellulosic biorefinery
process and numerous researches on grain-based generic fermentation feedstock
production in the Satake Centre for Grain Process Engineering (SCGPE). The origin
of the proposed research idea came from the modification of typical lignocellulose
pretreatment process, advantageous features of solid state fermentation and the
production of the generic feedstock from wheat grains and rapeseed meals. The
resultant study focused primarily on the development of a process for the
production of a generic microbial feedstock. This was based on hydrolysis and
fungal autolysis as a nutrient production method using fermented solids from solid
state fermentation of Trichoderma longibrachiatum on agricultural residues.
Chapter 3 Objective and research programme
53
Proposed bioprocess concept In conventional processing for lignocellulose, the pretreatment process involves
either physical, chemical or physico-chemical treatment. These are either energy
intensive or produce large effluent discharge problems. The use of a biological
treatment instead of other methods could reduce considerably the energy
consumption and effluent release of the process.
Attempts at bioprocessing involving submerged fermentation (SmF) using
filamentous fungi have suffered from viscosity problems and poor mass transfer
due to difficulty in controlling the fungal morphology during fermentation and
lignocellulose suspension in liquid medium (Lan et al., 2013). The inherent
advantage of solid state fermentation is that it provides growth conditions that
are similar to the environment to which filamentous fungi are naturally adapted.
Higher enzyme concentrations could be achieved because of better interaction
between the microorganism and the substrate in SSF.
This novel process concept is represented schematically in Figure 3.1. The
proposed process includes two key steps. In the first, a SSF with the fungus
Trichoderma longibrachiatum using sugarcane bagasse and soybean hull, is used
to simultaneously secrete cellulolytic enzymes and deconstruct recalcitrant
lignocellulose. This consolidated fungal fermentation not only eliminates the
expensive pretreatment stage but also provides abundant enzymes (cellulase,
beta-glucosidase, xylanse etc.) for further hydrolysis. At the end of the SSF, the
whole fermented solids, containing the enzymes, fungal mycelium and residual
substrate, are transferred and blended in sterile distilled water to cut the
mycelium and make the suspension homogeneous. The suspension is then stored
at higher temperature and oxygen-limited conditions to elevate enzyme activities
in the mash for hydrolysing the remaining substrates to produce sugars, free
amino nitrogen (FAN), inorganic phosphorous (IP) and other nutrients. The
purpose of using high temperature for hydrolysis is, not only to prevent fungi
growth, but also to achieve optimal enzyme activities for hydrolysis of the
Chapter 3 Objective and research programme
54
lignocellulose. Previous studies have shown that the growth of bacterial
contaminants was also effectively prevented at temperatures above 50 °C in
hydrolysis of fermented rapeseed meal (Wang et al., 2010). On the other hand,
oxygen-limited conditions could trigger fungal autolysis and release nutrients by
degrading cell walls to give the generic microbial feedstock more nutrients (sugars,
FAN, IP etc.) rather than sugar solution only. Fungal cell wall disruption and
macromolecules hydrolysis, therefore, lead to the production of a nutrient-rich
solution containing amino acids, peptides, nucleotides, phosphorous and vitamins
as well as other trace elements.
Figure 3.1 Proposed process for a generic microbial feedstock production from lignocellulose
The further fermentation step is carried out using the microbial feedstocks
obtained from the above proposed process. For example, submerged
Fermentation with yeast was used to test the suitability of the microbial
feedstocks for the production of the ethanol (reported in chapter 8).
Chapter 3 Objective and research programme
55
Research objectives In order to establish a process for the biological pretreatment of sugarcane
bagasse, four major objectives were identified. The first was to grow the fungus
on the substrate. Secondly, testing the enzymes produced during growth for the
ability to further hydrolyse the substrate was key to the process. The third and
fourth objectives were then to carry out trials in bioreactors and to use the
resultant medium for a subsequent fermentation process. Each objective
is described in more details in the following sections.
3.3.1 Growth and adaptation of T. longibrachiatum
The early part of the project was focused on acquiring knowledge about the fungi
growth on lignocellulosic material as carbon source. Trichoderma longibrachiatum
cultivation in varied conditions, particle size, C/N ratio, different proportions of
substrates, incubation time and environment moisture level, were investigated.
After solid state fermentation (SSF), the whole fermented solids were transfer to
sequential hydrolysis stage. The principal components in the nutrient medium
obtained from the solid state fermentation followed by further hydrolysis are
reducing sugars and free amino nitrogen (FAN). These total reducing sugars and
FAN were to be used as indicators to evaluate the performance of the bioprocess.
The objectives of these experiments were to identify a set of improved conditions
for the growth of T. longibrachiatum on sugarcane bagasse and soybean hull and
their effects on nutrients production.
3.3.2 Sequential hydrolysis of fermented solids
The aim of this part of the study was to investigate the optimum hydrolysis
conditions (solid loading ratio, temperature, pH, microbial inhibitor addition) for
efficient microbial feedstocks production from fermented solids of
T. longibrachiatum. Investigations of kinetics on further hydrolysis and fungal
autolysis help to understand the hydrolysis profile for further application.
Moreover, the characteristics of cellulase, xylanase and beta-glucosidase were
Chapter 3 Objective and research programme
56
investigated by means of determination of physico-chemical parameters
e.g. optimal temperature and pH.
3.3.3 Bioreactor Studies
This experimental stage involved studies of microbial feedstocks production
through solid state fermentation (SSF) in a single-layer tray bioreactor, a multi-
layer tray bioreactor and a packed-bed bioreactor. To study the growth kinetics of
the microorganisms in SSF, biomass content measurement, substrate
concentration and other process parameters are required. Because separating
filamentous fungi from solid substrates is difficult, indirect measurements such as
carbon dioxide, oxygen and glucosamine were employed to examine
quantitatively the fungal growth, substrate consumption kinetics in solid state
fermentation.
3.3.4 Ethanol production from bagasse derived feedstock
To produce ethanol, S. cerevisiae was used in the nutrient-rich medium produced
from sugarcane bagasse and soybean hull. Nutrient utilisation (sugars, FAN and
Inorganic phosphorous) as well as ethanol production were observed then
compared with that from glucose-based media (glucose and yeast extract). A mass
balance of overall ethanol production using feedstock derived from sugarcane
bagasse and soybean hull was presented. A preliminary consideration of economic
feasibility, based on the results of the study, provided a further objective.
Chapter 4 Materials and methods
57
CHAPTER 4 Materials and methods
Introduction The experimental methods in all practical work described in this thesis are
presented in this chapter, including fermentation methodology and product
analysis. Analytical methods are also given and theoretical information presented
where appropriate. Research results from experiments are presented in Chapters
5, 6, 7 and 8, where brief summaries of the relevant experimental methods are
also given.
Materials and microorganisms
4.2.1 Sugarcane Bagasse and Soybean Hull
Sugarcane bagasse was provided by Dr. R.F. Chang, Taiwan, packaged in vacuum-
packed bags to maintain freshness (Figure 4.1). To adjust the different particle
sizes for experiments, sugarcane bagasse was ground using a kitchen blender then
passed through three different sieves (1.4 mm, 0.85 mm and 0.5 mm) and the
fractions were stored in air-tight plastic containers until used.
Soybean hull was obtained from Cargill Plc, Liverpool, United Kingdom (Figure 4.1).
It was also ground using a kitchen blender then passed through a 2 mm sieve to
remove other debris and stored at room temperature in an air-tight plastic
container. After adjustment to the desired moisture content and mixed substrates,
the conditioned substrates were sterilised at 120°C in the autoclave for 20 min.
Prior to carrying out chemical analysis, both substrates were ground with a small
hammer mill (Glen Creston, DFH48) fitted with a 0.8 mm sieve for 3 times for
accurate sampling.
Chapter 4 Materials and methods
58
Figure 4.1 Appearance of (a) sugarcane bagasse and (b) soybean hull
4.2.2 Microorganisms and inoculum preparation
4.2.2.1 Microorganism used in solid state fermentation
Many microorganisms are known to produce different type of lignocellulolytic
enzymes. Among them, Trichoderma spp. are widely known as a lignocellulose
decomposer since they are filamentous and have the ability to produce prolific
spores and abundant lignocellulolytic enzymes which can invade substrates
quickly. In this PhD thesis, Trichoderma Longibrachiatum was used because of the
ability to generate high levels of both cellulolytic and xylanolytic enzymes. The
fungus, Trichoderma Longibrachiatum, obtained from the fermentation
engineering laboratory of the School of Chemical Engineering and Analytical
Science, Faculty of Engineering and Physical Science, University of Manchester, UK
was used throughout this study for solid state fermentation.
The spores were purified in order to ensure that the inocula used throughout the
research study were initiated, each time, from a single spore. First of all, a small
amount of the spore suspension (0.5 mL) obtained from previous culture was
spread on the surface of the 50 mL PDA medium (15 g/L agar, 20 g/L dextrose, and
4 g/L of potato extract) in 250 mL Erlenmeyer flasks, and the inoculated flasks were
incubated at 30°C for 7 days. Several glass beads (4-mm diameter) and 50 mL of
sterile 0.1% (v/v) Tween 80 were added to the flasks to extract and form a stock
spore suspension after gentle shaking. Monospore isolation was performed by
Chapter 4 Materials and methods
59
serially diluting the fungal spore suspension from 101 to 109. A 10 μL aliquot of
each serially diluted spore suspension was plated onto solid PDA medium and
incubated at 30°C for 7 days (Figure 4.2). Then an isolated colony was selected
from the Petri dish and transferred onto slopes of the same sporulation medium
in test tubes for spore formation and incubated for another 7 days at 30°C. After
this incubation period the slopes were transferred and stored at 4°C.
Figure 4.2 Trichoderma Longibrachiatum on PDA medium agar plate
(Growth of colony after 7 days at 30°C).
In these studies, untreated sugarcane bagasse and soybean hull, were used for
growing T. longibrachiatum. The ratio of sugarcane bagasse and soybean hull were
adjusted according to the purpose of experiments. For each set of preliminary
experiments in Petri dishes (chapter 5 and 6), a total weight of 4 g (dry weight) of
the mixed substrates were used by. All samples were put into 250 mL flasks and
sterilized at 121°C for 20 min. The substrates were allowed to cool to room
temperature before inoculating with T. longibrachiatum and moistened with an
amount of sterilized mineral salt medium (CaCl2‧2H2O: 0.5g/L; KH2PO4: 3g/L;
MgSO4: 0.5g/L) to obtain the desired moisture content. The spores were harvested
from the slope with 10 mL of sterile 0.1% (v/v) Tween 80. 1 mL of the spore
suspension (106 spores/mL) was inoculated and mixed into the substrate. Then
whole inoculated samples were transferred to petri dishes. All petri dishes were
Chapter 4 Materials and methods
60
incubated at 30°C for 5 days, and samples were taken to measure total reducing
sugar, free amino nitrogen (FAN), moisture content, weight loss and enzyme
activity (cellulase, xylanase and beta-glucosidase).
4.2.2.2 Microorganism used for ethanol fermentation
Yeast fermentations were carried out using Saccharomyces cerevisiae ATCC 22602
for the production of ethanol. Yeast cells were cultivated in standard YDX medium
(5 g/L yeast extract, 10 g/L xylose, and 5 g/L glucose) for inoculum preparation.
Reactor systems
4.3.1 Petri dish
Petri dishes of 10 cm of diameter and 1 cm height were used during the project to
carry out several preliminary experiments. These simple reactor systems which
have some similarities with tray bioreactors were filled with the inoculated
substrates and were placed in an incubator at 30°C where the cultivation was
carried out statically. Individual petri dishes, from a parallel set, were sacrificed for
analysis as required.
4.3.2 Circular tray bioreactor
Circular tray bioreactors were used as an extension to the simple petri dish system.
One of these is shown in Figure 4.3, and consisted of a single circular perforated
tray with 10.0 cm diameter. The bioreactors were made of polycarbonate and
were autoclave compatible. The system (Figure 4.4) was used to determine the
kinetic parameters for the fungal growth and the consumption of substrate. Due
to their similarity, five circular tray bioreactors could be operated in parallel, and
like petri dishes, one could be sacrificed per day for analysis (e.g. at specific time
24, 72, 96, 120, 168 h). The bioreactors were connected to humidifier and gas
distributor and fermented in the incubator at 30°C. The airflow of air supply was
maintained at 0.6 L/min per gram of substrate for each bioreactor. The outlet gas
was passed through a filter and silica gel tube to reduce the humidity and
Chapter 4 Materials and methods
61
impurities of exhaust gases before entering to a gas analyser (FerMac 368,
Electrolab, UK) to sample on-line CO2 and O2 data.
Figure 4.3 Left: A circular tray bioreactor. Right: A perspective view of bioreactor
Figure 4.4 A diagram of the system developed for on-line automated monitoring of solid state fermentation (1) Regulated pressure air inlet; (2) 0.2 µm filter (3) humidifier; (4) air distributor; (5) Circular tray fermenters; (6) Silica gel tube (7) Gas analyser.
Chapter 4 Materials and methods
62
4.3.3 Multi-layer tray bioreactor
As shown in Figure 4.5, the multi-layer tray bioreactor contained three sieve trays
with 10.0 cm diameter and 5.0 cm height, which were tightly stacked one over
another, and a bottom tray that was closed. This bottom tray, with the same
diameter and 3.5 cm height, acted as an air distributor. The forced air passed from
the bottom to the top tray by continuous aeration. The stacked trays were sealed
in such a manner that prevented air exchange from the outside to the inside
environment and vice versa. The bioreactor, adapted from a set of laboratory
sieves, was constructed from stainless steel and aluminium. The mesh bases, with
specific aperture sizes (500 and 600 μm) allowed uniform distribution of air, as
well as supporting the solid particles on the tray.
Figure 4.5 Left: A multi-layer trays bioreactor. Right: A perspective view of bioreactor
A schematic diagram of the multi-layer tray bioreactor system is shown in Figure
4.6. The compressed air entered the system at 0.6 L/min·g and was sterilised using
an air filter. Sterilized thermocouples were positioned at all three trays to record
the bed temperature using a multichannel temperature data logger (Microlab II,
Chapter 4 Materials and methods
63
Aglicon, UK). The outlet gas was passed through a filter and silica gel tube to
reduce the humidity and impurities of exhaust gases before entering to a gas
analyser (FerMac 368, Electrolab, UK) to sample on-line CO2 and O2 data.
Figure 4.6 Schematic diagram of the multi-layer tray bioreactor system (1) Compressed air, (2) air flow meter, (3) 0.2 µm air filter, (4) humidifier, (5) water bath at 30°C, (6) incubator at 30°C, (7) multi-layer circular bioreactor, (8) silica gel,
(9) thermocouples type K, (10) temperature data logger, (11) gas analyser, (12) computer
4.3.4 Packed-bed bioreactor
The generic microbial feedstocks production using T. longibrachiatum was
implemented in a 1.5 L packed bed bioreactor. After the inoculation, the
substrates were transferred to the sterilised reactor (1.5 h at 121°C), which was
set up as shown in Figure 4.7.
The reactor consisted of a sterilisable glass cylinder of 8 cm diameter and 30 cm
height, closed at both ends with stainless-steel plates. The bottom plate had a port
for air inlet (0.6 L/min·g) and the top plate had a port for the exhaust gas. Three
temperature probes (thermocouples type K) at 5.5, 13.8, and 21.9 cm from the
bottom and two sampling points at 6.4 and 17.4 from the bottom were located
through the glass cylinder. The reactor was placed into an incubator at 30°C
Chapter 4 Materials and methods
64
instead of using the water jacket. Sterilized thermocouples were used to record
the bed temperature using a multichannel temperature data logger (Microlab II,
Aglicon, UK). The outlet gas was passed through a filter and silica gel tube to
reduce the humidity and impurities of exhaust gases before entering to a gas
analyser (FerMac 368, Electrolab, UK) to sample on-line CO2 and O2 data.
Figure 4.7 Schematic diagram of the 1.5 L packed-bed bioreactor system (1) Compressed air, (2) air flow meter, (3) 0.2 µm air filter, (4) humidifier, (5) water bath at 30°C, (6) relative humidity data logger, (7) incubator at 30°C, (8) packed-
bed bioreactor, (9) thermocouples type K, (10) temperature data logger, (11) silica gel, (12) gas analyser, (13) computer
4.3.5 Enzyme hydrolysis and fungal autolysis system
At the end of solid state fermentations, the contents of petri dishes, bioreactors
or flasks were suspended in sterilised solution (citric buffer or water). The
suspension was stored in 500 mL Duran bottles under oxygen-limited condition at
specific temperature (40 to 60°C), 160 rpm for 48 hours to elevate the enzyme
activities in the solids for hydrolysing the remaining components.
4.3.6 Ethanol fermentation system
The strain Saccharomyces cerevisiae was used for ethanol production experiments.
Fermentation experiments were performed in Erlenmeyer flasks (250 mL)
containing 100 mL of hydrolysates without any supplement, at 32°C for 30 hours.
Chapter 4 Materials and methods
65
Analytical methods
4.4.1 Fungal spore count
The concentration of the spore suspension was measured using a
haemocytometer (Figure 4.8 and 4.9). The suspension should be dilute enough so
that the spores do not overlap each other on the grid of the haemocytometer and
should be uniformly distributed as it is assumed that the total volume in the
chamber represents a random sample. To prepare the haemocytometer, the
mirror-like polished surface is carefully cleaned with lens paper and 75% ethanol.
The spore suspension is introduced onto the surface using a Pasteur pipette (about
200 µL). The cover slip is then placed over the counting surface. The
haemocytometer is then placed on the microscope stage and the counting grid is
brought into focus at low magnification (100X) and then followed with higher
magnification at 400X. The spores are counted in selected squares so that the total
count is around 100 spores or so (number of spores needed for a statistically
significant count). The main divisions separate the grid into 9 large squares. Each
square has a surface area of 1.0 mm2. The depth of the chamber is 0.1 mm. Each
square of the haemocytometer with the cover slip in place represents a total
volume of 0.1 mm3. The subsequent spore concentration per mL can therefore be
determined (Figure 4.8). If there are less than 50, or more than 200 spores per
large square, the procedure is repeated by using an appropriate dilution factor.
The subsequent spore concentration per mL and the total number of spores is
determined using the following equation.
Number of Spores per mL = [the average count per square] x [dilution factor] x 104
Equation 4-1
Total spores number = spores per mL x the original volume of spores stock
suspension from which spores sample was removed.
Chapter 4 Materials and methods
66
The volume of suspension needed for the inoculation of the solid was calculated
for each experiment to reach a concentration of around 1 x 106 spores/g solid
substrate.
Figure 4.8 Spores counting on haemocytometer
Figure 4.9 Spores of Trichoderma Longibrachiatum on Haemocytometer grid (magnification 100X)
4.4.2 Analysis of the moisture content of materials
Moisture content of the samples was determined by weight loss after heating to
constant weight at 95°C. After being dried in the oven for 24 hours, the samples
were cooled in desiccators for 30 min before being weighed.
Chapter 4 Materials and methods
67
Moisture content was calculated as follows:
M(d.b)=(Wi−Wf)
(Wf−Wd)× 100 Equation 4-2
Where:
M(d.b): Moisture content in dry basis (g water/g dry solid) (on a dry solid basis)
Wi: Total weight of the dish with the materials before drying (g)
Wf : Total weight of the dish with the materials after drying (g)
Wd: Weight of the dish (g)
4.4.3 Properties of solid substrate
Following the procedure described by (The GLOBE program, 2002), the bulk
density, particle density and porosity of sugarcane bagasse and soybean hull were
measured using the soil particle density protocol. Density is measured as mass per
unit volume. Solid substrate density depends on the chemical composition and
structure of the solid substrate. Density of any materials (solid substrate) can be
divided into two categories: (1) bulk density and (2) particle density. Bulk density
refers to the volume of the solid portion of the solid substrate particles along with
the spaces where the air and water exist. In contrast, particle density is only
concerned with solid substrate particles and the pore spaces occupied within the
solid substrate. Bulk density is used along with particle density to calculate
porosity. Porosity is defined as the ratio of the volume of voids to the total volume
(void + solid) (Figure 4.10).
Chapter 4 Materials and methods
68
Figure 4.10 Porosity
Adapted from (The GLOBE program, 2002)
Bulk density
Solid substrates (sugarcane bagasse and soybean hulls) were oven dried at 50°C
for 24h, then poured into a measuring cylinder of known volume (50.0 mL) and
weighed to determine the bulk density. Bulk density of solid particles was
18.1: Millimolar extinction coefficient of p-nitrophenol under the assay condition
(cm2/micromole)
Chapter 4 Materials and methods
86
Reaction time: 15 min
4.4.13 Ethanol
Ethanol concentration was measured directly using an Analox GL6 analyser
(Analox, England).
Chapter 5 Production of a generic feedstock: solid-state fermentation
87
CHAPTER 5
Production of a generic feedstock:
solid-state fermentation
Introduction In the consolidated bioprocess, the solid material plays an important role as
physical support and nutrient supply. Consequently, the characteristics of the raw
material (chemical composition and physical properties such as bulk density,
particle density, porosity and water evaporation) will influence the microorganism
growth. A good knowledge of material characteristics could greatly help the
selection of substrates, as well as design and optimisation of the solid state
fermentation.
The ability of Trichoderma spp. to produce secondary metabolites, especially
cellulase and xylanase, is well known. However, the evaluation of performance in
this project is the production of nutrients (sugars, nitrogen and other minerals)
through the consolidated bioprocess rather than enzyme secretion from
fermentation. Trichoderma longibrachiatum has not been previously studied for
this consolidated process. First, a technique for studying a sequential bioprocess
using lignocellulose had to be developed. This was then used to collect data and
optimise conditions for further bioreactor application.
Several different experimental conditions reported for Trichoderma species in SSF
influence enzyme yield and deconstruction of lignocellulose, including particle size
of substrate, C/N ratio of substrate, and moisture content of substrate. Other
factors that affect Fungi growth during SSF include temperature, aeration and
incubation time. The study reported in this chapter was therefore aimed at
identifying suitable conditions for growth and nutrients production with a focus
Chapter 5 Production of a generic feedstock: solid-state fermentation
88
on the sugar and FAN. All results obtained were used as fundamental knowledge
for growing T. longibrachiatum in subsequent experiments.
The characteristics of the substrates
5.2.1 Chemical composition
The chemical composition of sugarcane bagasse and soybean hull is presented in
Table 5.1. Half of the chemical composition of sugarcane bagasse and soybean hull
is cellulose and hemicelluloses. Notably lignin content is quite different between
sugarcane bagasse (17.8) and soybean hull (8.4), causing the fungal growth to a
certain extent.
Table 5.1 Composition of sugarcane bagasse and soybean hull on dry basis
Type Cellulose Hemicellulose Lignin Protein* Others
Sugarcane
bagasse
30.9% 31.9% 17.8% 2.5% 16.9%
Soybean
hull
31.1% 21.9% 8.4% 11.9% 26.7%
*Protein content of sugarcane bagasse and soybean were obtained from references (Jenkins et al., 1998; Ontario minister of agriculture, food and rural affairs, 2011)
An important factor in solid-state fermentation is the ratio between carbon and
nitrogen (C/N). The ratio of C/N is most vital for a specific process to obtain desired
products. Sugarcane bagasse has higher cellulosic composition which is ideal for
good growth of fungal cultures and enzyme production. However, its low protein
content (2.5%) and high lignin concentration limit its efficiency of bioconversion
to produce value-added products (El-Sayed et al., 1994). Brijwani et al., (2010)
showed that the protein content of soybean hull (11.9%) is not as high as wheat
bran (16.29%). However, Soybean hull is still a good source of nitrogen compared
with sugarcane bagasse. Mixtures of sugarcane bagasse and soybean hull can,
Chapter 5 Production of a generic feedstock: solid-state fermentation
89
potentially, improve C/N to present ideal conditions for fungal growth and enzyme
production and this is explored below.
5.2.2 Bed Porosity
As has been defined, SSF involves a discrete solid phase in which microorganisms
grow on the surface of moist particles as well as inside and between them. The
space between particles is occupied by a continuous gas phase. And the size, shape,
and porosity of substrates could affect the gas phase in the SSF. Availability of
spaces between particles ensures the accessibility of oxygen for enzyme
production in aerobic fungal growth (Brijwani and Vadlani, 2011).
In order to assess the likely importance of particle size and type on bed porosity,
measurements were made of bulk density and particle density and these were
then used to calculate bed porosity for a range of different particle sizes of
sugarcane bagasse as well as for soy bean hull particles. The results, presented in
Table 5.2, show that there is a substantial difference in the porosity, estimated at
dry basis, for different particle sizes of sugarcane bagasse compared with soybean
hull (<2 mm). For sugarcane bagasse, with the increase of particle size, the porosity
raises gradually from 84.2% to 92.9%. The higher value of porosity means the more
open space available between substrate particles. Notably, the maximum
difference of bed porosity between soybean hull (<2 mm) and sugarcane bagasse
(2-1.4 mm) was close to 18%. These results match those observed in earlier studies
that porosity varies depending on several factors such as fibre bonds, moisture,
particle size and aggregation.
Chapter 5 Production of a generic feedstock: solid-state fermentation
90
Table 5.2 Bed porosity of Sugarcane bagasse and Soybean hull particles
Type Bulk Density (g/mL)
Particle Density (g/mL)
Porosity (%)
Sugarcane bagasse (2-1.4 mm)
0.05 0.704 92.9
Sugarcane bagasse (1.4mm-0.85 mm)
0.049 0.6 91.8
Sugarcane bagasse (0.85mm-0.5 mm)
0.058 0.401 85.5
Sugarcane bagasse (0.5mm-0.21 mm)
0.059 0.375 84.2
Soybean hull (< 2 mm)
0.435 1.787 75.6
Figure 5.1 shows photographs of sugarcane bagasse under a microscope at 50X
magnification. For the biggest size of sugarcane bagasse (2-1.4 mm), various
particle shapes are shown in Figure 5.1a. Because of the irregularly size particles,
particles could not pack the inter-particle void space and block the open pore for
air or water accessibility. Therefore, the void spaces between particles are larger
than for other particle sizes. As for the smallest one (0.5-0.21 mm), the quality and
shape of particles is more homogeneous (Figure 5-1d), causing the amount of
open space to reduce. The findings of the current study are consistent with those
of (Manickam and Suresh, 2011) who showed that the porosity is decreasing with
decreasing particle size for corn pitch.
Chapter 5 Production of a generic feedstock: solid-state fermentation
91
Figure 5.1 Sugarcane bagasse at different particle sizes under a USB microscope, 2X magnification (a) 2-1.4 mm (b) 1.4-0.85 mm (c) 0.85-0.5 mm (d) 0.5-0.21 mm
5.2.3 Water evaporation
The water evaporation rate of substrates is related to the efficiency of Solid-state
fermentation, especially for the issue of heat and mass transfer during
bioprocessing. One of the major barriers is the difficulty in controlling the water
content and temperature of the bed in large-scale bioreactors.
For investigating evaporation characteristics of sugarcane bagasse and soybean
hull, it could be useful to use dimensionless moisture ratio (MR) to represent
evaporation behaviour.
MR=(M-Me)
(Mo-Me) Equation 5-1
Where M is the moisture content of the product, Mo is the initial moisture content
of the product and Me is the equilibrium moisture content.
Chapter 5 Production of a generic feedstock: solid-state fermentation
92
The values of Me are relatively small compared to M and Mo for long drying times
and accordingly one can write:
MR=M
M0 Equation 5-2
The non-fermented sugarcane bagasse and soybean hull were dried at 30°C in an
oven, adopting thin-layer thickness of about 10 mm. The initial moisture content
of samples was about 0.75g water per g of dry matter. All samples were put in
petri dishes without lids. Using the moisture ratio to generalize the change of
moisture content is the most common method in drying process. As shown in
Figure 5.2, the moisture ratio versus drying time for sugarcane bagasse and
soybean hull at 30°C. According to the results obtained, the effect of various
particle size of sugarcane bagasse does not cause significant difference of water
evaporation.
Figure 5.2 The drying curve of sugarcane bagasse and soybean hull
The limitation of mass transport by diffusion plays an important role in solid state
fermentation especially when the substrate has a porous structure. Mass transfer
inside the substrate particle is limited to diffusion and because of consumption of
nutrients by the microorganisms, concentration gradients will go up within the
0
0.2
0.4
0.6
0.8
1
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Mo
istu
re r
atio
(M
R)
Time (hr)
bagasse (2-1.4mm)
bagasse (1.4-0.85mm)
bagasse (0.85-0.5mm)
soybean hull
Chapter 5 Production of a generic feedstock: solid-state fermentation
93
substrate. Therefore, understanding this characteristic of the substrate is one of
the crucial factors for the design of solid state fermentation. It has been accepted
that the drying characteristics of biological products in the falling rate period can
be described by using Fick's diffusion equation. Crank (1975) used various regularly
shaped bodies such as rectangular, cylindrical and spherical products, and the
form of Equation 5-3 to apply on particles with slab geometry, as is the case of the
sugarcane bagasse and soybean hull, by assuming uniform initial moisture
distribution (Crank, 1975; M.A. Mazutti et al., 2010).
MR =8
𝜋2∑
1
(2𝑛+1)2∞(𝜋=0) 𝑒𝑥𝑝 (−
(2𝑛+1)2𝜋2𝐷𝑒𝑓𝑓𝑡
4𝐿02 ) Equation 5-3
where Deff is the effective diffusivity (m2 s−1), L0 is the half thickness of slab (m). For
long drying periods, Equation 5-3 can be simplified to retain only the first term of
the series and re-writing to a logarithmic form as follows:
ln(MR) = ln(8
π2) −
π2Defft
4L02 Equation 5-4
Diffusivities are determined by plotting drying data in terms of ln(MR) versus time
in Equation 5-4, providing a straight line with the slope given by:
slope = −π2Deff
4L02 Equation 5-5
Table 5.3 Values of effective diffusivity obtained for different particle sizes of sugarcane bagasse and soybean hull
Sample Effective Diffusivity (m2s-1)
2-1.4 mm of sugarcane bagasse 6.36 x 10-11
1.4-0.85 mm of sugarcane bagasse 6.82 x 10-11
0.85-0.5 mm of sugarcane bagasse 6.59 x 10-11
Soybean hull (<2mm) 5.21 x 10-11
Chapter 5 Production of a generic feedstock: solid-state fermentation
94
The values of effective diffusivity (Deff) of sugarcane bagasse with various particle
size and soybean hull are presented in Table 5-3. The effective diffusivities for
sugarcane bagasse ranged from 6.82 x 10-11 to 6.36 x 10-11 m2s-1, whereas for
soybean hull the value was 5.21 x 10-11. The effective diffusivities of sugarcane
bagasse were, on average, 1.26 times higher than those found for soybean hull. A
possible explanation for this might be that the diffusion rate is proportional to the
porosity of the solids (Table 5.3), and as the drying process takes place the
structure hinders diffusion, diminishing the values of Deff.
Chapter 5 Production of a generic feedstock: solid-state fermentation
95
Effect of washing procedure on sugarcane bagasse
Sucrose is the prime constituent of sugarcane juice. Also, a variety of other
carbohydrates is found in mixed juice. (Walford, 1996) reported that the most
common consist of the monosaccharides glucose and fructose (3-6%), and the
dissacharides, sucrose (81-87%). Oligosaccharides and polysaccharides (0.2-0.8%)
may be present depending on the age of the cane when harvested and
deterioration during delays. In sugar processing, the first step is to shred
sugarcane to prepared cane. The cane juice is then extracted from the prepared
cane and collected through a series of solid-liquid separation processes. The
residue left after crushing is called ‘bagasse’. Sugarcane bagasse is the fibrous
material still containing a significant amount of sugar. In sugar processing plant,
there are five more crushing stages to extract the remaining sugar from bagasse
and the remained residue in the last extraction stage is called ‘final bagasse’
(Tewari and Malik, 2007). Since the bagasse in the experiments reported in this
thesis was collected from a traditional market, the sugarcane was extracted
through a simple sugarcane extruder. It was necessary, therefore, to explore the
influence of the remaining sugar in the bagasse on the solid state fermentation.
Several researches showed that microorganisms are able to use sucrose from
sugarcane bagasse during solid state fermentation (Kumar et al., 2003; Marcio A.
Mazutti et al., 2010). This study was carried out to investigate the influence of
washing the sugarcane bagasse on the sugar and FAN yield obtained from
sequential solid state fermentation (SSF) and hydrolysis.
The solids (sugarcane bagasse) were either used as non-washed or washed with
distilled water before inoculation of solid state fermentation. The amount of
distilled water used in the washing process was about 1 L per 25 g bagasse (dry
basis). The wash liquid was stored at 4°C for further analysis within 3 days.
Chapter 5 Production of a generic feedstock: solid-state fermentation
96
After being autoclaved at 121°C for 20 min, 4 g of non-washed and washed
sugarcane bagasse were then distributed into each of two 9 cm petri dishes and
incubated at 30°C for 120 h. The default culture condition was 65 % moisture
content using sterile mineral salt water, 10% (w/w) of yeast extract supplemented,
106 spores per gram substrate, and 7-day spores if it was not indicated specially.
At the end of solid state fermentation the contents of all Petri dishes were
suspended in pH 4.8 citric buffer. The suspension was stored in sealed Duran
bottles to elevate enzyme activities in the fermented substrate for hydrolysing the
remaining sugarcane bagasse components. All experiments were carried out with
agitation using a shaking incubator at 50°C, 160 rpm for 48 h.
From Figure 5.3a, it was noticed that there was nearly no growth of T.
longibrachiatum on washed sugarcane bagasse. Careful visual observation
showed that there were some mycelia formed on the substrate surface, which
means that T. longibrachiatum did grow on it, but with a slow rate. However, when
non-washed sugarcane bagasse was used for substrate (Figure 5.3b), the growth
was slightly improved.
Figure 5.3 SSF using Trichoderma longibrachiatum on (a) washed and (b) non-washed sugarcane bagasse after 5 days
The composition analysis of the resulting wash waters showed that reducing
sugars, sucrose and FAN were 0.04, 0.14 and 0.01 g per g of bagasse, respectively.
Of the nutrient compounds monitored, sucrose is retained in the highest
Chapter 5 Production of a generic feedstock: solid-state fermentation
97
concentration. This means that a large fraction of fermentable sugars remained in
the sugarcane bagasse that could be fed to the solid state fermentation.
It is clear that non-washed sugarcane bagasse led to a higher sugar and FAN yield
after sequential bioprocessing (Figure 5.4). Data for non-washed process are
almost twice as high as those for washed process. A possible explanation for this
result might be that abundant fermentable sugars were washed from sugarcane
bagasse before inoculation, resulting in insufficient nutrients and poor microbial
growth. These findings may help us to understand the effect of remaining sugars
in the bagasse on our sequential bioprocess. A reasonable approach for further
experiments is to choose non-washed bagasse to ensure enough sugars and
nitrogen for fungal growth.
Figure 5.4 Effect of washing process on nutrients production via solid state fermentation and subsequent hydrolysis
Effect of particle size on sugarcane bagasse Particle size of the substrate is related to substrate characteristics and system
capacity to interchange with microbial growth and heat and mass transfer during
SSF. Smaller particle size could provide larger surface area for microbial growth
and it is also beneficial for heat transfer and exchange of oxygen and carbon
0
50
100
150
200
250
300
0
1
2
3
4
5
6
7
washed non-washed
FAN
(m
g/L)
Red
uci
ng
suga
r (g
/L)
Reducing sugar
FAN
Chapter 5 Production of a generic feedstock: solid-state fermentation
98
dioxide between the air and solid matrix. However, too small particles may lead to
substrate agglomeration, which may interfere with microorganism respiration and
then result in poor cell growth (Pandey et al., 2000; Xin and Geng, 2010).
Three different particle sizes of sugarcane bagasse were carried out following the
method described in section 5.3. Figure 5.5 shows the characteristic development
of T. longibrachiatum under different particle sizes of substrates. The spores are
spread on the surface and inside the solid matrix, and hyphae forms a microscopic
network (mycelium) inside the substrate. Varying the growth condition, similar
patterns of morphological development are observed in Figure 5.5. However, the
smaller particles may result in agglomeration, causing the void space to decrease.
Therefore, the development of fungal growth would be a little different from the
others. As shown in the figure, the aerial hyphae intermeshed on the surface of
the substrate densely.
Figure 5.5 Images of fungi growth on different particle sizes of sugarcane bagasse with soybean hull (2X magnification, USB Microscope)
Chapter 5 Production of a generic feedstock: solid-state fermentation
99
Figure 5.6 Effect of particle size on nutrients production via solid state fermentation and subsequent hydrolysis
Figure 5.6 shows sugar and FAN production from sugarcane bagasse using T.
longibrachiatum under the sequential bioprocessing conditions described. When
sequential bioprocesses were carried out with different particle sizes of bagasse,
particle size of 1.4-0.85 mm supported maximal reducing sugar and FAN (9.4 g/L
and 670.7 mg/L). Generally, the particle size not only significantly affects the water
holding capacity of the substrate, but also influences the diffusion of nutrients and
the exogenous metabolic products to and from the microorganisms. According to
the result reported in section 5.2.2, the porosity of over 1.4 mm particle size (92.9)
is quite similar with the one of 1.4-0.85 mm (91.8). However, larger particle sizes
could present less surface area than smaller one. And it could affect the fungal
growth or enzyme production. In this case, sugarcane bagasse with 1.4-0.85 mm
particle size possibly provided sufficient surface area and aeration to T.
longibrachiatum for growth and enzyme production resulting in increased sugar
production.
0
100
200
300
400
500
600
700
800
0
2
4
6
8
10
12
2-1.4mm 1.4-0.85mm 0.85-0.5mm
FAN
(m
g/L)
Red
uci
ng
suga
r (g
/L)
Reducing sugar
FAN
Chapter 5 Production of a generic feedstock: solid-state fermentation
100
Influence of nitrogen supplement on SSF For fungi of the genus Trichoderma, better results of enzyme production in SSF
have been obtained with organic nitrogen sources, such as peptone, yeast extract
or corn-steep liquid, than with inorganic compounds, such as ammonium sulfate
or ammonium nitrate, as the sole nitrogen (Sun et al., 2010).
In the current project, the sugarcane bagasse contains 62.8% of total
carbohydrates and 0.4% of total nitrogen. The abundant amount of carbohydrates
in bagasse could be desirable for filamentous fungi utilisation. The smaller amount
of total nitrogen, 0.4%, on the other hand could not provide a sufficient source of
nitrogen to promote the fungal growth and facilitate enzyme production. A high
C/N ratio could mean that the nitrogen will be consumed before carbon is utilised.
In contrast, a low C/N ratio may provide excess nitrogen which tends to become
toxic to some microorganisms (Mital, 1997). Therefore, it is necessary to know
what level of C/N ratio could benefit nutrients production through T.
longibrachiatum. The effect of different amounts of nitrogen source on sugar and
FAN production was tested by yeast extract (Table 5.4). All of the experiments
were carried out in an incubator at 30°C for 120h for SSF and then transferred to
50°C for 48h for subsequent hydrolysis.
Carbon to nitrogen ratio (C/N) is a ratio of the mass of carbon to the mass of
nitrogen in a substance. In section 5.5, the carbon content and nitrogen content
in the sugarcane bagasse and yeast extract were based on referred data (Holwerda
et al., 2012; Kruesi et al., 2013).
Chapter 5 Production of a generic feedstock: solid-state fermentation
101
Table 5.4 C/N ratio of mixed-substrates culture medium
Nitrogen added (%, w/w) C/N ratio (g g-1)a
0 107.2
1.5 80.9
4.5 54.3
7.5 40.8
10.5 32.7
15 25.2
a Ratio between carbon in sugarcane bagasse and supplemented nitrogen
It was observed that different amount of yeast extract supplemented has a direct
effect on the growth of T. longibrachiatum on sugarcane bagasse (Figure 5.7).
Fungi did not grow well in the fermentation with nitrogen added below 7.5% (w/w).
It is known that the C/N ratio is one of the most important factors to balance
biomass and products production. The excess or lack of nitrogen content in the
substrate may inhibit fungal growth and is presumably the reason to hinder
enzyme production (Mantovani et al., 2007). In the solid state fermentative
process, even minor variations in the C/N ratios may result in quite distinct
responses from the biological system, since the local concentrations are greatly
superior to those of the submerged fermentation (Bertolin et al., 2003). Figure 5.7
shows that a green surface area appeared in plates where the yeast extract
supplemented is above 7.5% (w/w). In addition, droplets of condensed water were
observed on the internal surface of lids where prolific fungal growth was observed.
Chapter 5 Production of a generic feedstock: solid-state fermentation
102
Figure 5.7 SSF using Trichoderma longibrachiatum on different amount of nitrogen supplement on sugarcane bagasse after 5 days
Figure 5.8 shows that reduction in the dry weight of the substrate is directly
related to the consumption of the nutrients by the extent of fungal growth. Dry
weight loss is increasing in fermentations with nitrogen added contents between
0% and 15% (w/w). The highest dry weight loss (25% of initial dry weight) was
observed in the fermentation carried out using 15% of nitrogen supplement. The
dry weight loss seems be associated with the fungal growth from visual
observation (see Figure 5.7). However, it is worth mentioning that separation of
fungal mycelium from the solid particles of the substrate was practically
impossible, and therefore quantitative analysis of biomass from the remaining dry
fermented substance is very difficult.
Chapter 5 Production of a generic feedstock: solid-state fermentation
103
Figure 5.8 Dry weight loss of 5 days solid state fermentation with different amount of nitrogen supplement
The effect of different amounts of nitrogen added on sugar and FAN production is
presented in Figures 5.9 to 5.11. Enzyme production and substrate utilisation are
associated with fungal growth. Hydrolytic enzymes such as cellulase, xylanase,
beta-glucosidase and protease, which are secreted by the mycelium, diffuse to the
solid matrix and degrade macromolecules into smaller units for taking up
surrounding nutrients. In this study, though, cellulase, xylanase, beta-glucosidase
and protease production were not examined directly, but reducing sugar and FAN,
the products of the hydrolysis reactions, were measured. The results shown in
figures are the average values of duplicate samples.
The reducing sugar released after 120 h fermentation as a consequence of the
hydrolysis of substrate (cellulose and hemicellulose) during solid state
fermentation, provides available carbon source for fungi consumption. Figure 5.9
shows that no yeast extract supplement had the highest reducing sugar yield (35.8
mg/g substrate) after solid-state fermentation. However, substrates with high
sugar content may inhibit the microbial growth during SSF, resulting in low system
productivity.
0
5
10
15
20
25
30
0 1.5 4.5 7.5 10.5 15
Dry
mat
ter
we
igh
t lo
ss
(% o
f in
itai
al d
ry w
eig
ht)
Nitrogen supplemented (%, w/w)
Chapter 5 Production of a generic feedstock: solid-state fermentation
104
Figure 5.9 Effect of different amounts of nitrogen supplement on sugar production via solid state fermentation (SSF) or sequential bioprocessing (SSF + hydrolysis). Error bars indicate ranges between duplicate samples.
Subsequent hydrolysis of whole fermented substrates could provide partially
deconstructed materials and abundant in-situ enzyme complex, which further
reduce operating cost and nourish the hydrolysate with sugar, FAN and trace
elements. Figure 5.9 shows that 10.5% of yeast extract supplement had the
highest reducing sugar yield (120.9 mg/g substrate) after sequential bioprocessing
(SSF + hydrolysis), followed by the 15% nitrogen supplement (89.6 mg/g substrate).
The lowest sugar production was SSF with 4.5% of yeast extract supplement.
There is little apparent difference between the first three trials (0% to 4.5%) about
the effect of nitrogen addition on the sugar production of sequential bioprocessing
from Figure 5.9. To verify this observation, the analysis of variance (ANOVA) was
used to test the basic difference between the groups in this experiment. The
results of reducing sugar released during the hydrolysis of fermented solids
presented four significantly (p<0.05) different groups (a to d) as a function of
nitrogen supplemented (Figure 5.10). The maximum sugar yield was obtained on
the medium with 10.5% of yeast extract added (C/N = 32.7), 85.7 mg/g substrate.
It is interesting to note that sugar production of hydrolysis after SSF with a range
of nitrogen supplemented from 0 to 4.5% (C/N from 107.2 to 54.3) was constant,
0
20
40
60
80
100
120
140
0 1.5 4.5 7.5 10.5 15
Re
du
cin
g su
gar
(mg/
g su
bst
rate
)
Nitrogen supplement (%, w/w)
SSF
SSF + hydrolysis
Chapter 5 Production of a generic feedstock: solid-state fermentation
105
with no significant difference. Clearly, the sugars production and fungal growth
were strongly related to a low C/N ratio that resulted in better productivity and
biomass. The high ratio between carbon and nitrogen content in the medium led
to a low consumption of substrates and a decrease in the sugars production.
Figure 5.10 Average saccharification yield in subsequent hydrolysis (SSF + hydrolysis) with different proportion of nitrogen added. Letters a b c and d represent significantly different (p<0.05) groups of data
The production of FAN after 5 days of fermentation was also investigated (Figure
5.11). The positive effect of increasing nitrogen supplement was observed,
growing from 0.6 mg/g substrate obtained for no nitrogen supplement condition
to 7.5mg/g substrate (15% added yeast extract). Protease activity is associated
with FAN hydrolysis from protein, and It was found that the profile of protease
activity was very similar to that of fungal growth (Wang et al., 2005). From Figure
5.7, it can be clearly seen that FAN production could be related to visual
observation of solid state fermentation.
During the subsequent hydrolysis period, not only the enzymatic hydrolysis
reaction happened, but also the autolysis of the fungal cells occurred under an
oxygen limited condition. Many enzymes such as protease, phosphatase,
0
10
20
30
40
50
60
70
80
90
0 2 4 6 8 10 12 14 16
Suga
r yi
eld
in h
ydro
lysi
s
(mg/
g su
bst
rate
)
Nitrogen supplement (%, w/w)
c
d
a
b
Chapter 5 Production of a generic feedstock: solid-state fermentation
106
glucanase and chitinase are degrading fungal cell walls, cytoplasmic materials and
remaining protein of fermented solids. This process could release intracellular
nutrients (amino acid, peptide, phosphorous and vitamins etc.) and break down
protein remaining in fermented lignocellulose, to the hydrolysate. FAN production
after SSF and hydrolysis is presented in Figure 5.11. In SSF with nitrogen
supplemented from 7.5% to 15% (C/N ratios from 40.8 to 25.2), total FAN
production after sequential bioprocessing (SSF + hydrolysis) were higher than the
others. However, FAN yields in the subsequent hydrolysis in all trials were similar
from 1.31 mg/g to 6.83 mg/g, with no significant (p< 0.05) difference. Results from
this study suggest that the production of sugars and FAN are highly dependent on
C/N ratios.
Figure 5.11 Effect of different amounts of nitrogen supplement on FAN production via solid state fermentation (SSF) or solid state fermentation plus sequential bioprocessing (SSF + hydrolysis)
Effect of mixed substrates on SSF The composition of the cultivation medium is a key parameter in optimising SSF
processes since nutritional factors such as the carbon source, and the levels of
nitrogen and trace minerals, can influence the metabolites enzymes and spore
production. Despite the potential to produce value-added products through solid
state fermentation, supplementing with organic nitrogen sources can be costly. It
0
2
4
6
8
10
12
14
0 1.5 4.5 7.5 10.5 15
FAN
(m
g/g
sub
stra
te)
Nitrogen supplement (%)
SSF
SSF + hydrolysis
Chapter 5 Production of a generic feedstock: solid-state fermentation
107
is not an economical process for chemicals production. In this work, soybean hull
was used as alternative nitrogen supplement for solid state fermentation. Soybean
hulls are by-products from soybean processing, being considered as waste due to
anti-nutritional protein as the low-quality animal feed supplements (Zhang and Hu,
2012). However, with its low content of lignin and sufficient nitrogen content for
microorganisms, soybean hull can be a good choice for medium formulation.
As shown in Table 5.5, sugarcane bagasse (SB) and soybean hull (SH) were used in
varied SB:SH ratios (1:0, 8:2, 6:4, 5:5, 2:8 0:1) for solid state fermentation using T.
longibrachiatum. Carbon to nitrogen ratio is a ratio of the mass of carbon to the
mass of nitrogen in a substance. The carbon content and nitrogen content in the
substrates (sugarcane bagasse and soybean hull) were based on referred data
(Kruesi et al., 2013; Ontario minister of agriculture, 2011).
Table 5.5 Composition of mixed-substrates culture medium
Medium C/N Ratio Composition
A 107.2 100% SB (1:0, SB:SH)
B 63.4 80% SB + 20% SH (8:2, SB:SH)
C 44.9 60% SB + 40% SH (6:4, SB:SH)
D 39.2 50% SB + 50% SH (5:5, SB:SH)
E 28.3 20% SB + 80% SH (2:8, SB:SH)
F 23.8 100% SH (0:1, SB:SH)
The preliminary experiment was incubated at 30°C for 5 days for understanding
the effect of different mixed substrates on SSF at the beginning of the study. From
Figure 5.12, it is noticed that there was nearly no growth of Trichoderma
longibrachiatum (Figure 5.12a). However, when 80% of the sugarcane bagasse
was mixed with 20% of soybean hulls (Figure 5.12b), the growth was slightly
Chapter 5 Production of a generic feedstock: solid-state fermentation
108
improved. The centre area of mixed-substrate could be seen covered by the green
coloured fungus with spreading mycelia in the case of 60% bagasse and 40% hulls
(Figure 5.12c). There was good growth in parts of the plate with 50% bagasse with
50% hulls (Figure 5.12d). When only 20% bagasse was used for the substrate
(Figure 5.12e), abundant growth of mycelia and spores could be seen around the
substrate. The growth then got poorer as the proportion of bagasse to hulls got
smaller. In the last experiment, 100% soybean hulls (Figure 5.12f), careful visual
observation showed that there were some mycelia formed on the substrate, which
means that Trichoderma longibrachiatum did grow on the substrate. Nevertheless,
no spores could be seen in this case.
Figure 5.12 Growth of Trichoderma longibrachiatum on different mixed substrate ratios (a) 1:0, SB:SH (b) 8:2, SB:SH (c) 6:4, SB:SH (d) 5:5, SB:SH (e) 2:8, SB:SH (f) 0:1, SB:SH.
It was observed that some mixed substrates, for example, medium c (6:4, SB:SH)
shrank considerably during the fermentation. Also, it was observed that the
fermented solids inside the petri dish had formed a complete cake through the
extension of the fungal hyphae (Figure 5.13). Due to the extent of growth and cake
Chapter 5 Production of a generic feedstock: solid-state fermentation
109
formation, it was not easy to get representative samples to measure composition
of fermented substrate and enzyme activities from heterogeneous medium,
especially under time course experiment using the same bioreactor.
Figure 5.13 Growth of Trichoderma longibrachiatum on mixed substrates at a ratio of 6:4 (SB:SH)
Cultivation media were consumed for the metabolism of the fungus, and cell
formation during solid state fermentation. Consequently, the highest dry weight
loss should take place in the fermentation with highest cellular metabolic activity.
The dry matter weight loss during each fermentation is presented in Figure 5.14.
Chapter 5 Production of a generic feedstock: solid-state fermentation
110
Figure 5.14 Dry weight loss of the solids after 5 days fermentation with different mixed substrate ratios
As can be seen in Figure 5.14, the highest reduction in the dry weight was observed
in the fermentation with mixed substrate at a ratio of 5:5 (SB:SH). In this study,
ratios of 1:0, 8:2, and 6:4 (SB:SH) show increasing dry weight loss with increasing
soybean hull addition. Thereafter, the reduction of dry weight between 6:4 to 0:1
(SB:SH) was quite similar. The trend of weight reduction is similar to previous
experiments with supplemented nitrogen, which showed increased dry weight
loss with decreasing C/N ratio.
The major goal of implementing solid state fermentation using agricultural wastes,
in this study, was to produce sugars and FAN as a potential generic fermentation
feedstock. In this set of experiments, the effect of mixed substrate ratio on
nutrients production was investigated. Reducing sugars were measured after the
120 h of solid state fermentation and solid state fermentation with subsequent
hydrolysis, respectively. The experiments were carried out with mixed substrate
ratio between 1:0 to 0:1 (SB:SH) and results are presented in Figure 5.15.
0
5
10
15
20
25
30
35
1:0 8:2 6:4 5:5 2:8 0:1
Dry
mat
ter
we
igh
t lo
ss
(% o
f in
itai
al d
ry w
eig
ht)
Substrate ratio (SB:SH)
Chapter 5 Production of a generic feedstock: solid-state fermentation
111
Figure 5.15 Effect of mixed substrates ratio on sugar production via solid state fermentation (SSF) or sequential bioprocessing (SSF + hydrolysis)
As can be seen from Figure 5.15, an equal mixture of substrates 5:5 (SB:SH) ratio
gives the best result after 120 h of solid state fermentation, 202 mg/g of sugar
productivity. Using further hydrolysis at 50°C for 48 h, more sugars were released
from the substrate by enzymes already attached to cellulose and hemicellulose.
The mixed substrate ratio of medium directly affects sugar production after
sequential bioprocessing (SSF + hydrolysis). Sugar yields raised with increasing
soybean hull supplemented up to 50% (w/w) of addition, and decrease thereafter.
The highest sugar production, 197.4 mg/g substrate, was obtained in the
fermentation carried out at mixed ratio of 6:4 (SB:SH).
The carbon to nitrogen (C/N) ratio is always used as a measure of nutrient balance
in the fermentation process. Also, It is known that C/N ratio of substrates is a
critical factor that determines microbial growth, and metabolite formation in solid
state fermentations (Mantovani et al., 2007; Sun et al., 2010). According to Table
5.5, varied mixed-substrate trials can be presented in specific carbon to nitrogen
(C/N) ratios. Figure 5.16 presents five significantly (p<0.05) different groups (a to
e) of sugar yield during further hydrolysis with different C/N ratios. The highest
sugar production, 187.4 mg/g substrate, was obtained in the fermentation carried
0
50
100
150
200
1:0 8:2 6:4 5:5 2:8 0:1
Red
uci
ng
suga
r (m
g/g
sub
stra
te)
Substrate ratio (SB:SH)
SSF
SSF + hydrolysis
Chapter 5 Production of a generic feedstock: solid-state fermentation
112
out at mixed ratio of 6:4 (SB:SH). It is worthwhile to note that in substrate
formulations with C/N ratio ranging from 44.9 to 39.2 (SB: SH from 6:4 to 5:5),
sugar production was similar, with no significant (p<0.05) difference. There was a
low sugar production in the substrate with C/N ratios lower than 39.2. In the
medium with C/N ratio of 23.8 (100% of soybean hull) there was significant
(p<0.05) reduction of sugar yield when compared to the substrate with C/N ratio
of 28.3 (20% of sugarcane bagasse and 80% of soybean hull). The above results
suggest appropriate medium could provide prolific fungal growth and sufficient
enzymes to break down the tough lignocellulosic structure, leading to efficient
subsequent hydrolysis of fermented solids.
Figure 5.16 Average sugar yield in subsequent hydrolysis after 5 days of growth on substrate consisting of mixtures of sugarcane bagasse and soybean hull with different carbon/nitrogen ratios. Different letters represent groups with significant differences (p<0.05)
FAN production during various processes with different mixed substrate ratios was
compared as shown in Figure 5.17: solid state fermentation (SSF) and solid state
fermentation followed by subsequent hydrolysis (SSF + hydrolysis). During the
solid state fermentation, FAN production was low (below 2.5 mg/g substrate) with
soybean hull supplements from 0% to 80%. Interestingly, the FAN yield was
0
20
40
60
80
100
120
140
160
180
200
0 20 40 60 80 100 120
Suga
r yi
eld
in h
ydro
lysi
s (m
g/g
sub
stra
te)
C:N ratios
a
b
c
d
e
Chapter 5 Production of a generic feedstock: solid-state fermentation
113
markedly increased to 7 times using soybean hull only as substrate and reached
the maximum value of 18 mg/g substrate. It is likely that the high protein content
in soybean hull (11.9%) compared with sugarcane bagasse (2.5%), stimulates
higher protease activity production. As a result, a high level of FAN content was
observed in the case of fermentation using soybean hull only.
Figure 5.17 Effect of mixed substrate ratios on FAN production via solid state fermentation (SSF) or sequential bioprocessing (SSF + hydrolysis)
FAN production after sequential bioprocessing (SSF and hydrolysis) is affected
significantly by the ratio of sugarcane bagasse to soybean hull as shown in Figure
5.17. FAN concentration was greatly increased when the soybean hull was added
to the medium. With increased addition of soybean hull to the medium, FAN
reached the highest level (28.2 mg/g) when soybean hull was used alone. As
mentioned before, during the subsequent hydrolysis of fermented solids, not only
the enzymatic hydrolysis reaction happened, but also the autolysis of the fungal
cells occurred under an oxygen limited condition. Two sources to be converted to
FAN after the hydrolysis stage could contribute to (1) remaining substrate protein
and (2) fungal cells. In these experiments, insufficient remaining protein for
0
5
10
15
20
25
30
35
40
1:0 8:2 6:4 5:5 2:8 0:1
FAN
(m
g/g
sub
stra
te)
Substrate ratio (SB:SH)
SSF
SSF + hydrolysis
Chapter 5 Production of a generic feedstock: solid-state fermentation
114
subsequent hydrolysis could be the reason for low FAN content (ranging from 0%
to 50% of soybean hull supplemented), caused by microbial consumption and low
protein provided during solid state fermentation.
Effect of environmental humidity on SSF (in petri dishes)
Smits et al. (1999) explicitly describes the water balance of solid-state
fermentation, which consists of three important processes (1) metabolic water
production by the microorganism (2) evaporation of water (3) diffusion of water
vapour in the fermentation bed (Figure 5.18). Even though simulations with the
model suggested that the diffusion of water was not important for a tray
bioreactor incubated in a 98% relative humidity headspace, it might become
important if the tray were incubated within a drier atmosphere (e.g. lower
humidity). This might be done to promote an evaporative cooling effect, although
such a strategy would quickly lead to drying out of the fermentation bed (Mitchell
et al., 2003; Smits et al., 1999). Evaporation effect is a major concern in the
moisture control of solid-state fermentation, especially in bioreactors. It is
important to keep consistency of operating factors such as moisture content from
laboratory scale to large scale production. However, far too little attention has
been paid to environmental humidity influence (humidity of incubator) on solid-
state fermentation using petri dishes.
Chapter 5 Production of a generic feedstock: solid-state fermentation
115
Figure 5.18 Water balance in a solid state fermentation using petri dish system
Evaporation is the process by which liquid is vaporised from a surface into a
gaseous phase that is not saturated with the evaporating substance. Mitchell et al.
(2006) mentioned that the degree of evaporation depends on the saturation of
the air. If the air entering the bed is not saturated with water, it will promote
evaporation and dry out the bed, eventually decreasing the water activity to
values unfavourable for growth. Saturated air should therefore be supplied for air
inlet, so that the evaporation effect can be minimised. The same phenomenon
could be happening in the small-scale fermentation system (petri dish).
Petri dishes (9 cm diameter) containing 5 g of non-fermented mixed solid medium
substrates (6:4, SB:SH) were tested with different environmental relative humidity
(35% and 75%) at 30°C. The initial moisture content of solids was 65% (w/w, db).
All samples were put in the petri dishes and covered with lids without inoculating.
Using the moisture ratio to generalise the change of moisture content is the most
common method in the drying trial. Results for the test are shown in Figure 5.19.
Clearly, with a low environmental humidity (35%), the evaporation rate from the
non-fermented solids was higher than when the surroundings were more humid.
Chapter 5 Production of a generic feedstock: solid-state fermentation
116
According to the results obtained, the effect of environmental humidity on solids
in the petri dish did cause significant difference in water evaporation. The
moisture ratio observed for the non-fermented substrate under 35% of
environmental relative humidity after 7 days was around 0.29, while for the higher
relative humidity (75%) during the same period a value of around 0.71 was
reached. This is due to the higher moisture difference between the outside
environment and the head space within the petri dish, which provides the driving
force for the evaporation.
Figure 5.19 The moisture ratio of non-fermented solids under different
environmental relative humidity (35 and 75%)
In order to understand the effect of relative humidity on growth, the different
levels were tested (35% and 75%, relative humidity) with inoculated substrates.
The relative humidity of incubator can be regulated by placing an open container
filled with sterilised water at the bottom of the incubator. The environmental
relative humidity could achieved and maintained 75% within 2 hours in the
incubator at 30°C. On the contrary, the relative humidity of incubator was only
35% without providing sterilised water in the container. To make sure the
consistency of the relative humidity level, the portable humidity sensor was used
to monitor the variation during the solid-state fermentation.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 1 2 3 4 5 6 7
Mo
istu
re r
atio
Time (day)
35% 75%
Chapter 5 Production of a generic feedstock: solid-state fermentation
117
The required amount of spore suspension was poured into disposable Petri dishes
(9 cm diameter) and the solid fermentation was started by adding 5 g of the
sterilized mixed solid substrate (6:4, SB:SH). The pH was not controlled and the
plates were incubated under static conditions at 30°C. Plates from different
cultivation conditions were withdrawn for analysis at the desired intervals. The
initial moisture content of all substrates was adjusted to 65% (w/w). Results are
shown in Figure 5.20.
Figure 5.20 The moisture ratio of solid state fermentation under different environmental relative humidity (35 and 75%)
The decrease of moisture ratio for both conditions is quite similar during the first
2 days of incubation, where there appears to be a balance between water
production (via metabolism) and water loss (via evaporation). However, the
moisture ratio reduction for the low humidity case (35%) after 3 days is much
faster than that for the higher one (75%). The moisture ratio experimentally
obtained for the incubator with high humidity was still around 0.93 after 7 days,
while for the incubator with low humidity the value had fallen to around 0.47.
Clearly, in the high humidity case, there is a good balance between production and
loss. As in the non-fermented substrates test, Inspection of Figure 5.20 shows a
0
0.2
0.4
0.6
0.8
1
1.2
0 1 2 3 4 5 6 7
Mo
istu
re r
atio
(%
)
Time (day)
35% 75%
Chapter 5 Production of a generic feedstock: solid-state fermentation
118
significant difference between high and low humidity environments for solid state
fermentation using petri dishes.
The values of effective diffusivities (Deff) of fermented and non-fermented
substrates for the two humidity environments are presented in Table 5.6. As
expected, the values of Deff increased with increasing moisture difference between
environment and system (petri dish). It shows that the non-fermented substrate
of environment with low humidity affords higher values of Deff, about 2.87 times,
than that with high humidity. Furthermore, conducting solid state fermentation at
the same condition for moisture ratio resulted in a raise of about 15 times in Deff
values. The above results mean that the metabolic water production during
fermentation leads to effective diffusivities for the fermented solids that are
apparently lower than those for non-fermented solids.
Table 5.6 Values of effective diffusivities obtained with different environmental humidities on solid state fermentation
Sample Effective Diffusivity (m2s-1)
Non-fermented, 35% 1.278 x 10-11
Non-fermented, 75% 4.45 x 10-12
Fermented, 35% 9.02 x 10-12
Fermented, 75% 5.86 x 10-13
For mass and heat transport during heating process, external and internal
transport phenomena can be distinguished. External transport occurs from the
particle surface to the surrounding air and internal transport from the inner to the
outer layer of the particle. The difference of water concentration between the gas
and the solid phase, and the bulk of the air is the driving force for mass transfer
(Sun, 2007). When the water concentration gradient is increased in the gas phase
between system (petri dish) and environment (incubator), the moisture content
of substrate can reduce dramatically due to evaporation of the existing water in
the solids through metabolic heat evolution. It is worthwhile to note that the use
of high relative humidity does not prevent evaporation from occurring within the
Chapter 5 Production of a generic feedstock: solid-state fermentation
119
bed, but it does minimize evaporation compared to the use of unsaturated air. The
reason could be attributed to the metabolic water production during solid state
fermentation, which can maintain the desired level of moisture content when the
minor evaporation effect occurs under small water concentration difference
between system and environment.
As can be seen in Figure 5.21a, which shows petri dishes from above and below,
the fungi development during 3 days of fermentation with high relative humidity
(75%) occurs on both top and bottom surfaces. In addition, droplets of condensed
water were observed on the internal surface of the lid. The low concentration
gradient between environment (incubator) and system (petri dish) could leave
abundant water vapour inside the petri dish, leading to condensation when it
reached the lid. However, in the other experiment with lower relative humidity
(35%) lower growth was observed, with very poor growth on the bottom surface
(Figure 5.21b).
Chapter 5 Production of a generic feedstock: solid-state fermentation
120
Figure 5.21 Effect of the relative humidity on SSF after 3 days (a) 75% (b) 35% showing top (left) and bottom (right) views
After 5 days of fermentation it could be seen that fungal growth favoured the high
relative humidity environment (Figure 5.22). It was noticed that the entire surface
of the substrate was covered with spores and mycelium in the fermentation with
higher relative humidity (75%). In contrast, the lower relative humidity case (35%)
showed much less growth on the same mixed substrate. Despite having been
inoculated over entire petri dish, Figure 5.22b shows two quite distinct areas on
the surface of the substrate; a green surface covered by the prolific fungal growth
in the centre of the substrate (area 1); and poor development in the outer region
of the medium (area 2). The moisture content of the centre area was measured at
56%, whereas the outer region was only 19%. The moisture content of substrate
in the centre was therefore almost 3 times greater than that found for the outer
Chapter 5 Production of a generic feedstock: solid-state fermentation
121
region. This visual observation supports the phenomenon observed in Figure 5.20,
where the high moisture gradient between petri dish and incubator led to
increased moisture content reduction of substrate during solid state fermentation.
Figure 5.22 Effect of the relative humidity on SSF after 5 days (a) 75% (b) 35%
Further hydrolysis of fermented solids should be also investigated to understand
the effect of humidity level on sequential bioprocessing (SSF and in-situ enzyme
hydrolysis). Figure 5.23 indicates that the sugar yield was enhanced by increasing
environmental humidity level from 35% to 75%, and maximum reducing sugar
production yield of 263.1 mg/g substrate was obtained when the relative humidity
level was 75%. It is acknowledged that the water activity (aw) of the substrate is a
key factor affecting microbial activity and enzyme production, an optimal moisture
level has to be maintained during solid state fermentation (Molaverdi et al., 2013).
Strong evaporation effect and mass transfer work synergistically to promote water
vapour diffusion from substrate, leading to low moisture level of the substrate
(Figure 5.22b) to reduce fungal growth, enzyme activity and substrate
deconstruction. However, there is no significant difference between low humidity
level and high humidity level on FAN production, 13.4 mg/g substrate and 13.8
mg/g substrate, respectively. This is probably due to the same limited nitrogen
source supplied (SB:SH, 6:4) of the substrate for sequential bioprocessing.
Chapter 5 Production of a generic feedstock: solid-state fermentation
122
Figure 5.23 Effect of environmental humidity level on sugar and FAN production via sequential bioprocessing (SSF + hydrolysis)
In summary, relative humidity is one of the most important aspects of fungi
growth in petri dishes as it has a direct influence on evaporation. A low relative
humidity increases evaporation rate from the moist fermented medium.
Depending on the process, the results obtained in a preliminary test indicated the
necessity of a better control of the medium moisture content by maintaining a
high humidity content of the incubation environment. This measure could provide
similar and more reliable results for comparison with larger scale experiments.
Effect of incubation time on sequential enzyme hydrolysis
During solid state fermentation, enzyme activities and deconstruction level of
substrate are affected by the growth of the microorganism. Therefore, it is crucial
to monitor nutrients production to find the optimal time to terminate the
fermentation to have a proper enzyme complex and accessible substrate for
0
2
4
6
8
10
12
14
16
18
0
50
100
150
200
250
300
35% 75%
FAN
(m
g/g
sub
stra
te)
Re
du
cin
g su
gar
(mg/
g su
bst
rate
)
The relative humidity of Incubator
Reducing sugar
FAN
Chapter 5 Production of a generic feedstock: solid-state fermentation
123
further hydrolysis. Based on the outcomes of the pervious experiments, the size
of sugarcane bagasse was adjusted to 1.4-0.85 mm and mixed with soybean hull
in the ratio of 6:4 (SB:SH). Initial moisture content of substrate was set to 65%
(w/w), the relative humidity in the incubator was fixed to 75% and the
fermentations were carried out at 30°C. The time interval chosen for sampling was
every 24 hours.
After 1 day of solid state fermentation, further hydrolysis yielded only 5.6 mg
reducing sugar/g solid and 3.9 mg FAN/g solid (Figure 5.24). After 5 days of
incubation, the reducing sugars and FAN, after subsequent hydrolysis, were 226.3
mg/g and 7.7 mg/g respectively. Beyond 5 days of solid state fermentation, sugar
yield started to decrease and further incubation times were markedly affected. It
is believed that sugar production was considered to be associated with cellulase,
beta-glucosidase and hemicellulase activity that degraded cellulose and
hemicellulose remaining after SSF. The decreasing pattern, however, suggests that
the synergistic effect of lignocellulosic enzymes might reach highest level during 5
days of fermentation and thus hit the highest point of further hydrolysis
performance. As for FAN production, it was a result of the activity of both
extracellular protease released during the fermentation and intracellular protease
from autolysis (Wang et al., 2010). FAN production rose gradually until 5 days of
SSF and then reached a plateau around 7.8 mg/g substrate. This is probably due
to the limited nitrogen content of the substrate and the inhibition of enzymes by
the product.
Chapter 5 Production of a generic feedstock: solid-state fermentation
124
Figure 5.24 Effect of SSF incubation time on Sugar and FAN production via sequential bioprocessing (SSF + hydrolysis)
A further experiment was carried out to show the celluloytic enzyme activities
produced from Trichoderma longibrachiatum in SSF system (Table 5.7). Since
sequential bioprocessing development is our priority, further investigation in the
coming chapters focuses on the generic feedstock production rather than
optimisation of enzyme production.
Table 5.7 Celluloytic enzyme activities produced by T. longibrachiatum
Sources of enzymes Enzyme activities1
Exoglucanase Beta-glucosidase Xylanase
FPU/g U/g U/g
DEPOL 686L 24.9 144.5 382.4
1 enzymes production from Trichoderma longibrachiatum under 5 days of solid-state fermentation using mixed-substrates (SB:SH, 6:4)
0
1
2
3
4
5
6
7
8
9
10
0
50
100
150
200
250
300
1 2 3 4 5 6 7
FAN
(m
g/g
sub
stra
te)
Re
du
cin
g su
gar
(mg/
g su
bst
rate
)
Incubation time of SSF (day)
Reducing sugar
FAN
Chapter 5 Production of a generic feedstock: solid-state fermentation
125
Summary
The main objectives of the experiments reported in this chapter were to check the
feasibility of sugarcane bagasse and soybean hull for the growth of Trichoderma
longibrachiatum and subsequent hydrolysis. The influence of operational factors
for solid state fermentation and environmental variables on this system for sugar
and FAN production were also investigated. Preliminary studies of Trichoderma
longibrachiatum in mixed-substrates demonstrated its capability to produce a
generic microbial feedstock. This indicates that sequential bioprocessing could be
a competitive choice for microbial culture medium production (sugar, FAN and
other minerals) and a suitable alternative to traditional sugar production from
lignocellulose.
Chapter 6 Production of a generic feedstock: subsequent hydrolysis
126
CHAPTER 6
Production of a generic feedstock:
subsequent hydrolysis
Introduction Enzymatic hydrolysis efficiency has a great influence on the overall lignocellulose
bioconversion process and thus has been recognised as one of the bottlenecks in
lignocellulose biorefinery development during the past two decades. Besides the
technical challenge of enzyme hydrolysis, the cost of enzyme is another issue to
be solved. It is evident from the literature that enzymatic hydrolysis accounts for
up to around 66% of the cost on lignocellulose-based bioethanol (Du et al., 2014).
Currently, no report on direct utilisation of raw lignocellulose for microbial
feedstock using solid state fermentation and subsequent in-situ enzyme hydrolysis
with fungal autolysis is available. In chapter 5, the technical feasibility of utilising
the sugarcane bagasse and soybean hull as raw material for generic microbial
feedstock has been demonstrated through preliminary trials. The focus of this
chapter is the investigation of operation factors on the subsequent hydrolysis with
fungal autolysis of fermented solids. The effect of solid loading and reaction
conditions (temperature, pH, as well as inhibitor addition) on hydrolysis was
closely studied and enzyme activities were explored under different pH and
temperature. Additionally, specific attention was given to fungal autolysis in the
submerged cultivation in order to verify that fungal autolysis occurred in the
oxygen-limited environment.
Chapter 6 Production of a generic feedstock: subsequent hydrolysis
127
Solid to liquid ratio effect Unlike starch hydrolysis, being an insoluble and highly heterogeneous substrate,
lignocellulosic materials pose several challenges in conversion to fermentable
sugars (Kristensen et al., 2009). Increasing a solid loading in hydrolysis is one of
the most important challenges to make biofuels production more economical due
to the reduction of water usage and ethanol distillation cost (López-Linares et al.,
2014; Modenbach and Nokes, 2013). In most of the lignocellulose-based ethanol
production, the final ethanol titer should be above 4% (v/v), which is the minimum
for economic feasibility (Huang et al., 2011; Jørgensen et al., 2007b; Wingren et
al., 2003). To reach an ethanol concentration higher than 4%, a promising
approach is to increase substrate concentrations up to 20% (w/w) during enzyme
hydrolysis. However, High solids concentrations would lead to larger levels of end-
product inhibition (Andrić et al., 2010), limitation of enzyme diffusion (Lee and Fan,
1982; Wang et al., 2011) and viscosity increase, as well as stirring and mixing
limitations (López-Linares et al., 2014). To reduce these negative effects and
facilitate lignocellulose conversion, different strategies such as fed-batch mode or
diverse pretreatment process have been proposed for subsequent hydrolysis at
high substrate concentrations.
In this study, fermented solids need to be prepared for further hydrolysis. The 10
g of mixed substrates (6:4, sugarcane bagasse: soybean hull) for each solid state
fermentation were prepared. The substrate and mineral salt solution were then
autoclaved at 121 ° C for 20 min. After the substrates had cooled to room
temperature, they were moistened with the mineral salt solution, subsequently
inoculated with T. longibrachiatum spore suspension and mixed with the substrate
thoroughly to reach a concentration of 1x106 spores/g substrate and final
moisture of 65% (g water per 100 g wet substrate). The solid state fermentations
using petri dishes under 75% of environmental humidity condition were carried
out for 5 days. After 5 days of fermentation, enzymatic hydrolysis and fungal
autolysis were carried out at pH 4.8 using 0.05 M sodium citrate buffer and 50°C
in an incubator shaker at 160 rpm for 48 h under oxygen-limited condition. The
Chapter 6 Production of a generic feedstock: subsequent hydrolysis
128
experiments were carried out in 500 mL Duran bottles containing a total dry mass
of 10 g fermented solids. Considering the type of hydrolysis step used in our study,
approximately 12-15% of insoluble solids represent the upper limit at which
lignocellulose can be mixed effectively in stirred-tank reactors (Hodge et al., 2009).
Therefore, different fermented solids loading at 2%, 4%, 6%, 8%, 10% and 12%
(w/w) were investigated in the further hydrolysis experiments after solid state
fermentation. All experiments were conducted in duplicate.
As shown in Figure 6.1, the amount of free liquid water decreased as the solid
loading increased (2 to 12%, w/w). Further increase of the bagasse concentration
was attempted, however, the bagasse powder is so bulky and puffy that complete
wetting in solution could not be achieved. Thus, 12% was considered as the highest
loading ratio for the hydrolysis experiment at lab scale. The material with 12%
loading initially appeared as only slightly moist and very little free water could be
observed in the bottle. The materials with lower solid loading (2 to 8%, w/w) were
transformed into a paste-like matter after around 20 h. However, materials with
initial 12% (w/w) were still dense and had a clay-like appearance. It is well known
that end-product inhibition plays an important role in enzymatic hydrolysis and
particularly cellulase, beta-glucosidase and cellobiohydrolase (Xiao et al., 2004).
Figure 6.1 Post-fermentation hydrolysis of bagasse at different solid loadings (2, 4, 6, 8, 10 and 12 g) into 100 mL citric buffer solution
Chapter 6 Production of a generic feedstock: subsequent hydrolysis
129
During the hydrolysis step, the yield of sugar production gradually decreased (33
to 11%) as the solids concentration increased, the sugar yield was reduced to
almost one third at 12% (w/w) of solid loading compared with that at 2% (Figure
6.2). On the other hand, sugar concentration increased (6.5 to 13.14 g/L) as solid
loading decreased. For free amino nitrogen (FAN) production, both yield and
concentration decreased with increasing solid loading (Figure 6.3).
The process used in this experiment is quite distinct from previous lignocellulose
hydrolysis studies, though they have quite similar hydrolysis kinetics. In this
research, instead of additional commercial or crude enzymes, the in-situ enzymes
were transferred with fermented solids and used for further hydrolysis.
Furthermore, secreted enzymes have probably already attached to the surface of
the substrate (cellulose, hemicellulose) as the mycelium penetrated into the
bundle of lignocellulosic fibres during the solid state fermentation. Once
fermented, these are transferred directly to the hydrolysis step where, unlike with
the use commercial enzymes, they are already absorbed into the substrate.
Figure 6.2 Effect of substrate concentration on sugar production in subsequent hydrolysis of fermented bagasse/SBH solids
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
3
5
7
9
11
13
15
0.02 0.04 0.06 0.08 0.1 0.12
Yiel
d (
g/g
pre
trea
ted
su
bst
rate
s)
Red
uci
ng
suga
r (g
/L)
Solid loading ratio in enzyme hydrolysis (g/g)
Reducing sugar
Yield
Chapter 6 Production of a generic feedstock: subsequent hydrolysis
130
Figure 6.3 Effect of substrate concentration on FAN production in subsequent hydrolysis of fermented bagasse/SBH solids
The results confirm that the in-situ enzymes enable saccharification of
lignocellulosic materials at 12% (w/w) of solid concentration, and probably even
higher. However, the general trend appears to be a significant decrease in
conversion yield of sugar with increasing solid content. The reduced efficiency at
higher solid loadings appears to be a general effect as previous studies
investigating the relationship between solids concentration and conversions
showed a similar effect. Since the problem of enzyme diffusion is minimised in the
process proposed in this thesis, part of the reason for the decreased yield might
be end-product inhibition, which intensifies as the substrate concentration
increases (Jørgensen et al., 2007b). Unlike sugar production, the decreasing FAN
concentration with increasing loading could be attributed to the ineffective mixing,
which is a crucial factor influencing the hydrolysis efficiency. Further and more
specific testing in the stirred reactor would be required to clarify the limitation of
solid loading in subsequent hydrolysis. Because different types of analysis (sugars,
FAN, moisture content and composition analysis) need to be carried out within
limited raw materials (sugarcane bagasse), the 4% solids loading was chosen
instead of the higher solids loading in the next experiments.
0
0.002
0.004
0.006
0.008
0.01
0.012
0.014
100
150
200
250
300
0.02 0.04 0.06 0.08 0.1 0.12
Yiel
d (
g/g
pre
trea
ted
su
bst
rate
)
FAN
(m
g/L)
Solid loading ratio in enzyme hydrolysis (g/g)
FAN
Yield
Chapter 6 Production of a generic feedstock: subsequent hydrolysis
131
Temperature and pH effect A lot of reported studies on enzymatic hydrolysis of cellulose or lignocelluloses
using commercial Trichoderma reesei cellulase were conducted at pH 4.8 with
temperature near 50°C, and it was found that activity would be maximised within
a pH range between 4.8 to 5.0 and decrease out of this range (Lan et al., 2013).
However, the optimal range could be slightly different between varied
lignocellulose and pure cellulose. One of the reasons for the above difference is
the amount of existing lignin in lignocellulose to show a surface hydrophobic
property and cause nonspecific binding to the cellulase or other enzymes
(Mansfield et al., 1999). The pH value can influence substrate surface charge
through surface functional groups to change hydrophobicity of the surface. Also,
the pH-induced lignin surface charge may also affect electrostatic interactions
between enzymes and lignin. This surface charge variation can lead to various
effects on hydrophobic/hydrophilic and perhaps electrostatic interactions
between cellulase and lignin in reducing cellulase non-productive binding to lignin
and therefore enhancing saccharification among different substrates. The enzyme
activity could be adversely affected by temperature through protein denaturation
and adsorption (Lan et al., 2013; Seiji Nakagame et al., 2011). Moreover, crude
multi-enzyme solution from the specific microorganism could present different
behaviour rather than single enzyme shown in previous research. To explore the
effect of pH and temperature on enzyme hydrolysis using in-situ multi-enzymes
could provide further knowledge to optimise the process design.
The fermented solids for further hydrolysis were prepared following the
procedure in section 5.8. The objective of the present work was to investigate the
effect of pH (4-8.5) and temperature (40-60°C) on enzymatic hydrolysis using the
in-situ enzymes with whole fermented media of Trichoderma longibrachiatum
cultivated on substrates.
The optimum temperature and pH for subsequent hydrolysis and autolysis were
determined by conducting experiments at 40, 50, 60°C from pH 4 to pH 8.5.
Chapter 6 Production of a generic feedstock: subsequent hydrolysis
132
Reaction temperature was controlled at three levels in order to investigate the
possible growth of contaminants and fungal growth. The maximum sugar
concentration of hydrolysis (9.1 g/L) was achieved at 50°C, pH 6 (Figure 6.4). The
large yield at 50°C, pH 6 could be attributed to the fact that this temperature and
pH combination is close to the optimum one for these multi-enzymes (e.g.
cellulase, xylanase and beta-glucosidase). Overall, the enzymatic hydrolysis below
60°C could achieve better sugar production in the experiments. By comparing the
results presented in Figure 6.4 and Figure 6.5 for the experiments measured for
FAN concentration, the profile is slightly different. The FAN production increases
gradually from pH 4 to pH 6 at different temperature levels except in the 50°C trial.
It can be observed that the FAN production hit the peak at a pH value of pH 7 then
declined at pH 8.5. These results differ from previous study that showed the pH
effect (3-6.5) did not considerably affect the degree of fungal autolysis
(Koutinas et al., 2005). A possible explanation for this might be that the specific
characteristics were of multi-enzymes presented from a different fungus (i.e. T.
longibrachiatum). The fungal mycelium autolysis and hydrolysis requires the
synergism of many enzymes because of the complexities of cell walls and
substrates. Each of the enzymes secreted from different microorganisms catalyses
a specific reaction in a particular pH range. Another possible reason for this is that
the substrate used for hydrolysis and autolysis could affect the efficiency of
enzyme-binding in varied pH environments (Lan et al., 2013).
Chapter 6 Production of a generic feedstock: subsequent hydrolysis
133
Figure 6.4 Effect of temperature and pH on sugar production during 48h further hydrolysis after solid state fermentation
Figure 6.5 Effect of temperature and pH on FAN production during further hydrolysis after solid state fermentation
Growth of contaminants never occurred in any experiments implemented at
temperatures above 50°C. However, fungal re-growth could happen at 40°C during
enzyme hydrolysis and fungal autolysis under oxygen-limited conditions. Figure
6.6 shows that at 40°C some white specks formed in the hydrolysate after 48 h.
0
1
2
3
4
5
6
7
8
9
10
3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9
Red
uci
ng
suga
r (g
/L)
Reaction pH
40℃
50℃
60℃
0
100
200
300
400
500
600
3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9
FAN
(m
g/L)
Reaction pH
40℃
50℃
60℃
Chapter 6 Production of a generic feedstock: subsequent hydrolysis
134
This may be attributed to insufficient environment stress during the fungal
autolysis, so the fungi kept growing and formed mycelia in the nutrient solution.
Many fungi can grow in submerged fermentation in different forms ranging from
dispersed filaments to pellets, depending on culture conditions and the strain of
organism used. Different pellets could be observed from loosely packed hyphae,
forming ‘‘fluffy’’ pellets, to tightly packed, compact, dense pellets (Ferreira et al.,
2009; Papagianni, 2004).
Figure 6.6 possible fungal growth during enzymatic hydrolysis and autolysis at 40°C, pH 6
The effect of microbial inhibitor on further hydrolysis
The implementation of enzyme hydrolysis with fungal autolysis above 50°C under
oxygen-limited environment is very crucial in the process developed during this
project. Not only did it accelerate enzyme activities to produce sugars, nitrogen
and other nutrients, it also inhibited fungal growth and destroyed the mycelium
through the action of its own enzymes (autolysis). However, because no
sterilisation was carried out when the fermented solids were transferred to the
further hydrolysis steps, there is a risk of contamination in the subsequent
hydrolysis under particular conditions. Numerous studies have used sodium azide
Chapter 6 Production of a generic feedstock: subsequent hydrolysis
135
to prevent microbial growth during enzyme hydrolysis of lignocellulose (Kuo and
Lee, 2009; Mussatto et al., 2008). Because the influence of microbial inhibitor on
further hydrolysis is unknown, especially in the oxygen-limited condition, it is
necessary to test microbial addition effect on further hydrolysis. In the present
study, the effect of microbial inhibitor addition was investigated under the same
hydrolysis procedure as mention in section 6.1. The solid loading was 4% (w/w) at
50°C, 160 rpm under oxygen-limited condition in citric buffers at two pH values
(4.8 and 6) and sterilised tap water (6.5), respectively.
Figure 6.7 and 6.8 show that the sugar and FAN hydrolysis profile varied under the
different conditions used. It is immediately apparent that the overall sugar
production is more or less the same between the hydrolysis with sodium azide
addition and those without addition (Figure 6.7). The sugar concentration rose
steadily without sodium azide addition from pH 4.8 to pH 6, and achieved highest
while using sterilised water. The above results demonstrate that the further
hydrolysis at 50°C under oxygen-limited condition could suppress microbial
activity effectively.
Figure 6.7 Effect of microbial inhibitor on sugar production during further hydrolysis after solid state fermentation
0
1
2
3
4
5
6
7
8
9
pH 4.8 pH 6 Sterilised water (pH 6.5)
Re
du
cin
g su
gar
(g/L
)
without addition
sodium azide addition
Chapter 6 Production of a generic feedstock: subsequent hydrolysis
136
Figure 6.8 Effect of microbial inhibitor on FAN production during further hydrolysis after solid state fermentation
The influence of sodium azide addition on FAN production after 48 h of further
hydrolysis is shown in Figure 6.8. In general, subsequent hydrolysis with the
addition of sodium azide were higher than without sodium azide. Both conditions
could achieve the highest FAN concentration in pH 6. It is interesting to note that
sodium azide is one of the non-specific autolysis inducers to inhibit cytochrome
oxidase and ATPase to trigger autolysis reaction (Chung et al., 2009). This could be
the reason that hydrolysis with sodium azide could reach higher FAN
concentration in some specific conditions but more or less the same sugar
concentration compared to the trials without sodium azide addition. However,
sodium azide may affect some microorganisms such as Saccharomyces cerevisiae
and Candida albicans (Fales, 1953; Rikhvanov et al., 2002). This effect must be
considered when the hydrolysate is used for subsequent fermentations. The
results shown in this section indicate the possibility of sterilised tap water used in
hydrolysis to substitute citric buffer to reduce production cost.
0
50
100
150
200
250
300
350
400
450
pH 4.8 pH 6 Sterilised water (pH 6.5)
FAN
(m
g/L)
without addition
sodium azide addition
Chapter 6 Production of a generic feedstock: subsequent hydrolysis
137
Fungal autolysis Fungal autolysis is the natural process of self-digestion of aged hyphal cultures,
happening as a result of hydrolyase activity, leading to disruption of organelles and
cell wall structure to release cytoplasmic proteins and vacuolation (White et al.,
2002). Numerous previous studies have been carried out on the susceptibility to
autolysis and lysis of fungal cell walls. For example. In many bioprocesses fungi
may be exposed to either nutrient limitation or physical stress leading to
morphological and/or physiological changes (Lahoz et al., 1986; Pollack et al.,
2008). Environmental stress induced changes may cause damage in some
bioprocesses such as penicillin production (Harvey et al., 1998). But for some
processes, the autolysis could significantly benefit outputs through rational
process integration. A series of studies in the SCGPE has demonstrated that fungal
autolysis can yield beneficial results, for example, nutrient supplements
production from fermented residues (Dorado et al., 2009; Du et al., 2008a;
Koutinas et al., 2005; Wang et al., 2010).
Microscopic observations showed that cell walls were difficult to solubilise
completely and break down at temperatures between 50 and 60°C. There are
similarities between the results presented in the present study and those
described by Koutinas et al. (2005). Figure 6.9 shows the degradation of released
cytoplasm and some of them can be observed around the hyphae. Overall, the
damage to cell wall structure was limited because autolysis could continue for over
30 days once the environmental stress was imposed (Lahoz et al., 1986). The
reported results about the resistance of fungal cell walls could be consistent with
the observation of partial degradation in the first two days of the autolysis process.
Chapter 6 Production of a generic feedstock: subsequent hydrolysis
138
Figure 6.9 Cytoplasm degradation during the autolysis/hyrolysis of fermented solids at 50°C, pH 4.8
In this study, solid state fermented solids were used as substrate for further
hydrolysis to produce generic microbial feedstock. However, it is not easy to
confirm that the autolysis occurred from data since cellulose and hemicellulose
were hydrolysed by in-situ enzymes simultaneously to increase sugar and FAN
concentration. Therefore, a fungal autolysis test should be examined by the simple
sugars used to rule out the contribution of enzymatic hydrolysis. In this
experiment, Submerged fermentation of T. longibrachiatum (10 g/L glucose, 5 g/L
yeast extract, pH 6 citric buffer) was carried out at 30°C, 160 rpm for 120 h. At the
end of fermentation, the whole medium was transferred to the Duran bottle and
the temperature increased to 50°C, 160 rpm for 48 h, causing fungal autolysis
under environmental stress (high temperature and oxygen-limited condition).
The fungal autolysis trial was tested in an oxygen-limited environment using the
remaining glucose of a submerged fermentation as carbon source. Under
Chapter 6 Production of a generic feedstock: subsequent hydrolysis
139
environmental stress condition reducing sugar (glucose) remained unchanged
after 48 h of autolysis (Table 6.1). Meanwhile, FAN concentration increased from
4.7 mg/L up to 13.2 mg/L at the end of reaction. The fungal autolysis was increased
from the results obtained in the early solid state experiment (Figure 6-9). A direct
comparison between the microscopic study of fungal autolysis and nutrient
measurements confirmed that the environmental stress does limit fungal activity
during further hydrolysis.
Table 6.1 Broth composition after 48 h of fungal autolysis
Reducing sugar (g/L) FAN (mg/L)
0 h 1.93 4.7
48 h 1.94 13.2
Kinetics of further hydrolysis The hydrolysis kinetics of fermented substrates with in-situ enzyme complex was
studied. When a suspension of 4% (w/w) fermented solids was transferred at 50°C,
pH 4.8 citric buffer and the production of reducing sugars, FAN (free amino
nitrogen) and IP (inorganic phosphorous) was initiated (Figure 6.10). At the
beginning of further hydrolysis, the mixture contained 25 mg reducing sugar/g
solid, 1.03 mg FAN/g solid and 2.1 mg IP/g solid. At the end of the experiment,
reducing sugars, FAN and IP were 112.48 mg/g, 7.51 mg/g and 9.61 mg/g
respectively. Figure 6.10 shows the production of these nutrient components
during the further hydrolysis and fungal autolysis. The concentration of reducing
sugar rose sharply during the first 10 h then reached the plateau until the end of
reaction. This was probably due to the product inhibition and the enzyme activity
decrease with time, leading to a decrease in hydrolysis efficiency. As for FAN and
IP production, it was a result of protease enzyme and autolytic enzyme activity
that digested fungal cell walls and little nitrogen remained after solid state
fermentation (Koutinas et al., 2005; Wang et al., 2010).
Chapter 6 Production of a generic feedstock: subsequent hydrolysis
140
Figure 6.10 Profiles of further hydrolysis of the fermented solids at 50°C. The solid lines are generated by equation 6-1 to predict the production of reducing sugar, FAN and IP
Simultaneous hydrolysis and fungal autolysis involve many undefined enzymatic
hydrolysis reactions. The deconstructed cellulose, hemicellulose and remaining
protein from fermented solids were hydrolysed by various in-situ enzymes
secreted during the solid state fermentation. Meanwhile, disintegration of cellular
components occurred in the cytoplasm and cell wall (McIntyre et al., 2000). In
order to predict the production of reducing sugar, FAN and IP, an equation
proposed by Koutinas et al. (2003) was used (Equation 6-1).
𝑋𝑖 = 𝑋𝑖∞(1 − 𝑒−𝑘𝑖𝑡) Equation 6-1
Where 𝑋𝑖is the instantaneous concentration of component i (mg/g solid); 𝑋𝑖∞is
the equilibrium concentration of component i (mg/g solid); ki is the first-order
reaction constant (h-1); and t is the reaction time (h).
The application of nonlinear regression software (Origin 8.0) was used to fit
equation 6-1 to the reducing sugar, FAN and IP production profiles of further
hydrolysis at 50°C and generated the empirical constants and regression
coefficients (Table 6.2).
0
1
2
3
4
5
6
7
8
9
0
10
20
30
40
50
60
70
80
90
100
0 10 20 30 40 50
FAN
an
d IP
pro
ud
ctio
n (
mg/
g, d
b)
Tota
l red
uci
ng
suga
rs p
rou
dct
ion
(m
g/g,
db
)
Reaction time (h)
sugar
FAN
IP
Chapter 6 Production of a generic feedstock: subsequent hydrolysis
141
Table 6.2 kinetic model of further hydrolysis
model R2
Reducing sugar Xs=88.87 (1-e-0.209t) 0.95
FAN XFAN=6.03 (1-e-0.224t) 0.96
IP XIP=7.653 (1-e-0.411t) 0.99
The remaining lignocellulose after fungal autolysis and hydrolysis are shown in
Figure 6.11. The fibres consisted of undigested cellulose and lignin to support the
whole structure. Because of the liberation of cellulose and hemicellulose, a large
number of lignocellulose fragments are shown and seemed clearly hollowed out.
Figure 6.11 Lignocellulose degradation after 48 h autolysis/hydrolysis of fermented solids at 50°C, pH 4.8 citric buffer
Chapter 6 Production of a generic feedstock: subsequent hydrolysis
142
Characterisation of optimal reaction temperature and pH of the crude enzymes from SSF
Several hydrolytic enzymes were produced by T. longibrachiatum during solid
state fermentation of sugarcane bagasse with soybean hull. The complete
degradation of lignocellulose to sugars or oligosaccharides required synergistic
reaction between different enzymes, therefore the characterisation of each or co-
related enzymes was crucial to determine the hydrolysis yield (Ang et al., 2013).
The aim of this work was to characterise cellulase, beta-glucosidase and xylanase
produced by T. longibrachiatum culture under solid state fermentation. Optimal
pH and reaction temperature of the crude enzyme were determined (Figure 6.12
and 6.13).
In order to determine the optimal reaction pH of crude enzymes, assays were
carried out at various pH levels from 4 to 7 and the results are shown in Figure
6.12. The cellulase exhibited high activity in the range 4.8 - 6 with optimum activity
at pH 4.8 which is the same as that suggested for cellulase activity assay (Wood
and Bhat, 1988). The optimum pH for xylanase activity was at pH 5.4, and over
75% of the peak activity was displayed between pH 4 and 7. As for the beta-
glucosidase, the maximal enzyme activity was obtained at pH 4.8, indicating it was
the optimal condition for cellobiose hydrolysis in this system. Different pH stability
was observed for cellulase (maximum stability at pH 4.8), beta-glucosidase
(maximum stability at pH 4.8) and xylanase (maximum stability at pH 5.4).
Crude enzymes assay was conducted at different temperatures as indicated in
Figure 6.13 from 30 - 80°C with 10°C increment at pH 4.8 in order to determine
the optimal reaction temperature. Optimum temperature of crude cellulase for
hydrolysis of cellulose was found to be 50°C which showed good similarity with
other reported assay protocols (Saleh et al., 2014; Wood and Bhat, 1988). The
optimum temperature of beta-glucosidase and xylanase was 60°C which agreed
with other previous reports (Ang et al., 2013; D and M, 1994). Further increasing
Chapter 6 Production of a generic feedstock: subsequent hydrolysis
143
temperature greatly reduced the enzyme activity most likely because of enzyme
denaturation.
Figure 6.12 Effect of pH at 50°C on crude enzymes activity
Figure 6.13 Effect of temperature at pH 4.8 on crude enzymes activity
All hydrolytic enzymes presented an optimal pH between 4.8 and 5.4, while the
optimum temperature was 50°C for cellulase and 60°C for the other enzymes
0
20
40
60
80
100
4 4.8 5.4 6 7
Rel
ativ
e ac
tivi
ty (
%)
pH
cellulase
Beta-gulcosidase
Xylanase
0
20
40
60
80
100
30 40 50 60 70 80
Rel
ativ
e ac
tivi
ty (
%)
Temperature (C)
cellulase
Beta-gulcosidase
Xylanase
Chapter 6 Production of a generic feedstock: subsequent hydrolysis
144
(Table 6.3). The optimal conditions (temperature and pH) of crude enzymes
studied is not surprisingly slight different to those of subsequent hydrolysis of
fermented solids. The reason is due to complex polymers and several enzymes
involved in the complicated hydrolysis.
Table 6.3 Characteristics of hydrolytic enzymes from T. longibrachiatum grown on sugarcane bagasse and soybean hull
Chapter 7 Production of a generic feedstock: bioreactor studies
166
Heat loss through the walls of the bioreactor
The amount of heat that is removed through the wall is calculated from:
𝑄𝑤𝑎𝑙𝑙 =𝑘𝑚 ∙ 𝐴 ∙ (𝑇𝑏 − 𝑇𝑤𝑎𝑙𝑙) [W] Equation 7-19
with
km = 1.4, heat transfer coefficient of the wall [W ∙ (𝑚2 ∙ °C)−1]
A= heat transfer area of the reactor [m2]
Twall = average temperature of the wall [°C]
7.2.2.2.3 Model validation
Samples were taken during fermentation instead of measuring the whole
fermenting bed content in the end. Validation of the developed model was
investigated by comparing its estimations with the results of fermentation carried
out in the packed-bed bioreactor in this study (Figure 7.9 to 7.11). The predictions
of the model agree well with the measured values except some visible deviations
as fermentation time progressed. In all cases the model slightly under-predicts
total dry weight change, the water content of the bed and the average
temperature of the bed. Simulation of the dry matter loss fits the real data with
an acceptable discrepancy (Figure 7.9). The deviation found could be explained by
the variation of the conversion factor (kg) between the dry matter consumption
and the CO2 evolution. And this parameter used in this study was obtained from
the experiments in Circular tray fermenters instead of the packed-bed bioreactor.
Over the whole period of the fermentation process, the water balance model
showed a good fit to measured data of water content in the bed (Figure 7.10).
Some variance was observed between predicted and experimental values, but
they were not very remarkable. This deviation could be the value of kw used in our
study, using a simple relationship between glucose and CO2 instead of complex
carbohydrate and CO2. Predicted temperature of bed followed the trend of
qualitatively experimental results (Figure 7.11). This is apparently as a
Chapter 7 Production of a generic feedstock: bioreactor studies
167
consequence of an under-estimation of the heat release rate, which in turn may
result from the heat capacity of the control volume and value of 𝑌𝑞 𝑐𝑜2⁄ in
metabolic heat production. Heat capacity of control volume could be influenced
by the discrepancy between theoretical and experimental dry weight loss, while
𝑌𝑞 𝑐𝑜2⁄ could vary with different conditions and needs to be calibrated (Peña y Lillo
et al., 2001). Also, the measured data shown was the average value of
temperature in different positions of the packed-bed bioreactor and could be a
factor to cause the deviation of model accuracy. Alternatively, a different stage of
fermentation should be modelled and additional terms should be incorporated to
fit data well, however, this might make the model overcomplicated.
Figure 7.9 Comparison between predicted (line) and experimental (symbol) dry weight change in the packed-bed bioreactor during the course of fermentation
10
20
30
40
50
60
70
80
0 20 40 60 80 100 120
Dry
mat
ter
mas
s (g
)
Time (h)
Chapter 7 Production of a generic feedstock: bioreactor studies
168
Figure 7.10 Comparison between predicted (line) and experimental (symbol) average water content of bed during the course of fermentation
Figure 7.11 Comparison between predicted (dotted line) and experimental (solid line) average bed temperature during the course of fermentation
0.5
1
1.5
2
2.5
3
0 20 40 60 80 100 120
Wat
er c
on
ten
t o
f th
e b
ed (
g/g
)
Time (h)
25
27
29
31
33
35
37
39
41
0 20 40 60 80 100 120
Tem
per
atu
re( ℃
)
Time (hr)
Chapter 7 Production of a generic feedstock: bioreactor studies
169
Growth kinetics in SSF systems Knowledge of the growth during fungal fermentation is crucial in sequential
bioprocessing including solid state fermentation and further hydrolysis. In the case
study of this project, though, the biomass is not the final product. Also, it is
generally considered difficult to measure fungal growth in SSF due to the
adherence of the fungus on the heterogeneous and insoluble solid particles.
Because the fungal biomass cannot be separated easily from the fermenting
substrate, and because of the intimate interactions between microorganism and
the solids, it is not possible to obtain a homogeneous sample of biomass during
the fermentation. Kinetic models provide the basic knowledge needed in the
design and optimisation of fermentation processes. To accomplish a solid state
fermentation process based on lignocellulosic biorefinery, fungal cell
measurement throughout the fermentation is highly desirable. Since the use of a
direct measurement method such as the dry weight method (including the cells
and the solid particles) is impractical, the use of an indirect measurement could
be an alternative (Gretty K. Villena, 2011; Koutinas et al., 2003). Indirect
estimation methods can be mainly categorised into two groups as follows:
1. Off-line indirect measurement: The estimation of biomass components
could be determined either within the cell suspended in the cytoplasm,
such as DNA and RNA, or constituent materials of the cell walls, such as
ergosterol, glucosamine or chitin. Unfortunately, the content of all of these
components within the biomass cell can vary with culture conditions and
with the age of the mycelium. And most of components such as DNA, RNA,
glucosamine and chitin, could be measured in both, the fungal cells and the
substrate (Koutinas et al., 2003; Mitchell et al., 2006).
2. On-line indirect measurement: This method relies on detecting metabolic
activities of the biomass. Of these, CO2 evolution rate (CER) and O2
consumption rate (OUR) are most important. Production of CO2 and O2
consumption take place due to fungal growth and maintenance. Thus,
Chapter 7 Production of a generic feedstock: bioreactor studies
170
automatic on-line analysis of exit gases from SSF allows real time
information on the physiological state of the cultures to be correlated with
cell concentration (Koutinas et al., 2003; Machado et al., 2004).
In the following sections a model for Trichoderma longibrachiatum growth in SSF
is adapted to provide comprehensive information on the kinetics using, firstly, off-
line measurements of glucosamine, and secondly,. metabolic data from
respiratory analysis to determine carbon dioxide evolution rate (CER).
During experiments, both glucosamine and CER were monitored. The CO2
evolution and O2 consumption by cultures were measured continually using a gas
analyser. At specific times (24, 72, 96, 120, 168 h) the bioreactors were also taken
off to analyse biomass concentration, dry weight loss and substrate composition.
Cell concentration was obtained by the glucosamine method and total
carbohydrate of the substrates was determined by the Anthrone method. Each
analysis was conducted in duplicate. A schematic diagram of the bioreactor system
is shown in Figure 4.4.
7.3.1 Glucosamine production
As mentioned above, the time course of solid state fermentation by T.
longibrachiatum was monitored along with the dry matter weight loss of solids,
the total carbohydrate of substrates and the glucosamine concentration. At the
beginning of the fermentation (0-24 h), only 3.86% of the dry matter weight was
lost. During 24-120 h of incubation, rapid increase of dry matter weight loss to
20.71% occurred, attaining 26.76% after 168 h fermentation (Figure 7.12).
Generally speaking, the more carbohydrate is consumed, the more the
microorganism grows, and the more the dry matter weight loss (Weng and Sun,
2006). Only a little total carbohydrate was utilised during the first 72 h
fermentation, 4.74% of total carbohydrate loss. Thereafter (72-120 h), large
amounts (15.5%) of total carbohydrate consumption occurred. The total
carbohydrate utilisation reached 19.91% by the end of fermentation. Figure 7.12
Chapter 7 Production of a generic feedstock: bioreactor studies
171
shows a short lag phase of biomass growth during the first 24h, followed by
exponential growth until 120 h, where the stationary growth phase began.
Figure 7.12 Time course of solid state fermentation by T. longibrachiatum, using a circular tray bioreactor system
Based on the data collected for glucosamine concentration during the 7-day solid
state fermentation, a series of mathematical models were developed to describe
the correlation between total carbohydrate consumption and carbon dioxide
evolution. Linear, exponential, logistic, and deceleration models have been
reported in various SSF systems. Among these, the logistic equation is most
commonly used due to its relatively simple form and good fit to around 75% of the
literature growth profiles obtained in SSF (Viccini et al., 2001). The model
describes growth in microbial population depending on maximum biomass density,
specific growth rate, and time as described in equation 7-20.
𝑑𝑋
𝑑𝑡= 𝜇𝑚 (1 −
𝑋
𝑋𝑚)𝑋 Equation 7-20
where μm is the maximum specific growth rate (h-1) and Xm is the maximal
biomass concentration reached when dX/dt = 0 for X>0. This equation does not
consider the biomass death.
0
10
20
30
40
50
60
70
80
90
0.002
0.003
0.004
0.005
0.006
0.007
0.008
0.009
0.01
0.011
0.012
0 24 48 72 96 120 144 168
Tota
l car
bo
hyd
rate
(%
, w/w
)D
ry m
atte
r w
eigh
t lo
ss (
%, w
/w)
Glu
cosa
min
e (g
/g s
ub
stra
te)
Time (h)
Glucosamine
Dry matter weight loss
Total carbohydrate
Chapter 7 Production of a generic feedstock: bioreactor studies
172
Integration used for performing model fitting and parameter estimation in this
study, is presented by equation 7-21.
𝑋 =𝑋𝑚
1+((𝑋𝑚𝑋0
)−1)𝑒−𝜇𝑚𝑡
Equation 7-21
whereX0 is the initial biomass concentration at t=0.
In general, the carbon source, oxygen consumption rates, and CO2 evolution rates
can be described with the linear growth model of Pirt (Gelmi et al., 2002; Lareo et
al., 2006; Ooijkaas et al., 2000).
−𝑑𝑆
𝑑𝑡=
1
𝑌𝑋 𝑆⁄
𝑑𝑋
𝑑𝑡+𝑚𝑆𝑋 Equation 7-22
where S is the carbohydrate fraction (g/g substrate), Yx/s is the yield of the biomass
(g biomass/g total carbohydrate) and ms is the specific maintenance rate (g of total
carbohydrate/g of biomass·h). The second term of the equation, maintenance,
refers to the energy required for cell’s survival or preservation of a certain state,
which are not directly related to cell division (Bailey, 1986; Lareo et al., 2006).
Using the boundary condition at t0=0, X=X0, S=S0, Equation 7-22 can be integrated
to giving the following Equation 7-23 (Zhu et al., 2014).
𝑆(𝑡) = 𝑆0 − (1
𝑌𝑋 𝑆⁄) ∙ (𝑋(𝑡) − 𝑋0) − 𝑚𝑠 ∙
𝑋𝑚
𝜇𝑚∙ ln(1 −
𝑋0∙(1−𝑒𝜇𝑚∙𝑡)
𝑋𝑚)
Equation 7-23
The model has been successfully used to simulate growth in the reactor in the
experiment. Assuming a constant ratio of glucosamine to cell mass (i.e. X can be
represented directly by the measured glucosamine) and using MATLAB
(MathWorks), the differential equations described above were solved and fitted
to the data using nonlinear least squares. Figure 7.13 depicts the glucosamine level
(biomass concentration) obtained during circular tray bioreactor fermentations
and the estimated growth derived using the Equation 7-21. This shows a very good
fit between measurements and the logistic model (R2=0.99) as follows:
Chapter 7 Production of a generic feedstock: bioreactor studies
173
𝐺(𝑋) =0.011
1+1.971𝑒−0.022𝑡 Equation 7-24
Where G(X) is the glucosamine concentration
Assuming a constant value of glucosamine to cell mass ratio of 0.12 (Raimbault,
1998), equation 7-24 becomes, in terms of actual biomass:
𝑋 =0.088
1+1.95𝑒−0.023𝑡 Equation 7-25
Figure 7.13 Experimental (symbols) and predicted (line) for glucosamine
(biomass concentration) during 7 days solid state fermentation
As shown in Figure 7-13, the logistic equation predicted the fungal biomass fairly
accurately, with a short lag phase during the first 24 h, an exponential growth
phase that last until 120 h of incubation and a deceleration growth phase.
Maximum specific growth rates ( μm ) depend on the particular SSF system,
substrate type and microorganism used. The μm and Xm for T. longibrachiatum
cultivated in the Circular tray bioreactor system here were calculated to be 0.023
h-1 and 0.088 g/g substrate, respectively (Equation 7-24).
A wide range of maximum specific growth rates has been reported, from
0.027 h-1 (Santos et al., 2003) to 0.05 h-1 (Membrillo et al., 2011) using sugarcane
0.002
0.003
0.004
0.005
0.006
0.007
0.008
0.009
0.01
0.011
0 20 40 60 80 100 120 140 160 180
Glu
cosa
min
e (g
/g s
ub
stra
te)
Time (h)
Chapter 7 Production of a generic feedstock: bioreactor studies
174
bagasse as substrate for solid state fermentation. This indicates that the
utilisations of recalcitrant lignocellulose like sugarcane bagasse as a carbon source
is slower than other agricultural wastes, for example, wheat bran, where 0.15 h-1
was obtained for Trichoderma reesei (Smits et al., 1998).
Figure 7.14 shows the relation between the carbohydrate content and biomass
generation. It can be seen that the carbohydrate content is inversely proportional
to biomass production. The carbohydrate content of substrate was consumed
slowly while biomass generation was only 0.0009 g/g substrate during the initial
24 h. A plateau phase occurred when fungal biomass entered the log phase
(0.0009 to 0.0051 g/g substrate) from 24-96 h. Considering the heterogeneous
nature of the lignocellulosic feedstock, T. longibrachiatum may have to shift its
substrate utilisation between lignin and carbohydrate (cellulose and hemicellulose)
after depleting a fraction of the easily-accessible cellulose and hemicellulose
during early stage (Shi et al., 2012). Thereafter the total carbohydrate content
declined distinctly from 0.57 to 0.48 g/g substrate with gradual increase of
biomass concentration (0.0051 to 0.0065 g/g substrate).
Figure 7.14 Correlation between total carbohydrate loss and biomass generation (glucosamine) during 7 days solid state fermentation
0.4
0.45
0.5
0.55
0.6
0.65
0 0.001 0.002 0.003 0.004 0.005 0.006 0.007
Tota
l car
bo
hyd
rate
(g/
g su
bst
rate
)
Glucosamine (g/g substrate)
Chapter 7 Production of a generic feedstock: bioreactor studies
175
Total carbohydrate (cellulose and hemicellulose) served as the main carbon and
energy sources for supporting T. longibrachiatum growth. The estimated
parameters (μm and Xm) by nonlinear regression of equation 7-21 were used into
equation 7-23 to calculate the yield of the biomass (Yx/s) and the specific
maintenance rate (ms). Results presented in Figure 7.15 showed that the model
proposed from equation 7-23 in this research could predict carbohydrate
consumption well (R2=0.93). The estimated coefficient for total carbohydrate used
for cell growth (Yx/s) reached 0.19 (g biomass/g total carbohydrate) while the
maintenance coefficient (ms) is 0.12 (g total carbohydrate/g biomass*h). The value
of Yx/s was higher than those reported by Zhu et al. (2014) at 0.16 (g biomass/g
total carbohydrate) estimated for the solid-state fermentation using rice straw.
The reduction of carbohydrate content occurred slowly during the initial stage of
fermentation. The reason might be attributed to complex carbohydrate polymer
(cellulose and hemicellulose) and lignin interconnected by a great variety of
linkages which need to be broken down by a series of biochemical enzyme
reactions. Therefore, it takes time to secrete enzymes from fungi during the
beginning of fermentation (Crestini et al., 1998). The total carbohydrate trajectory
in Figure 7.15 agrees with that found by Padma and Singhal (2010).
Figure 7.15 Experimental (Symbols) and predicted (line) total carbohydrate consumption during 7 days solid state fermentation
0.002
0.102
0.202
0.302
0.402
0.502
0.602
0.702
0 20 40 60 80 100 120 140 160 180
Tota
l car
bo
hyd
rate
(g/
g su
bst
rate
)
Time (h)
Chapter 7 Production of a generic feedstock: bioreactor studies
176
Base on measured data, a straight line was found between dry matter weight loss
and biomass concentration during fermentation. The linear fit was X (glucosamine
where ∆𝐶𝑂2 is the amount of CO2 evolved (g/g initial dry substrate), and ∆𝐷𝑊 is
the amount of dry matter weight loss (g/g initial dry substrate). Several studies
showed a similar trajectory for Trichoderma species growth in the solid state
fermentation (Flodman and Noureddini, 2013; Smits et al., 1998, 1996). And the
all correlation coefficients of determination (R2) for the fit were from 0.87 to 0.89.
According to previous studies and the above experiment results, the
measurement of CO2 evolution is a potential way to model the fungal growth
during fermentation.
Figure 7.17 Correlation between dry matter weight loss and cumulative CO2 evolution
7.3.2 The respiratory gas model
Another alternative method to estimate biomass growth during solid state
fermentation is the use of CO2 evolution and O2 consumption. On-line analysis of
CO2 and O2 in the exhaust gases allows real time information of the physiological
state of the microorganisms to be collected. On-line derived variables such as CO2
y = 0.0642x + 0.0008R² = 0.897
0
0.002
0.004
0.006
0.008
0.01
0.012
0.014
0.016
0.018
0.02
0 0.05 0.1 0.15 0.2 0.25 0.3
CO
2 e
volu
tio
n (
g/g
sub
stra
te)
Dry matter weight loss (g/g substrate)
Chapter 7 Production of a generic feedstock: bioreactor studies
178
evolution rate (CER), O2 uptake rate (OUR) and the respiratory (RQ) could be used
to estimate the specific biomass growth rate on-line and assess the bioreactor
performance (Machado et al., 2004; Mitchell et al., 2006; Ooijkaas et al., 1998;
Rodriguez-Leon et al., 2008).
Figure 7.18a presents the time course of carbon dioxide evolution rate (CER) and
cumulative CO2 during fermentation using the circular tray bioreactors. A lag
phase in respiration was observed within the first 16 h of fermentation. The CO2
evolution was too low to be measured in this period. This might be due to an
adaptation of the microorganisms to the medium and to spore germination (de
Carvalho et al., 2006). After 16 h of fermentation, the CER increased dramatically
when biomass generation took place. The highest value of CER was detected at 27
h (0.27 mg/h·g). After this point, the exponential growth phase is suddenly
stopped when the oxygen uptake rate (OUR) dropped slightly from 1.9 to 1.5
mg/g·h and the highest temperature was reached (32.4°C). This means that fungal
cells were influenced by the temperature and oxygen level effect in the bioreactor.
The constant value of CER could be explained by the fact that the fungal cells
maintain a constant metabolic activity to utilise the remaining substrates.
Comparing the biomass growth and the accumulated CO2 evolution curve in Figure
7.18, it can be observed the similar trajectory from lag phase, exponential growth
phase and stationary phase during the fermentation. The profile of total CO2
evolution increased, corresponding to the beginning of the exponential growth
phase. After 113 h of cultivation, the beginning of the stationary phase as the total
CO2 evolution became flatter. The correlation coefficients (R2=0.99) confirmed a
close relationship between the total amount of CO2 evolved and cell growth during
solid state fermentation. This result indicated that the metabolic data can be used
to estimate indirectly the fungal growth through respiratory gas.
With the respirometric data, the maximum specific growth rate (μm) may also be
determined from the slope of the natural logarithm of the accumulated gases (CO2
and O2) (Gelmi et al., 2000; Machado et al., 2004). However, it is difficult to choose
Chapter 7 Production of a generic feedstock: bioreactor studies
179
the exact time interval of exponential growth phase, leading to a variation of the
maximum specific growth rate estimated easily.
Figure 7.18 Profile for T. longibrachiatum growth during 7days solid state fermentation (a) Accumulated CO2 and CO2 evolution rate (b) biomass growth (glucosamine)
The system simultaneously analyses both CO2 and O2 concentrations. In this
project, the choice of CO2 concentration, instead of O2, was due to the higher
sensitivity of the near infrared sensor. Considering that a CO2 evolution rates
which indicates biomass formation kinetics by combining growth associated and
0
0.05
0.1
0.15
0.2
0.25
0.3
0
2
4
6
8
10
12
14
16
18
0 20 40 60 80 100 120 140 160 180
CER
(mg/
h*g
)
Cu
mu
lati
ve C
O2
(mg/
g)
Time (h)(a)
cumulative CO2
CER
4
5
6
7
8
9
10
11
0 20 40 60 80 100 120 140 160 180
Glu
cosa
min
e (m
g/g
sub
stra
te)
Time (h)(b)
Chapter 7 Production of a generic feedstock: bioreactor studies
180
non-growth associated kinetics, thus, the following equation can written for CO2
evolution:
𝐶𝐸𝑅 =𝑑𝐶𝑂2
𝑑𝑡=
1
𝑌𝑋 𝐶𝑂2⁄
𝑑𝑋
𝑑𝑡+𝑚𝐶𝑂2𝑋 Equation 7-27
where CER is the carbon dioxide evolution rate (g of CO2/g of substrate·h), 𝑌𝑋 𝐶𝑂2⁄ is
the yield coefficient of cells with respect to CO2 (g of biomass/g of CO2) and 𝑚𝐶𝑂2is
the maintenance coefficient (g of CO2/g of biomass·h).
Equation 7-27 allows the estimation of the quantity of biomass that exists at a
particular time of the fermentation from the data of evolved CO2 by finite
determination of this evolution at particular time intervals. Equation 7-28 can be
integrated to obtain:
∫𝑑𝐶𝑂2
𝑑𝑡𝑑𝑡
𝑡=𝑛
𝑡=0= ∫
1
𝑌𝑋 𝐶𝑂2⁄
𝑡=𝑛
𝑡=0𝑑𝑡 +∫ 𝑚𝐶𝑂2𝑋𝑑𝑡
𝑡=𝑛
𝑡=0 Equation 7-28
Term ∫𝑑𝐶𝑂2
𝑑𝑡
𝑡=𝑛
𝑡=0 in equation 7-28 indicates that evolved CO2 in time t=n can be
evaluated from the relation between the carbon dioxide evolution (CER) and time.
Applying the Trapeze rule to the numerical integration of each term in the
expression, the following equation was obtained:
𝑋𝑛=(𝑌𝑋 𝐶𝑂2⁄ ∆𝑡(
1
2((
𝑑𝐶𝑂2𝑑𝑡
)𝑡=0
+(𝑑𝐶𝑂2𝑑𝑡
)𝑡=𝑛
)+∑ (𝑑𝐶𝑂2𝑑𝑡
)𝑡=𝑖
𝑖=𝑛−1𝑖=1 )+(1−
𝑎
2)𝑋0−𝑎∑ 𝑋𝑖
𝑖=𝑛−1𝑖=1 )
(1−𝑎
2)
where 𝑎 = 𝑚𝐶𝑂2𝑌𝑋 𝐶𝑂2⁄ ∆𝑡 Equation 7-29
To determine the carbon dioxide evolution ( 𝑌𝑋 𝐶𝑂2⁄ ) and biomass maintenance
coefficient (𝑚𝐶𝑂2), equation 7-27 can be rearranged to give:
𝐶𝐸𝑅
𝑋=
1
𝑋
𝑑𝑋
𝑑𝑡
1
𝑌𝑋 𝐶𝑂2⁄+𝑚𝐶𝑂2 Equation 7-30
The 𝑌𝑋 𝐶𝑂2⁄ and 𝑚𝐶𝑂2 can be estimated by linear regression through the plot of Y-
axis (𝐶𝐸𝑅
𝑋) versus X-axis (
1
𝑋
𝑑𝑋
𝑑𝑡). From the above calculation, bioprocess parameters
Chapter 7 Production of a generic feedstock: bioreactor studies
181
such as yield coefficient based on the carbon dioxide evolution (𝑌𝑋 𝐶𝑂2⁄ ), 0.071 g
biomass/g CO2, and biomass maintenance coefficient (𝑚𝐶𝑂2), 0.0092 g of CO2/g of
biomass·h, were determined. Figure 7.19 shows the estimated biomass and
measured one during the fermentation. It can be seen that the simulation could
fit the fungal growth during the first 120 h of fermentation. Towards the end of
fermentation the model estimated a bit higher growth than that measured.
Several studies have identified that accumulated CO2 evolution increases with
time. Therefore, a possible explanation for this might be that towards the end of
fermentation, the CO2 is more related to cell maintenance than to cell growth. In
general, therefore, it seems that fungal growth can be reliably assessed by
evaluating the CO2 evolved during the fermentation.
Figure 7.19 Profile of T. longibrachiatum growth, as glucosamine concentration (symbols) and simulation (line) during 7days solid state fermentation using circular tray bioreactor
The same approach used for the circular tray experiments was adopted to obtain
growth profile during fermentation in the multi-layer tray bioreactor and 1.5 L
packed bed reactor for the production of the microbial feedstock. Figure 7.20 to
7.21 show the estimated biomass kinetics based on the CO2 evolution for multi-
layer bioreactor and packed-bed bioreactor, respectively. The data fit the model
0.02
0.025
0.03
0.035
0.04
0.045
0.05
0.055
0.06
0 20 40 60 80 100 120 140 160 180
bio
mas
s (g
)
Time (h)
Chapter 7 Production of a generic feedstock: bioreactor studies
182
satisfactorily in both bioreactor systems. The linear regression using equation 7-
30 is applied to calculate carbon dioxide evolution (𝑌𝑋 𝐶𝑂2⁄ ), 0.8 g biomass/g CO2,
and biomass maintenance coefficient (𝑚𝐶𝑂2), 0.001 g of CO2/g of biomass·h, in
multi-layer bioreactor system. And in packed-bed bioreactor system, carbon
dioxide evolution (𝑌𝑋 𝐶𝑂2⁄ ), 0.19 g biomass/g CO2, and biomass maintenance
coefficient ( 𝑚𝐶𝑂2 ), 0.001 g of CO2/g of biomass·h, were determined.
The 𝑌𝑋 𝐶𝑂2⁄ and 𝑚𝐶𝑂2 values obtained from three different fermenters using
equation 7-30 were lower than those reported by Lareo et al. (2006) for C.
minitans (𝑌𝑋 𝐶𝑂2⁄ = 0.98, 𝑚𝐶𝑂2=0.0058). However, the biomass profile obtained
describes the experimental biomass production reasonably well (Figure 7.19 to
7.21).
Figure 7.20 Profile of T. longibrachiatum growth, as glucosamine concentration (symbols) and simulation (line) during 5 days solid state fermentation using multi-layer tray bioreactor
0
0.05
0.1
0.15
0.2
0.25
0.3
0 20 40 60 80 100 120
Bio
mas
s (g
)
Time (h)
Chapter 7 Production of a generic feedstock: bioreactor studies
183
Figure 7.21 Profile of T. longibrachiatum growth, as glucosamine concentration (symbols) and simulation (line) during 5 days solid state fermentation using packed-bed bioreactor
The experimental cell growth profile in the packed-bed bioreactor was strongly
correlated to the prediction of the model compared to that obtained in the multi-
layer bioreactor. There are several possible explanations for this result. One might
be the lack of adequate sampling data from the multi-layer bioreactor. Due to the
design of multi-layer bioreactor, it is impractical to take samples during the
fermentation, leading the error of estimated yield coefficient and maintenance
coefficient. Another reason for this is that the metabolic information (CO2 and O2)
of multi-layer tray bioreactor is the sum amount from biomass growth in three
different layers. Therefore, varied fungal growth profile could not be easy to
describe by the model using the total respiratory data.
Indirect methods to measure fungal growth like CO2 evolution may be useful in
solid-state fermentation by applying mathematical model. However, the
estimation of biomass through CO2 requires 𝑌𝑋 𝐶𝑂2⁄ and 𝑚𝐶𝑂2, which is strongly
dependent on specific growing conditions and fermentation systems. Therefore,
individual investigations for different fermentation conditions are necessary to
identify case-specific kinetic parameters.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0 20 40 60 80 100 120
Bio
mas
s (g
)
Time (h)
Chapter 7 Production of a generic feedstock: bioreactor studies
184
Characteristics of fermented solids during SSF Understanding changes in the structure of lignocellulose during solid state
fermentation could provide more valuable insights into the nature of this complex
bioprocessing. Microscopic observation, both optical and scanning electron
microscopy (SEM), can be a useful way to see changes in the substrates and fungal
growth in solid state fermentation. Moreover, thermogravimetric (TG) and
differential thermogravimetric (DTG) analysis of fermented substrates can also
provide information about the thermal decomposition profiles of lignocellulosic
components.
7.4.1 Microscopic observation
A fungal growth typically begins with the mixing of a spore inoculum with
substrate particles. Each particle initially has some spores attached to it. The
spores take around 10 hours to germinate while the substrate bed is under the
optimal conditions (water activity, temperature and pH etc.). As shown in Figure
7.22a, once each spore germinates, a germ tube extends away from the spore and
branches to give daughter hyphae, which extend and further branch to give an
expanding micro-colony. The original extension of the germ tube is fuelled by
reserves in the spores, but the continued expansion and growth depends on
nutrients from the substrate. Hyphae from near micro-colonies meet one another,
which cause interactions (Mitchell et al., 2006). During the growth period some
hyphae will also extend above the surface (aerial hyphae), and others will have
penetrated into the substrate (Figure 7.22b).
Chapter 7 Production of a generic feedstock: bioreactor studies
185
Figure 7.22 Change in biomass distribution during a static SSF process with a fungus. (a) Growth to cover the particle surface during the early stages of the fermentation, shown with an overhead view of the particle surface. (b) Development of aerial and penetrative hyphae during the late phase of the fermentation, shown with a side view of a cut through two particles. (adapted from (Mitchell et al., 2006)
Sampling is difficult throughout SSF experiments and it is not easy to take
homogeneous samples from the bioreactor during fermentation. Therefore, to
observe fully the fungal morphology in the bioreactor is almost impossible during
solid state fermentation., In the experiments reported here, when the bioreactor
was disassembled at the end of the fermentation, it was observed that fungal
growth was almost uniform throughout the solids and formed a “cake” as shown
in Figure 7.23a. In Figure 7-23b it can be clearly seen that aerial hyphae from
different micro-colonies meet and intermesh as suggested by Mitchell et al.,
(2006). Since the moist air supplies sufficient water to avoid drying by evaporation
during fermentation, the whole cake seems to be colonised and there is not the
suboptimal growth at the top of the substrate that is often associated with dry
systems. The experiment showed that lignocellulosic waste could be fermented
successfully in a bioreactor without mixing when moist air is supplied.
Chapter 7 Production of a generic feedstock: bioreactor studies
186
Figure 7.23 Fermented solids of T. longibrachiatum after 5 days fermentation using a multi-tray bioreactor with moist air aeration (a) fungal cake (b) aerial hyphae intermeshed above the surface
The untreated sugarcane bagasse exhibited rigid and highly ordered fibrils in the
SEM image (Figure 7.24). The ability of the fungus (Trichoderma longibrachiatum)
to penetrate into the fibres can be appreciated in the micrographs taken using the
SEM (Figure 7.25 and 7.26) which show high growth in the sugarcane bagasse. The
spores were distributed over the surface of the fibres, and the hyphae clearly form
a microscopic network (mycelium) inside the substrate, through which the
nutrients are absorbed. After solid-state fermentation, the fibres can be seen to
be partly degraded and separated by mycelium, creating a larger accessible
0.5 cm
Chapter 7 Production of a generic feedstock: bioreactor studies
187
surface area to facilitate subsequent enzymatic hydrolysis. Unfortunately, it was
not possible to remove the fungi in order to take pictures of samples before and
after fungal pretreatment, so these effects could not be compared.
Figure 7.24 SEM image of a sugarcane bagasse structure, showing rigid and highly ordered fibrils
Chapter 7 Production of a generic feedstock: bioreactor studies
188
Figure 7.25 SEM image of mixed sugarcane bagasse and soybean hull after 120 h of fermentation covered by hyphae of Trichoderma longibrachiatum
Chapter 7 Production of a generic feedstock: bioreactor studies
189
Figure 7.26 SEM image of mixed sugarcane bagasse and soybean hull after 120 h of fermentation covered by spores of Trichoderma longibrachiatum
7.4.2 Thermogravimetric analysis
The thermogravimetric (TG) and derivative thermogravimetric (DTG) analysis of
the raw materials (sugarcane bagasse and soybean hull) and fungal pretreated
materials are shown in Figure 7.27. In general, the thermal decomposition of these
fibres consists of four stages: stage 1 – moisture evaporation; stage 2 –
decomposition of lignocellulosic components of hemicelluloses; stage 3 –
decomposition of cellulose and lastly; stage 4 – decomposition of lignin (Ishak et
al., 2012; Yang et al., 2007). Stage 1 ranges from 45 to 150°C. With temperature
Chapter 7 Production of a generic feedstock: bioreactor studies
190
increase, hemicellulose starts its decomposition (stage 2) easily, between 220 and
315°C. Cellulose pyrolysis (stage 3) is focused at a higher temperature range (315–
400°C) with the maximum weight loss rate. Lignin is the most difficult to
decompose (stage 4). It decomposes slowly in the range from 300°C to 500°C
(Zhang et al., 2014).
Since all samples were dried prior to the TGA, the weight loss before stage one
was ignored. From the TGA distributions, it can be seen that the difference among
the materials is insignificant. With further examination upon the DTG curves, the
distribution of the raw materials is characterized by a multiple-peak distribution.
The first and the second peaks, developing at 236 and 365°C, respectively,
represent the thermal degradations of hemicellulose and cellulose in the
substrates. Once the materials undergo fungi pretreatment, the first peak
disappears, suggesting that hemicellulose has been removed to a great extent. The
DTG curves also revealed that the impact of the pretreatment on cellulose is
slightly varied in the second peak. According to the thermogravimetric analysis of
samples, the pyrolysis profile of fungi treated sugarcane bagasse and soybean hull
shifted to lower temperatures and with weaker major peaks. This result indicates
that the fungal pretreatment improved the efficiency of pyrolysis by reducing the
demand for a high thermal degradation temperature.
Chapter 7 Production of a generic feedstock: bioreactor studies
191
Figure 7.27 TG and DTG analysis of (a) untreated substrates (b) treated substrates
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0
10
20
30
40
50
60
70
80
90
100
75 175 275 375 475 575
DTG
(-d
m/d
t)
We
igh
t (%
)
Temperature (℃ )
TGA
DTG
Cellulose
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0
10
20
30
40
50
60
70
80
90
100
75 175 275 375 475 575
DTG
(-d
m/d
t)
We
igh
t (%
)
Temperature ( ℃)
TGA
DTG
Cellulose
(b)
Hemicellulose
(a)
Chapter 7 Production of a generic feedstock: bioreactor studies
192
Summary The object of the study reported in this chapter was the development of a
consolidated bioprocess using solid-state bioreactors to produce a generic
microbial feedstock. Experiments were conducted using two different types of
bioreactor (multi-layer tray bioreactor and packed-bed bioreactor). Results from
experiments confirmed that the upper layer of packed-bed bioreactors and
bottom tray of multi-layer tray bioreactor can have better sugar and FAN
production through subsequent hydrolysis. For the packed-bed bioreactor, an
equation based on material and energy balances was proposed for predicting and
simulating average bed temperature and average water content of the bed. A
microbial growth model based on metabolic data has been proposed for the
estimation of fungal growth in the solid-state fermentation. Yield of fungal cells
based on CO2 evolution and maintenance coefficients were estimated from
fermentation. This method can help to estimate and monitor cell growth during
fermentation from the on-line indirect measurement.
Several methods were also used to analysis the chemical and physical change of
substrates during fermentation. The SEM images showed the hyphae granule
penetration and the variation of substrate structure. Thermogravimetric analysis
also demonstrated that significant hemicellulose degradation occurred during the
solid-state fermentation.
Chapter 8 Evaluation of the generic feedstock
193
CHAPTER 8
Evaluation of the generic feedstock
Introduction The feasibility of a generic feedstock obtained from sugarcane bagasse and
soybean hull by the process developed in this project was evaluated. The
application of these media as the sole nutrient source for the production of
ethanol has been carried out. Based on the results reported in chapter 5 and in
this chapter, a mass balance for ethanol production from raw materials and a
preliminary economic assessment were conducted and are presented in section
8.3 and 8.4.
Ethanol fermentation Many lignocellulosic materials have been tested for developing large-scale ethanol
production during the past two decades. These processes provide several
desirable features: a secure source of energy supply, limited conflict with land use
for food production, and reduction of fossil fuel demand (Margeot et al., 2009). In
order to test the potential of the environment-friendly process developed in this
project, a number of experiments were carried out using the generic feedstock
produced by sequential bioprocessing for ethanol production.
Yeast strain, Saccharomyces cerevisiae was cultivated on slant of YDX medium
(yeast extract: 5 g/L, glucose: 5 g/L and xylose: 10 g/L) for pentose adaption. To
prepare inoculum the yeast were culture in 100 mL of YDX medium in 250 mL
flasks at 30°C on a 160 rpm shaker for 30 h. After being washed twice with sterile
distilled water to remove nutrient residuals, yeast cells harvested from 50 mL of
broth by centrifugation at 4000 rpm for 5 min were concentrated in to 4 mL. The
100 mL of hydrolysate were inoculated with 4 mL of a suspension of S. cerevisiae
giving a concentration of around 106 cell/mL. Ethanol fermentations were carried
Chapter 8 Evaluation of the generic feedstock
194
out anaerobically at 32°C, 180 rpm in 250 mL Duran bottles fitted with “swan-neck”
caps to allow release of fermentation gases. As YD media, fermentations of
solutions containing 15 g/L glucose and 10 g/L yeast extract were performed. After
the fermentation, samples were taken for sugar, FAN and ethanol analysis and
results are shown in Table 8.1 and 8.2. The theoretical ethanol production could
be calculated from the reducing sugar, assuming that the theoretical ethanol yield
for fermentation is 0.511 g per g of reducing sugars (Wongwatanapaiboon et al.,
2012). The percentage of theoretical maximum ethanol yield was calculated by
equation 8-1.
%𝑜𝑓𝑡ℎ𝑒𝑜𝑟𝑒𝑡𝑖𝑐𝑎𝑙maximum𝑒𝑡ℎ𝑎𝑛𝑜𝑙𝑦𝑖𝑒𝑙𝑑
=𝑔𝑜𝑓𝑒𝑡ℎ𝑎𝑛𝑜𝑙𝑜𝑏𝑡𝑎𝑖𝑛𝑒𝑑
𝑔𝑜𝑓𝑟𝑒𝑑𝑢𝑐𝑖𝑛𝑔𝑠𝑢𝑔𝑎𝑟𝑠𝑐𝑜𝑛𝑠𝑢𝑚𝑒𝑑×0.511× 100
Equation 8-1
The hydrolysate resulting from the hydrolysis and autolysis of fermented
substrates for mixed substrates (sugarcane bagasse and soybean hull, 6:4) was
used as the feedstock for ethanol fermentation. The composition of the
hydrolysates and that of the YD media, as well as the corresponding ethanol
concentration, are presented in Table 8.1. The initial fermentation broth contains
2.53 g/L of glucose, 7.42 g/L of xylose and 1.01 g/L of arabinose. Hemicelluloses
contain most of the pentose sugars (xylose, arabinose mannose, galactose and
rhamnose), and occasionally small amounts of hexose sugar (glucose) as well. In
contrast, cellulose contains only anhydrous glucose. Through the sequential
bioprocess in this project. The abundant sugars come from the hemicellulose
instead of cellulose. Large amounts of hemicellulose sugars, making up over half
of the total reducing sugars, were produced after hydrolysis. After an anaerobic
incubation time of 30 h, the ethanol concentration of 3.14 g/L was obtained. Table
8.2 shows that ethanol yield based on consumed total sugars was 0.31 g/g (61.4%
of theoretical ethanol maximum), as 69% of total sugars, 62.6% of pentose sugar
and 100% of glucose were consumed respectively. It is not surprising that all
glucose was exhausted and about 37% of the pentose remained in the broth. The
Chapter 8 Evaluation of the generic feedstock
195
naturally occurring Saccharomyces yeasts that are currently used by industry lack
the ability to utilise xylose because of the absence of genes required for
assimilation of these molecules (Hector et al., 2011). The yeast does not have the
ability of fermenting pentose sugars efficiently even though it was cultured and
adapted in high xylose content medium.
Table 8.1 Ethanol fermentations of S. cerevisiae on different media