Bioresource Technology 233 (2017) 342–352

Contents lists available at ScienceDirect

Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech

Fermentation and crystallization of succinic acid from Actinobacillus

succinogenes ATCC55618 using fresh cassava root as the main substrate
Nguyen Thi Huong Thuy a, Artit Kongkaew a, Adrian Flood c, Apichat Boontawan a,b,⇑
a
School of Biotechnology, Institute of Agricultural Technology, Suranaree University of Technology, 111 University avenue, Muang district, Nakhon Ratchasima 30000, Thailand
b
Cassava Research Center, Suranaree University of Technology, 111 University avenue, Muang district, Nakhon Ratchasima 30000, Thailand
c
Department of Chemical and Biomolecular Engineering, Vidyasirimedhi Institute of Science and Technology (VISTEC), Wangchan District, Rayong 21210, Thailand

h i g h l i g h t s

Fresh cassava root was used as the main substrate for succinic acid fermentation.

Viscosity reduction was necessary before the saccharification process.

(NH4)2HPO4 was supplemented as a nitrogen source to substitute yeast extract.

Pre-treatment using nano-filtration enhanced the purity of the crystal.

High seeding strategy resulted in the highest purity of succinic acid crystal.

a r t i c l e i n f o a b s t r a c t

Article history: The fermentation of succinic acid from fresh cassava root using Actinobacillus succinogenes ATCC55618,
Received 16 December 2016 and the recovery of the product using crystallization were investigated. Fresh cassava root is an ideal suc-
Received in revised form 23 February 2017 cinic acid feedstock due to its low price and high starch content. Saccharification was carried out using
Accepted 24 February 2017
commercially available enzymes and diammonium phosphate was used as an inexpensive nitrogen
Available online 28 February 2017
source. Different fermentation modes were compared in terms of product yield and productivity.
Results for fed-batch fermentations showed that a succinic acid titer of 151.44 g/L, with yield and produc-
Keywords:
tivity of 1.51 gSA/gglucose and 3.22 g/L/h could be obtained. Seeded batch cooling crystallization was inves-
Succinic acid
Actinobacillus succinogenes
tigated after pre-treatment using nanofiltration. A succinic acid crystal purity of 99.35% with a relative
Cassava crystallinity of 96.77% was obtained from high seeding experiments. These results indicated that fresh
Fermentation cassava roots could be an economically alternative feedstock for a high quality succinic acid production.
Crystallization Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction of SA has become an attractive option due to the decreasing avail-

Succinic acid (SA) is considered as one of twelve compounds
that are used as organic building blocks, and has drawn worldwide
concerns for its application in biodegradable polymers, food, cos-
metic, fine chemical, and pharmaceutical industries (Li et al.,
2010). Moreover, green technology is becoming more of a driving
force because the traditional production of SA mainly depends on
hydrogenation of petroleum-derived maleic anhydride, which
requires high temperature, high pressure, costly catalysts, and
may cause environmental problems (Chen et al., 2014; Zeikus
et al., 1999). The use of renewable feedstocks for the production
⇑ Corresponding author at: School of Biotechnology, Institute of Agricultural
Technology, Suranaree University of Technology, Nakhon Ratchasima 30000,
Thailand.
E-mail address: apichat@sut.ac.th (A. Boontawan).
ability of petroleum and its derivatives. Bio-based SA is currently
produced industrially by companies such as BioAmber, Reverdia,
PTT-MCC Biochem, and Myriant (Lu et al., 2015; Cok et al., 2013).
SA can be produced by many microbial strains such as Basfia
succiniciproducens (Danavia et al., 2016), Mannheimia succinicipro-
ducens (Lee et al., 2006), Anaerobiospirillum succiniciproducens
(Bretz, 2015), and Escherichia coli (Carel and Willie, 2013) respec-
tively. However, these bacteria have more complex requirements,
including high nutritional level, and they are also difficult to be
modified. Among these strains, Actinobacillus succinogenes is a nat-
urally occurring succinic acid producer, and is also one of the most
promising strains because of its ability to ferment high succinic
acid titer from a wide variety of relevant sugars including glucose,
sucrose, and fructose. This is an excellent trait for an industrial suc-
cinic acid production process based on renewable and low-cost
feedstock (Carvalho et al., 2016; Lu et al., 2015). Recent studies

http://dx.doi.org/10.1016/j.biortech.2017.02.114
0960-8524/Ó 2017 Elsevier Ltd. All rights reserved.
 N.T.H. Thuy et al. / Bioresource Technology 233 (2017) 342–352
have been reported on succinic acid production from cheap and
readily available cellulosic sources; for example, carob pods, corn
stover (Zheng et al., 2010), and sugar cane bagasse (Elcio and
Nei, 2011). In addition, cost effective nitrogen sources have been
investigated such as corn steep liquor (Xi et al., 2013), and spent
yeast cells (Chen et al., 2011a,b). However, technical problems still
exist; pretreatment steps are difficult, while the obtained nutrient
content is still low resulting in a low concentration of final succinic
acid product. Cassava root is the world’s sixth most important crop,
and is grown in almost all tropical countries including Thailand
(Collares et al., 2012). Cassava root contains a starch content of
approximately 25–30 wt%, and is employed as a raw material for
various industrial applications such as animal feed, food, modified
starch, and ethanol fermentation. It is available all year round with
a low price of approximately 0.057 $/kg. Recently, the application
of cassava root for bio-based succinic acid production is strongly
increasing because of its high carbohydrate content as well as
important nutrients such as carotenoids and vitamin A, particu-
larly on the root peels (Brimer, 2015).
In addition to fermentation, downstream processing plays an
important role in obtaining the final succinic acid product at a high
yield and with the desired purity. Crystallization is considered as
an effective recovery step for the separation of solid-liquid mix-
tures due to its ability to provide a high purity product (Li et al.,
2010). For organic species, batch crystallizations seeded with a
small amount of crystal of the desired product are most often used
for product recovery. Seed loading in crystallization is regarded as
the main variable that can be used for controlling and optimization
of the crystal size, shape, morphology, crystallinity, and purity.
This has previously been investigated on acetylsalicylic acid
(Eder et al., 2011), protein (Liu et al., 2010), ammonium sulfate
(Hojjati and Rohani, 2005), and glycine (Doki et al., 2004).
In this work, the succinic acid fermentation by A. succinogenes
ATCC 55618 was investigated using fresh cassava roots as the main
substrate. The medium was prepared from enzymatic hydrolysis of
the cassava mash, and fermentations were carried out with and
without use of di-ammonium phosphate (DAP) as a cheap nitrogen
supplement. Fed-batch fermentations were attempted to increase
the succinic acid titer, yield, and volumetric productivity. After fer-
mentation, cell removal was carried out by centrifugation followed
by nanofiltration to remove impurities including proteins, macro-
molecules, and especially multivalent ions. Recovery of succinic
acid from the mixture of organic acid by-products was investigated
using seeded batch cooling crystallization. It was the final step to
maximize the purity of the succinic acid product as all other
organic acid by-products were present in the mother liquor. The
effectiveness of seeding strategies was compared in terms of the
recovery and purity of the obtained succinic acid crystal.
2. Materials and methods
2.1. Raw materials
Fresh cassava roots (variety Rayong 80 at 8 months harvesting
time) were obtained from Sanguan Wongse Industries Co., Ltd.
(Nakhon Ratchasima, Thailand). Proximate analysis was carried
out for moisture, carbohydrate, protein, lipids, and ash, respec-
tively. Analytical grade succinic acid, acetic acid, lactic acid, and
formic acids were purchased from Sigma (Singapore). Commercial
enzymes including LiquozymeÒ (a-amylase), SpirizymeÒ Fuel
(gluco-amylase), and ViscozymeÒ (a mixture of cellulase, xylanase,
hemicellulase, and endo-b-glucanase) were obtained from Novo-
zymes (Bagsvard, Denmark). All others chemicals used for medium
preparation were obtained from HiMedia Laboratories (Mumbai,
India). Actinobacillus succinogenes ATCC 55618 was obtained from
343
the American Type Culture Collection (Virginia, USA). The strain
was maintained in 10% skim milk at 20 °C.
2.2. Liquefaction and saccharification
Pre-weighed fresh cassava root was washed, sliced, and minced
using a grinder without peeling. DI water was added at a 1:2 wt
ratio to produce the cassava mash. The first step in processing
was viscosity reduction because non-starch polysaccharides create
high viscosity solutions which are detrimental in terms of power
consumption in the mixing process. Typically, the cassava mash
(450 g) was placed in a 5-L jacketed glass reactor, and was mixed
with various concentrations of ViscozymeÒ at 50 °C using a
laboratory mixer (IKA, Germany). After 150 min, the mash was
iquefied by an addition of 0.05 wt% of thermostable a-amylase
(TermamylÒ) at the working temperature of 90 °C for 2 h. The tem-
perature of the liquefied mash was then decreased to 60 °C before
being saccharified by an addition of SpirizymeÒ Fuel at 0.05 wt%
for 12 h. Since the liquefaction and saccharification process fol-
lowed a previous report (Poonsrisawat et al., 2014) process opti-
mization was not investigated in this work. The reducing sugar
obtained after the saccharification process was approximately
255 g/L with a protein content of 12.5 g/L. After centrifuge at
10,000 for 10 min, the supernatant was supplemented with
1.0 wt% of di-ammonium phosphate (DAP) in order to provide
assimilable nitrogen for the bacteria. Supplementing the super-
natant with this inorganic nitrogen has advantages because it
can replace expensive yeast extract. DAP use does not result in sig-
nificant problems in the downstream processing. For fed-batch fer-
mentation, this supernatant (cassava hydrolysate) was further
concentrated by using a rotary vacuum evaporator until glucose
concentration reached approximately 1000 g/L, and was used as
the feeding solution.
2.3. Fermentations and pre-treatment steps
A. succinogenes ATCC 55618 seed cultivation was prepared by
the transfer of 1 mL of the stock to 20 mL of seed culture medium,
and this mixture was anaerobically incubated for 12 h at 35 °C as
reported previously (Lubsungneon et al., 2014). The medium for
seed culture contained the following (per liter): 17.0 g tryptone,
3.0 g peptone, 2.5 g glucose, 5.0 g NaCl, and 2.5 g K2HPO4, respec-
tively. Subsequently, the seed culture was transferred to 200 mL
of the seed culture medium in a 500 mL shake flask. The flask
was first flushed with CO2 to remove oxygen before being closed
with a rubber stopper. The standard batch fermentation synthetic
medium consisted of (per liter): 150 g glucose, 15 g yeast extract,
3.0 g KH2PO4, 1.5 g K2HPO4, 3.0 g (NH4)2PO4, 1.0 g NaCl, 0.3 g
MnCl26H2O, 0.3 g CaCl22H2O, 0.07 g MnCl2, and 0.5 g anti-foam
agent. The fermentation process was carried out in a 5 L fermenter
(BiostatÒ B plus, Sartorius, Germany) with a working volume of
4.0 L. After sterilization, an anaerobic condition was ensured by
flushing with CO2. The temperature was controlled at 35 °C, and
the pH was automatically adjusted to 6.8 by the addition of
40 wt% MgCO3. The agitation speed was fixed at 200 rpm. For
fed-batch fermentation, the initial volume and glucose concentra-
tion were 2.7 L, and 70 g/L, respectively. The feeding solution
(1000 g/L glucose) was intermittently fed into the bioreactor. The
residual glucose was maintained in the range between 30 and
40 g/L.
The broth was acidified to pH 3.0 with H2SO4 after the end of
the fermentation process to liberate the organic acids. The super-
natant was subjected to further purification using a 400 Da ceramic
nanofiltration system (Fraunhofer IKTS, Germany) to remove pro-
teins, multivalent ions, and macromolecules (Lubsungneon et al.,
2014). The NF permeate containing succinic acid and other organic
344 N.T.H. Thuy et al. / Bioresource Technology 233 (2017) 342–352
acid by-products was further concentrated to approximately
250 g/L SA concentration using a rotary vacuum evaporator (Buchi,
Switzerland), and was kept in a cold room for crystallization
experiments.
2.4. Crystallization experiment
The cooling crystallization experiments were performed in a
500 mL jacketed glass vessel with a working volume of 250 mL
for each run. The temperature of the solution was controlled using
a programmable refrigerated circulator (Julabo F25, Germany). A
two-blade marine type impeller operated at 400 rpm was used to
agitate the system. This agitation speed was chosen to be high
enough to guarantee that all particles were well suspended
throughout the process, but low enough to avoid attrition of crys-
tals or entrainment of bubbles due to vortex formation. Three ser-
ies of experiments were conducted to characterize the dynamics of
the process when applying the seed loading. The seeding strategies
were classified as un-seeded (0 wt%), low seeding (5 wt%) and high
seeding (10 wt%) of the final isolated crystal mass, respectively.
Seed succinic acid crystals of different sizes were prepared using
a standard sieve analysis using sieve sizes between 100 and
1000 mm. The product obtained on the 650 mm sieve was collected
for seeding. The temperature was lowered from 70 °C to 30 °C
using a linear cooling profile with a rate of 2.5 °C/h. Crystallization
of succinic acid commenced by addition of the seed succinic acid
crystals into the solution when the temperature of the solution
reached 50 °C. Subsequently, the cooling rate was increased to
obtain the temperature of 4 °C within 1 h. This temperature was
maintained for another 4 h. The succinic acid crystal was separated
from the mother liquor by means of filtration. Finally, the obtained
succinic acid crystal was washed with cold DI water (1 °C) to
remove any impurities on the surface of the product, and was dried
at 40 °C for 12 h.
2.5. Analytical methods
Cell concentrations were measured using a spectrophotometer
at the wavelength of 660 nm (OD660). Cell concentration was calcu-
lated from a calibration curve between the optical density and cell
dry weight (1 OD660 = 0.34 g/L). This value was obtained from a
previous work (Lubsungneon et al., 2014). The organic acid concen-
trations were analyzed by HPLC (Thermo Scientific, USA) using a
ZORBAX SB-Aq (4.6 mm  150 mm) column. Quantification was
carried out by UV detection at a wavelength of 210 nm. The mobile
phase was 1% acetonitrile +99% 20 mM Na2HPO4 (pH 2.0) at a flow
rate of 1 mL/min. The injection volume was 20 mL. Viscosity mea-
surements were carried out using a viscometer (Brookfield DV2T,
USA). Protein and reducing sugar concentrations were estimated
using the Bradford (Bradford, 1976), and DNS method, respectively.
Succinic acid yield and productivity of the fermentation was calcu-
lated based on the ratio of the mass of succinic acid produced (g) to
the mass of glucose consumed (g), and the amount of succinic acid
produced (g) per volume (L) per time (h). The recovery and purity
of succinic acid crystal product were calculated by the followed
equations (Wong et al., 2012):
Experimental yield ðgÞ
Recovery ð%Þ ¼
Initial mass of SA in the broth ðat t0 Þ ðgÞ þ seed mass ðgÞ
Dry weight of SA in crystals recovered ðgÞ
Purity ð%Þ ¼  100%
Dry weight of crystals recovered ðgÞ
ð2Þ
The glucose and protein removal rate were calculated as follows
(Sun et al., 2014):
 
MT
DG=P ¼ 1   100% ð3Þ
Mw
where DG/P: Glucose or protein removal rate (%), MT: mass of glu-
cose or protein in the top phase or still remaining in the obtained
crystals (g), and Mw: The total mass of glucose or protein in the
broth before crystallization (g), respectively.
The particle size distribution (PSD) of succinic acid crystals was
measured by laser scattering via a particle size distribution ana-
lyzer (Horiba LA-950V2, Kyoto, Japan). The crystal morphology
was examined by a Stereoscopic Light Microscope (Olympus
SZX7, Japan) with an Olympus SC100 camera directly attached at
1.25 magnification. The characterization of crystalline products
was also carried out with a D2 phaser X-ray diffractometer (XRD)
with CuKɑ diffraction k = 1.5406 Å) operating at 30 kV, and 10 mA
with a graphite monochromator in the diffracted beam path (Bru-
ker AXS GmbH, Germany). The step time and step size were
selected as 0.2 s and 0.02°, respectively. The relative crystallinity
of SA crystal products was estimated based on the X-ray diffraction
data focusing on the three most intense peaks, which occurred at
the 2h values of 20.0°, 25.5° and 31.5°, respectively. The sample
with highest crystallinity was used as the reference to calculate
the relative crystallinity as follows: (%) relative crystallinity = (total
area of crystalline peaks)/(total area of crystalline peaks for the
highest crystallinity sample) (Emrani et al., 2011). In addition,
the diffracted intensity algorithm, DIFFaX, was used for calculation
of phase purity by succinic acid to magnesium sulfate ratio. A scan-
ning electron microscope (JSM – 6010LV, Tokyo, Japan) was used to
observe the cross-section of fresh cassava root, and to confirm the
size and morphology of the crystalline products, and the presence
of imperfections (Emrani et al., 2011).
3. Results and discussion
3.1. Viscosity reduction and saccharification
Cassava is one of the cheapest carbon sources, and is readily
available in Thailand. The chemical composition of the cassava
feedstock depends on its variety, age, and soil condition. In this
work, proximate analysis for the chemical composition of the fresh
cassava root was water 60.44 wt%, carbohydrate 31.23 wt%, lipid
1.41 wt%, protein 1.32 wt%, fiber 1.82 wt%, and others 3.78 wt%,
respectively. For fresh cassava root, the problem of high viscosity
has been encountered during preparation of fermentation broths
due to the presence of non-soluble lignocellulosic materials. Disad-
vantages of high viscosity broths were observed, including high
power for mixing, high water consumption, resistance to soli-
liquid separation, and incomplete hydrolysis of starch resulting
in a low fermentation yield. Thus, mash with high dissolved solid
content and reduced viscosity is required prior to the mash enter-
ing the saccharification process. This can be achieved by using
enzymatic pre-treatment to degrade the cell wall of the cassava
roots after the mashing step. According to the manufacturer, the
major enzyme activity from ViscozymeÒ is via endo-b-glucanase
 100% ð1Þ
that hydrolyzes (1,3)- or (1,4)-linkages in b-D-glucans; it also has
the activity of xylanase, cellulase, and hemicellulase. Fig. 1 shows
that a reduction in the viscosity of the cassava mash was observed
for all dosages of the enzyme. For the control experiment where no
 N.T.H. Thuy et al. / Bioresource Technology 233 (2017) 342–352
Fig. 1. Change in the viscosity of the cassava mash during the enzymatic pre-treatments with ViscozymeÒ at 50 °C.
ViscozymeÒ was added, the viscosity of the cassava mash was con-
stant throughout the 150 min of operation. In contrast, experimen-
tal results showed that an increase in enzyme dosage caused an
increase in the rate at which the viscosity is reduced. The maxi-
mum enzyme loading of 32 fungal b-glucanase units (FBG) per g
cassava mash resulted in the highest viscosity reduction rate,
approximately 14.83 mPa.s/min during the first 60 min. After this
period, the viscosity reduction rate gradually reduced until it had
nearly reached a plateau at the end of the operation. A final
viscosity of 710 mPa.s was obtained resulting in a 72.16% reduction
in the viscosity. For the enzyme loading of 8, 16, and 24 FBG per g
cassava mash, the reduction of viscosity was also observed, but at a
lower rate in comparison to 32 FBG per g cassava mash. Accord-
ingly, this enzyme loading was selected as the optimal dosage for
subsequent studies. Higher dosage could result in a shorter opera-
tion time; however, it increases the cost of the enzyme. Optimiza-
tion for this operating step should be considered depending on the
economics of the process. In addition, a change in the morphology
of the cassava cell wall was observed using SEM. The starch gran-
ules were confined within the cell-wall matrix in fresh cassava
root. However, after the treatment with ViscozymeÒ, breakdown
of the cell-wall due to the hydrolytic activity of the enzymes was
clearly observed.
3.2. Fermentations
3.2.1. Batch fermentation of SA
The objective of this part was to compare the fermentation per-
formance between the standard batch synthetic medium and the
saccharified fresh cassava root (cassava hydrolysate). Since the ini-
tial glucose concentration of the cassava medium was too high
(255 g/L), it was diluted to 150 g/L before the fermentation process
to avoid the substrate inhibition effect (Lin et al., 2008). Fig. 2
revealed that the A. succinogenes ATCC55618 was able to utilize
sugar from the cassava hydrolysate under anaerobic conditions.
The succinic acid concentrations obtained from the cassava med-
ium (Fig. 2A), and the synthetic medium (Fig. 2B) were 93.34 g/L,
and 73.00 g/L with yield of 0.77, and 0.60 gSA/gglucose, respectively.
The succinic acid productivities obtained from each batch were
345
1.87 and 1.46 g/L/h, respectively. The above results showed that
cassava medium has a high potential to produce succinic acid at
high titer, yield and productivity in comparison to the synthetic
medium. The results obtained in this experiment were comparable
to a previous report using only glucose as the sole carbon source to
produce succinic acid (Jantama et al., 2008). Succinic acid fermen-
tation from waste bread by A. succinogenes using batch separate
hydrolysis fermentation has also been investigated (Leung et al.,
2012). A succinic acid concentration of 47.3 g/L with a yield of
55 gSA/100 g waste bread was reported in that study. In addition,
an investigation on the simultaneous saccharification and fermen-
tation (SSF) for succinic acid production from corn stover using A.
succinogenes has been reported (Zheng et al., 2010). A succinic acid
concentration of 47.4 g/L with a yield of 72 gSA/100 g corn stover
were attained. Wang et al. studied succinic acid production from
corn stalk hydrolysate product using E. coli SD121. A succinic acid
concentration of 57 g/L with a yield of 87 gSA/100 g corn stalk was
reported (Wang et al., 2011). The results in these experiments were
very encouraging to continue towards succinic acid production
from the cheap and high potential material of fresh cassava roots
used in this study, especially since there is no need for the supply
of yeast extract component in the fermentation medium. Finally, a
synthetic medium using glucose as the main carbon source supple-
mented with yeast extract as a nitrogen source was investigated
(Zhu et al., 2012). However, the final succinic acid concentrations
were low, in the range of 36.4; 41.7 and 48.4 g/L, corresponding
to the ratio of glucose mass (g) and yeast extract mass (g) at
90:4; 85:7 and 80:10, respectively. This indicated that the succinic
acid production decreased with a decrease of the nitrogen source
from the yeast extract. The nitrogen source, thus, plays an impor-
tant role for the microorganism growth and product formation.
Besides that, many other factors also affect the succinic acid pro-
duction performance. One of the factors that must be mentioned
is the neutralization reagent. The high succinic acid result in this
experiment was probably due to the efficient utilization of
MgCO3 + CO2 as a neutralization reagent source which simultane-
ously ensured the anaerobic condition as well as CO2 formation
(Van der Werf et al., 1997; Xi et al., 2011). The dissolved CO2 can
directly permeate into the cell membrane, and is used as an impor-
346 N.T.H. Thuy et al. / Bioresource Technology 233 (2017) 342–352

Fig. 2. Time profile of glucose ( ), cell ( ), succinic acid ( ), lactic acid ( ), acetic acid ( ), formic acid ( ), and pyruvic acid ( ) during batch
fermentation using A. succinogenes ATCC 55618 from cassava medium (A), and synthetic medium (B). Comparison of productivity, titer and yield at the end of fermentation
between cassava medium and glucose (C).

tant substrate for the carboxylation of phosphoenolpyruvate to
oxaloacetate, which is subsequently converted to succinic acid by
the reductive tricarboxylic acid cycle, and menaquinone systems.
This characteristic allows a highly efficient production of succinic
acid. However, CO2 is not only a necessary substrate for SA
metabolism from glucose, but is also good for cell growth,
especially for the indirect CO2 source from the carbonate supple-
mentation; the same as the dissolved CO2 in the culture medium
(Hong et al., 2004; Van der Werf et al., 1997; Xi et al., 2011). It
was reported that within the carbonate supplementations, MgCO3
was considered the best neutralizing reagent for enhancing the
succinic acid production of A. succinogenes (Zou et al., 2011).
Moreover, MgCO3 prevented flocculation, prolonged the steady
phase, and improved succinic acid fermentation compared to other
neutralizing agents such as NaOH or Na2CO3 (Xi et al., 2011). Li
et al. also tested different neutralizing reagents for the succinic
acid production from glucose as a carbon source using A. succino-
genes. The results showed that MgCO3 could act as one of the best
neutralizers to help to produce succinic acid at high titers and
yield. Simultaneously, it can form CO2, which is also useful to
 N.T.H. Thuy et al. / Bioresource Technology 233 (2017) 342–352 347

Fig. 3. Time profile of glucose ( ), cell ( ), succinic acid ( ), lactic acid ( ), acetic acid ( ), formic acid ( ), and pyruvic acid ( ) during fed-batch
fermentation using A. succinogenes ATCC 55618 from cassava medium (A), and synthetic medium (B). Comparison of productivity, titer and yield after fermentation between
cassava medium and glucose (C).

increase the titers and yield of the product (Li et al., 2010). Thus, an
optimized fermentation of A. succinogenes could be carried out
with MgCO3 as the neutralizer, and continuous CO2 sparging.
3.2.2. Fed-batch fermentation of SA
The batch fermentation experiments showed that substrate
limitation in the fermentation resulted in a maximum succinic acid
production of not more than 95 g/L. As demonstrated in a
fed-batch polysaccharide and ganoderic acid fermentation using
Ganoderma lucidum (Liu et al., 2008a), the purpose of the sugar
feeding strategy is to prevent the cell growth limitation, and inhi-
bition by the glucose substrate. Therefore, in this study, the sub-
strate feeding via fed-batch fermentation was investigated to
enhance the succinic acid production. Both standard glucose and
cassava medium were compared, which allowed an evaluation of
the efficiency of the different substrate sources. Fig. 3 shows that
during the first 12 h, the glucose content was rapidly utilized from
the initial concentration of 70 g/L to approximately 30 g/L for both
cases. This corresponded to a higher amount of cell growth com-
pared to the batch run. The highly concentrated glucose solution
was intermittently fed into the bioreactor using a peristatic pump
during the period of 12–24 h to maintain a glucose concentration
within the range of 30–40 g/L. After 6 h of fermentation, the cell
concentration considerably increased, and achieved the maximum
value of 3.99 and 4.24 g/L for synthetic glucose and cassava med-
ium, respectively. In addition, the succinic acid concentrations
348 N.T.H. Thuy et al. / Bioresource Technology 233 (2017) 342–352

reached maximum values of 151.44 and 100.54 g/L for the cassava
medium and standard glucose medium, respectively. Conversely,
other organic acid by-products only slightly increased, especially
lactic acid and formic acid, which almost disappeared after the first
glucose feeding. However, there was a formation of a low amount
of pyruvic acid at a concentration up to 1.0 g/L for both cases.
Productivities in the fermentation from the synthetic glucose and
cassava hydrolysate were obtained at 1.70 and 3.22 g/L/h, respec-
tively. In comparison to batch fermentation, the succinic acid pro-
duction in fed-batch fermentation increased 37.77% and 62.84% for
standard glucose medium, and cassava medium. Thus, the success
of fed-batch fermentation was the succinic acid production at high
titers and productivity, and the high succinic titer was among the
highest published result. It was particularly significant that this
experimental result showed that the succinic acid production from
the cassava medium was more efficient than from the standard
glucose medium. The succinic acid titers and productivity were
higher than the standard glucose medium by 33.61% and 47.20%,
respectively. This result can be explained by the presence of some
vitamins and minerals in the cassava hydrolysate, which can
contribute to the cell growth; whereas the synthetic fermentable
medium does not contain these species. The results obtained in
this experiment were compared with the results reported in previ-
ous work in Table 1. Succinic acid production from A.s succinogenes
CGMCC1593 was reported at 60.2 g/L which was an increase of
55.5% in comparison to the batch fermentation of 38.7 g/L (Liu
et al., 2008a). M. succiniciproducens LPK7 was employed to utilize
glucose as a carbon source for succinic acid production in a
fed-batch fermentation, and this resulted in a 2.9-fold increase in
succinic acid concentration compared to the batch run. For the
fed-batch system, the succinic acid concentration and productivity
were 52.43 g/L and 1.8 g/L/h while the values for batch fermenta-
tion were 13.4 g/L and 0.46 g/L/h, respectively (Lee et al., 2006).
In summary, fed-batch fermentation was developed to enhance
high succinic acid product titer, which has advantages for the
sequential purification in downstream processing.
The experiment results suggested that fresh cassava root
possesses a high potential for succinic acid production without
the need for expensive auxiliary components in the fermentation
medium preparation. This, combined with the low price of fresh
cassava root means that low cost bio-based succinic acid produc-
tion is a very promising process. This is due to the presence of valu-
able nutrient components in the cassava roots, as previously
mentioned. These rich nutrient components exist principally in
the root peels utilized in this experiment. This is a major difference
of the current study when compared to previous research using
only cassava starch or cassava pulp; cassava root peels contain a
large amount of the minerals and vitamins present in the cassava.
The crude protein increases up to an average of 4.9 wt% in compar-
ison to 2.3 wt% in the tuber. The peel also contains some other
vitamins such as vitamin A, riboflavin, and vitamin C. In addition,
cassava root meal also contains some mineral components such
as nitrogen (0.73%), phosphorous (0.58%), calcium (0.15%), magne-
sium (0.12%), and potassium (0.82%) (Abu et al., 2010). This result
was comparable with some previous research for succinic acid pro-
duction from different cassava-derived substrates, and is similar to
the other substrates presented in Table 2. For example, cassava
pulp was employed for SA production using E. coli KJ122, and this
resulted in the titer, yield, and productivity of 98.63 g/L, 0.71, and
1.03 g/L/h, respectively (Sawisit et al., 2015). Sugar cane molasses
was investigated for succinic acid fermentation by A. succinogenes
in fed-batch fermentation, but the obtained succinic acid titer after
48 h was only 55.2 g/L corresponding to the obtained productivity
of 1.15 g/L/h. In addition, the system still required yeast extract as
a nitrogen source for the cell growth (Liu et al., 2008b).
3.2.3. Succinic acid crystallization from pre-treated fermentation broth
The distribution (dÞ of carboxylic acid products into acid form/
anion form will be different depending on the dissociation constant
(Ka) of each compound; for instance the dissociation constant Ka is
6.2  105 for acetic acid, 1.37  104 for lactic acid, and
2.8  103 for pyruvic acid. Succinic acid, being a dicarboxylic acid,
has two dissociation constants (Ka1 = 2.1  104, Ka2 = 2.3  106).
For the monoacid (Eq. (4)) and diacid (Eq. (5)), respectively:
½HA ½Hþ  ½A 
dHA ¼  ¼ dA ¼
þ
½HA þ ½A  ½H  þ ½Ka  ½HA þ ½A 
½Ka 
¼ ð4Þ
½Ka  þ ½Hþ 
2
½H2 A ½Hþ 
dH2 A ¼ ¼ ð5Þ
½H2 A þ ½HA  þ ½A2  2
½Hþ  þ Ka1 ½Hþ  þ Ka1 Ka2
½HA  Ka1 ½Hþ 
dHA ¼ ¼
½H2 A þ ½HA  þ ½A2  2
½Hþ  þ Ka1 ½Hþ  þ Ka1 Ka2
dA2 ¼ 1  dHA  dH2 A
where, ½Hþ  is the concentration of Hþ which is controlled by the pH
value of the solution. Therefore, at a particular pH value there is a
proportion of the species in the dissociated form.

Table 1
Comparison of succinic acid production in batch and fed-batch fermentations with previous works.

Substrate Micro-organism Fermentation strategy Succinic acid production Reference

Titers (g/L) Yield1 (g/g) Productivity2 (g/L/h)
Glucose A .succinogenes Batch 38.7 0.82 1.0 Liu et al. (2008a,b)
Fed-batch 60.20 0.75 1.30
Glucose M. succiniproducens Batch 13.40 0.64 1.22 Lee et al. (2006)
Fed-batch 52.40 0.76 1.80
Glucose A .succinogenes Batch 73.00 0.60 1.46 This study
Fed-batch 100.54 1.00 1.70
Rapeseed meal A .succinogenes Batch 15.50 0.12 0.22 Chen et al. (2011a, 2011b)
Fed-batch 23.40 0.12 0.33
Raw carob pods A .succinogenes Batch 9.41 0.54 1.32 Carvalho et al. (2016)
Fed-batch 18.97 0.94 1.43
Corn straw A .succinogenes Batch 45.50 0.81 0.95 Zheng et al. (2009)
Fed-batch 53.20 0.83 1.21
Cassava roots A .succinogenes Batch 93.34 0.77 1.87 This study
Fed-batch 151.44 1.51 3.22
1
The SA yield was calculated as SA concentration obtained divided by utilized substrate concentration during fermentation.
2
The SA productivity was calculated as SA concentration obtained divided by overall fermentation time.
 N.T.H. Thuy et al. / Bioresource Technology 233 (2017) 342–352
Table 2
Comparison of succinic acid production from cassava-derived substrates and other materials in previous works.
Substrates Microorganism Fermentation strategy
Cassava
Cassava powder E. coliNZN111 Fed-batch SSF
Cassava starch E. coliNZN111 Fed-batch SSF
Cassava pulp E. coliKJ122 Fed-batch SSF
Other substrates
Cane molasses A .succinogenes Fed-batch
Whey A. succiniproducens Fed-batch
Corn stover A .succinogenes Batch SSF
Waste bread A .succinogenes Batch SSF
Wheat A .succinogenes Batch SSF
1
The SA yield was calculated as SA concentration obtained divided by the utilizing substrate concentration during fermentation.
2
The SA productivity was calculated as SA concentration obtained divided by overall fermentation time.
Previous research has shown that at pH 2.0, all the carboxylic
acids are in their free acid form with a succinic acid solubility of
73 g/L and approximately 80 g/L at pH 2.0 and pH 3.0 respectively.
At these conditions the other carboxylic acids are still miscible in
the aqueous broth at pH 1.0–14 and temperatures above 0 °C (Li
et al., 2010). The solubility of succinic acid decreases with decreas-
ing temperature, and is only 30 g/L at 4 °C and pH 2.0. Crystalliza-
tion of succinic acid can be easily carried out at this condition.
Based on the speciation and solubility of the species involved,
the crystallization process was performed at 4 °C, pH 3.0.
The succinic acid fermentation in this study was operated at pH
6.8, which is above the pKa of the acids produced; the final acid
products are mostly in their salt forms rather than as free acids.
Therefore, when the clarified fermentation broth was adjusted
pH to 3.0 by H2SO4, succinic acid could be selectively crystallized,
while the other organic acid by-products were still fully soluble in
the broth in their free acid or salt forms. In this circumstance,
crystallization could be regarded as a powerful tool to separate
the target acids from the broth. However, the degree of purity of
the broth has an extremely important role in determining the size,
morphology, purity, and structure of the crystal product. It can be
recognized that the presence of a small amount of certain impuri-
ties can have substantial effects on the crystallization process
(Nagy et al., 2008a) and thus an NF membrane was used to sepa-
rate the main impurities before the crystallization. The NF perme-
ate (decolorized and purified SA) was subjected to an evaporation
step to eliminate some volatile compounds such as acetic acid and
lactic acid. The evaporation contributes to the removal of fermen-
tation byproducts that are detrimental to the crystallization
because of the differences in the boiling points of the species
(Tb(SA) = 235 °C; Tb(AA) = 118 °C; Tb(LA) = 122 °C); this indicates
that the impurities are far more volatile than the succinic acid
product. The removal of the acetic acid by evaporation was far
more effective than the removal of lactic acid. The evaporation also
enhances the product titers in the final broth which is helpful for
the crystallization process. It has been reported that a high titer
fermentation broth is one of the parameters that could make the
development of a biological route to succinic acid production
economically viable, and this was possible by simple water evapo-
ration and acidification of the cell free broth concentrate stream
(Meynial-Salles et al., 2008).
A key objective of the current work is to determine the effect of
seed loading on the succinic acid crystallization process. The
particle size distributions and visual appearance of the final crystal
products depended very strongly on the seed loading. Seeding
results in an increase in the particle size compared to the un-
seeded run, and high seeding results in large particle sizes with a
clear shift of the particle size distribution to the right. For the
low seeding run, the smaller particle size and narrower distribu-
349
Succinic acid production Reference
1 2
Titers (g/L) Yield (g/g) Productivity (g/L/h)
106 0.66 2.54 Chen et al. (2014)
127 0.71 1.77 Chen et al. (2014)
98.63 0.71 1.03 Sawisit et al. (2015)
55.2 0.81 1.15 Liu et al. (2008b)
34.7 0.91 1.02 Samuelov et al. (1999)
47.4 0.72 0.98 Zheng et al. (2010)
47.3 0.55 1.12 Leung et al. (2012)
64 0.4 1.06 Du et al. (2008)
tion is closer to the commercial succinic acid crystal product, with
the median size being 431 mm. In contrast, the crystal product from
the un-seeded batch was not only smaller, having a median size of
260 mm, it also had a wider distribution with a span of 1.24 (Fig. 4).
The unseeded batch also resulted in crystals having a thin-glass
slice-like morphology, and resulted in a few needle-shaped
particles, rhombuses, and spherical aggregates with low quality
crystals, as seen in Table 3. The X-ray diffraction results show that
these crystals have very low peak intensity and relative crys-
tallinity. Table 3 shows that the relative crystallinity of the product
crystals from the unseeded experiments was just 23.37% in com-
parison with the standard succinic acid crystals, which requires
further refining or purification to obtain the reagent grade succinic
acid crystals. The nucleation rate for seeded operation is much
lower than for un-seeded operation because of crystal growth onto
the seed crystals rather than the creation of nuclei (Nagy et al.,
2008a,b). This means that the un-seeded experiment had large
amounts of nucleation, producing both a smaller and wider
particle size distribution, with a lower crystallinity of the product.
This phenomenon is common in industrial crystallization, and the
reason that most batch crystallizations are seeded.
In spontaneous nucleation, solid surfaces (dust particles or the
container wall, etc.) are regarded as key to reduce the surface
energy required in the creation of new nuclei. This was easily
observed under the microscope; aggregates of succinic acid crys-
tals were also stuck onto the surface of each other in the un-
seeded experiment. This illustrated that the agglomerates formed
during un-seeded crystallizations consist of aggregates of crystals
formed into a large mass. This is similar to the results of Sander
and Kardum (2012); in this case the agglomerates of Pentaerythri-
tol also formed from the very small particles formed during the
spontaneous nucleation. This is one of the factors that facilitate
the fouling of the crystal product with impurities. Crystals with
smaller particle size have a higher probability of entrapping
mother liquor leading to lower purity (Abidin et al., 2009). The
other residual carboxylic acids in the mother liquor have a strong
affinity to the surface of the succinic acid crystals, while the smal-
ler crystal sizes have a higher contacting capacity. Therefore, the
seeding strategy was quite successful in suppressing the primary
nucleation in these experiments and avoiding the agglomeration
of the crystals (Ferguson et al., 2014; Nagy et al., 2008a,b).
Additionally, the crystals produced by seeded batches were
found to have a difference in both the morphology and the
structural quality (crystallinity) compared to those generated by
primary nucleation. In fact, one might expect that a high seeding
rate should yield larger crystals than that with a low seeding. This
is because a larger seed crystal amount suppresses secondary
nucleation more effectively and minimizes the total number of
crystals in the system. Consequently, the growth rate is dominant
350 N.T.H. Thuy et al. / Bioresource Technology 233 (2017) 342–352

Fig. 4. Comparison of the particle size distribution of obtained succinic acid crystals with commercial succinic acid.

Table 3
The effect of various seeding strategies on the relative crystallinity, phase purity, and morphology of the obtained succinic acid crystal.

Sample Relative crystallinity(%) SA-MgSO4 (%) Crystal shape

SA Standard 100 100–0 Hexagonal board, spherical, short cylinder
Low seeding 96.77 100–0 Hexagonal board, spherical, short cylinder
High seeding 92.92 100–0 Hexagonal board, spherical, short cylinder
Unseeded 23.37 100–0 Glass slice, rhombus, needle, spherical

rather than the nucleation rate (Hermanto et al., 2013). When more
seed crystals are used, more supersaturation is consumed by crystal
growth and therefore there is not a high enough supersaturation to
form a large number of secondary nuclei. In addition, larger crystals
will lead to an increase in purity (Table 4) due to their available sur-
face area being lower than that of smaller crystals, therefore limit-
ing entrapment of the impurities from the mother liquor. Table 3
shows a higher relative crystallinity of the larger crystals that were
produced with a higher seed loading than those produced using a
smaller seed loading. This result was also confirmed by the SEM
images; the smaller crystals from the low seeding had rough sur-
faces and many small pores. This is different to the high seeding
sample which has an angular appearance and fewer pores. Thus,
seeding is one of the methods used to control the succinic acid crys-
tallization to generate product crystals that are more uniform in
size and shape, and also to suppress the nucleation rate - especially
the primary nucleation rate, in order to optimize the particle size
distribution. When seeded, the seed crystals enter into a growth
Table 4
Result of the succinic acid crystallization trials.
Low seeding High seeding
(5 wt%) (10 wt%)
Recovery rate (%) 93.47 91.86
Purity (%) 99.18 99.35
Glucose removal (%) 99.65 99.94
Protein removal (%) 99.85 99.97
phase via the incorporation of solute (in this case succinic acid) onto
the surface of the crystals which allows them to grow to larger sizes.
Seeding at a particular time during the cooling crystallization is
also very important for the control of the succinic acid crystalliza-
tion. It is important to seed at a suitable temperature during the
cooling process, and this point is called the seed temperature. If a
high seed temperature is used, then the seed crystals will be easily
dissolved; in contrast, if the seed temperature is too low then
nucleation will occur as a large degree of supersaturation has been
created before the addition of the seed. The solubility of succinic
acid decreases with decreasing temperature. After evaporation at
80 °C the solute concentration was around 21 Brix; the saturation
temperature for this succinic acid concentration was found to be
approximately 50 °C. It is optimal to seed the batch as soon as
the succinic acid has become saturated, and hence the batch was
seeded at 50 °C. For low seeding, the superaturation is still high
enough for secondary nucleation to occur when the cooling rate
becomes faster at the second stage of cooling (30 °C down to
4 °C). Secondary nucleation did not occur to a significant extent
in the high seeding experiment where the amount of supersatura-
tion was just enough for the growth of the available crystals. This
was because the supersaturation at any point is controlled by the
Unseeded rate of cooling and the rate of consumption of the solute by the
(0 wt%) growth of the seed crystals added. This was reason for the lower
92.04 recovery rate of the high seeding run than that of low seeding
96.86 run as presented in Table 4. If the cooling rate was too fast the rate
98.71
of generation of supersaturation may be too fast, leading to the
99.50
entrapment of impurities. Nonetheless, in contrary to the seeded
 N.T.H. Thuy et al. / Bioresource Technology 233 (2017) 342–352
batches, the un-seeded batch had a larger metastable zone width
which meant it underwent excessive spontaneous nucleation,
leading to lower purity and quality crystal than the seeded
experiments, with small size, wide distribution and non-uniform
product (Nagy et al., 2008b). In conclusion, crystallization is not
only a widely used separation method for the solid-liquid mixtures
based on its capacity to produce high purity product, but it is also
the final refining and purification step for the production develop-
ment of the crystalline compound with the desired crystal proper-
ties of purity, shape, morphology and size distribution, which was
assisted by seeding and also the removal of impurities using NF
membrane separation and evaporation, and high fermentation effi-
ciency resulting in high titer.
When using an agitation rate high enough to ensure that crystal
growth is not limited by mass transfer, large amounts of seed crys-
tals, and a more uniform supersaturation distribution, then sponta-
neous nucleation can be suppressed, resulting in larger SA crystals
being produced.
4. Conclusion
SA was produced from fresh cassava root with impressive titer,
yield, and productivity in an economical, and feasible medium
using separate hydrolysis fed-batch fermentation. MgCO3 was used
as neutralization reagent with its main benefits of preventing
flocculation, prolonging the steady phase, and enhancing SA pro-
duction. The fermentation performance at high titer facilitated
the crystallization step to separate, purify and refine the SA for
the production of high purity crystalline product, and the expected
crystal properties of shape, morphology, and size distribution were
obtained. These results suggest the potential of industrial SA pro-
duction from cheap material until the final crystal product.
Acknowledgements
This research work was financially supported by the Thailand
Research Fund (TRF) – Thailand under the contract number
RSA5780051.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.biortech.2017.02.
114.
References
Abidin, S.Z., Ling, G.K.F., Abdullah, L.C., Ahmad, S., Yunus, R., Choong, T.S.Y., 2009.
Effects of temperature and cooling modes on yield, purity and particle size
distribution of dihydroxystearic acid crystals. J. Sci. Res. 33, 471–479.
Abu, O.M.G., Sanni, L.O., Erondu, E.S., Akinrotimi, O.A., 2010. Chemical composition
and cyanide levels of hybrid catfish fed whole cassava root meal in replacement
of maize. J. Food Technol. 8, 52–57.
Brimer, L., 2015. Cassava production and processing and impact on biological
compounds. In: Preedy, V. (Ed.), Processing and Impact on Active Components
in Food. Academic Press, Amsterdam, pp. 81–87.
Bradford, M.M., 1976. A rapid and sensitive method for the quantitation of
microgram quantities of protein utilizing the principle of protein-dye binding.
Anal. Biochem. 72, 248–254.
Bretz, K., 2015. Succinic acid production in fed-batch fermentation of
anaerobiospirillum succiniciproducens using glycerol as carbon source. Chem.
Eng. Technol. 38, 1659–1664.
Carel, D.V.H., Willie, N., 2013. Continuous and batch cultures of Escherichia coli
KJ134 for succinic acid fermentation: metabolic flux distributions and
production characteristics. Microb. Cell Fact. 12, 1–10.
Carvalho, M., Roca, M., Reis, M.A.M., 2016. Improving succinic acid production by
Actinobacillus succinogenes from raw industrial carob pods. Bioresour. Technol.
218, 491–497.
Chen, C., Ding, S., Wang, D., Li, Z., Ye, Q., 2014. Simultaneous saccharification and
fermentation of cassava to succinic acid by Escherichia coli NZN111. Bioresour.
Technol. 163, 100–105.
351
Chen, K.Q., Li, J., Ma, J.F., Jiang, M., Wei, P., Liu, Z.M., Ying, H.J., 2011a. Succinic acid
production by Actinobacillus succinogenes using hydrolysates of spent yeast cells
and corn fiber. Bioresour. Technol. 102, 1704–1708.
Chen, K., Zhang, H., Miao, Y., Wei, P., Chen, J., 2011b. Simultaneous
saccharification and fermentation of acid-pretreated rapeseed meal for
succinic acid production using Actinobacillus succinogenes. Enzyme Microb.
Technol. 48, 339–344.
Cok, R., Tsiropoulos, I., Rose, A.L., Patel, M.K., 2013. Succinic acid production
derived from carbohydrates: an energy and greenhouse gas assessment of a
platform chemical toward a bio-based economy. Biofuels Bioprod. Biorefin. 8,
16–29.
Collares, R.M., Miklasevicius, L.V.S., Bassaco, M.M., Salau, N.P.G., Mazutti, M.A.,
Bisognin, D.A., Terra, L.M., 2012. Optimization of enzymatic hydrolysis of
cassava to obtain fermentable sugar. J. Zhejiang Univ. Sci. B 13, 579–586.
Danavia, S., Holly, S., Peter, C.S.H., Ali, M., Darren, J.P., Brenna, A.B., Nancy, D., Gregg,
T.B., 2016. Succinic acid production from lignocellulosic hydrolysate by Basfia
succiniciproducens. Bioresour. Technol. 214, 558–566.
Doki, N., Seki, H., Takano, K., Asatani, H., Yokota, M., Kubota, N., 2004. Process
control of seeded batch cooling crystallization of the metastable a-form glycine
using an in-situ ATR-FTIR spectrometer and an in-situ FBRM particle counter.
Cryst. Growth Des. 4, 949–953.
Du, C., Lin, S.K.C., Koutinas, A., Wang, R., Dorado, P., Webb, C., 2008. A wheat
biorefining strategy based on solid-state fermentation for fermentative
production of succinic acid. Bioresour. Technol. 99, 8310–8315.
Eder, R.J.P., Schmitt, E.K., Grill, J., Radl, S., Woelfler, G.H., Khinast, J.G., 2011. Seed
loading effects on the mean crystal size of acetylsalicylic acid in a continuous-
flow crystallization device. Cryst. Res. Technol. 46, 227–237.
Elcio, R.B., Nei, P.J., 2011. Succinic acid production from sugarcane bagasse
hemicellulose hydrolysate by Actinobacillus succinogenes. J. Ind. Microbiol.
Biotechnol. 38, 1001–1011.
Emrani, P., Shohresh, F., Siamak, A.T., 2011. Effect of synthesis parameters on phase
purity, crystallinity and particle size of SAPO-34. Iran. J. Chem. Chem. Eng. 30,
29–36.
Ferguson, S., Morris, G., Hao, H., Barrett, M., 2014. Automated self seeding of batch
crystallizations via plug flow seed generation. Chem. Eng. Res. Des. 92, 2534–
2541.
Hermanto, M.W., Phua, A., Chow, P.S., Tan, R.B.H., 2013. Improved C-control of
crystallization with reduced calibration effort via conductometry. Chem. Eng.
Sci. 97, 126–138.
Hojjati, H., Rohani, S., 2005. Cooling and seeding effect on supersaturation and final
crystal size distribution (CSD) of ammonium sulphate in a batch crystallizer.
Chem. Eng. Process. 44, 949–957.
Hong, S.H., Kim, J.K., Lee, S.Y., In, Y.H., Choi, S.S., Rih, J.K., Kim, C.H., Jeong, H., Hur, C.
G., Kim, J.J., 2004. The genome sequence of the capnophilic rumen bacterium
Mannheimia succiniciproducens. Nat. Biotechnol. 22, 1275–1281.
Jantama, K., Zhang, K., Moore, J.C., Shanmugam, K.T., Svoronos, S.A., Ingram, L.O.,
2008. Eliminating side products and increasing succinate yields in engineered
strains of Escherichia coli C. Biotechnol. Bioeng. 101, 881–893.
Lin, S.K.C., Du, C., Koutinas, A., Wang, R., Webb, C., 2008. Substrate and product
inhibition kinetics in succinic acid production by Actinobacillus succinogenes.
Biochem. Eng. J. 41, 128–135.
Nagy, Z.K., Chew, J.W., Fujiwara, M., Braatz, R.D., 2008a. Comparative performance
of concentration and temperature controlled batch crystallizations. J. Process
Control 18, 399–407.
Nagy, Z.K., Fujiwara, M., Braatz, R.D., 2008b. Modelling and control of combined
cooling and antisolvent crystallization processes. J. Process Control 18, 856–
864.
Lee, S.J., Song, H., Lee, S.Y., 2006. Genome-based metabolic engineering of
Mannheimia succiniciproducens for succinic acid production. Appl. Environ.
Microbiol. 72, 1939–1948.
Leung, C.C.J., Cheung, A.S.Y., Ahang, A.Y.Y., Lam, K.F., Lin, C.S.K., 2012. Utilisation of
waste bread for fermentative succinic acid production. Chem. Eng. J. 65, 10–15.
Li, Q., Wang, D., Wu, Y., Li, W.L., Zhang, Y., Xing, J., Su, Z., 2010. One step recovery of
succinic acid from fermentation broth by crystallization. Sep. Purif. Technol. 72,
294–300.
Liu, Z.Z., Ma, C.Y., Hu, Y.D., Wang, X.Z., 2010. Effect of seed loading and cooling rate
on crystal size and shape distributions in protein crystallization—a study using
morphological population balance simulation. Comput. Chem. Eng. 34, 1945–
1952.
Liu, Y.P., Zheng, P., Sun, Z.H., Ni, Y., Dong, J.J., Wei, P., 2008a. Strategies of pH control
and glucose-fed batch fermentation for production of succinic acid by
Actinobacillus succinogenes CGMCC1593. J. Chem. Technol. Biotechnol. 83,
722–729.
Liu, Y.P., Zheng, P., Sun, Z.H., Ni, Y., Dong, J.J., Zhu, L.L., 2008b. Economical succinic
acid production from cane molasses by Actinobacillus succinogenes. Bioresour.
Technol. 99, 1736–1742.
Lu, C., Xiong, Y., Zhao, J., Wang, J., 2015. Fermentation production of succinic acid
from sucrose by engineered Escherichia coli JH208. Adv. Mater. Res. 1088, 503–
506.
Lubsungneon, J., Srisuno, S., Rodtong, S., Boontawan, A., 2014. Nanofiltration
coupled with vapor permeation-assisted esterification as an effective
purification step for fermentation-derived succinic acid. J. Membr. Sci. 459,
132–142.
Meynial-Salles, I., Dorotyn, S., Soucaile, P., 2008. A new process for the continuous
production of succinic acid from glucose at high yield, titer, and productivity.
Biotechnol. Bioeng. 99, 129–135.
352 N.T.H. Thuy et al. / Bioresource Technology 233 (2017) 342–352
Poonsrisawat, A., Wanlapatit, S., Paemanee, A., Eurwilaichitr, L., Piyachomkwan, K.,
Campreda, V., 2014. Viscosity reduction of cassava for very high gravity ethanol
fermentation using cell wall degrading enzymes from Aspergillus aculeatus.
Process Biochem. 49, 1950–1957.
Sander, A., Kardum, J.P., 2012. Pentaerythritol crystallization-Influence of the
process conditions on the granulometric properties of crystals. Adv. Powder
Technol. 23, 191–198.
Samuelov, N., Datta, R., Jain, M.K., Zeikus, J.G., 1999. Whey fermentation
by Anaerobiospirillum succiniciproducens for production of a succinate-based
animal feed additive. Appl. Environ. Microbiol. 65, 2260–2263.
Sawisit, A., Jantama, S.S., Kanchanatawee, S., Jantama, K., 2015. Efficient utilization
of cassava pulp for succinate production by metabolically engineered
Escherichia coli KJ122. Bioprocess Biosyst. Eng. 38, 175–187.
Sun, Y., Yan, L., Fu, H., Xiu, Z., 2014. Salting-out extraction and crystallization of
succinic acid from fermentation broths. Process Biochem. 49, 506–511.
Van der Werf, M.J., Guettler, M.V., Jain, M.K., Zeikus, J.G., 1997. Environmental and
physiological factors affecting the succinate product ratio during carbohydrate
fermentation by Actinobacillus sp. 130Z. Arch. Microbiol. 167, 332–342.
Wang, D., Li, Q., Yang, M., Zhang, Y., Su, Z., xing, J., 2011. Efficient production of
succinic acid from corn stalk hydrolysates by a recombinant Escherichia coli
with ptsG mutation. Process Biochem. 46, 365–371.
Wong, S.Y., Bund, R.K., Connelly, R.K., Hartel, R.W., 2012. Designing a lactose
crystallization process based on dynamic metastable limit. J. Food Eng. 111,
642–654.
Xi, Y.L., Chen, K.Q., Dai, W.Y., Ma, J.F., Zhang, M., Wei, P., Ouyang, P.K., 2013. Succinic
acid production by Actinobacillus succinogenes NJ113 using corn steep liquor
powder as nitrogen source. Bioresour. Technol. 136, 775–779.
Xi, Y.L., Chen, K.Q., Fang, X.J., Zheng, X.Y., Sui, S.S., Jiang, M., Wei, P., 2011.
Optimization of culture conditions in CO2 fixation for succinic acid
production using Actinobacillus succinogenes. J. Ind. Microbiol. Biotechnol.
38, 1605–1612.
Zeikus, J.G., Jain, M.K., Elankovan, P., 1999. Biotechnology of succinic acid
production and markets for derived industrial products. Appl. Microbiol.
Biotechnol. 51, 545–552.
Zheng, P., Fang, L., Xu, Y., Dong, J.J., Ni, Y., Sun, Z.H., 2010. Succinic acid production
from corn stover by simultaneous saccharification and fermentation using
Actinobacillus succinogenes. Bioresour. Technol. 101, 7889–7894.
Zheng, P., Dong, J.J., Sun, Z.H., Ni, Y., Fang, L., 2009. Fermentative production of
succinic acid from straw hydrolysate by Actinobacillus succinogenes. Bioresour.
Technol. 100, 2425–2429.
Zhu, L.W., Wang, C.C., Lia, R.S., Li, H.M., Wan, D.J., Tang, Y.J., 2012. Actinobacillus
succinogenes ATCC 55618 fermentation medium optimization for the
production of succinic acid by response surface methodology. J. Biomed.
Biotechnol. 2012, 1–10.
Zou, W., Zhu, L.W., Li, H.M., Tang, Y.J., 2011. Significance of CO2 donor on the
production of succinic acid by Actinobacillus succinogenes ATCC 55618. Microb.
Cell Fact. 10, 1–10.