Appl Microbiol Biotechnol (2011) 89:1255–1264

DOI 10.1007/s00253-010-3076-3

BIOENERGY AND BIOFUELS

Ethanol production from biodiesel-derived crude glycerol

by newly isolated Kluyvera cryocrescens
Won Jae Choi & Maria Regina Hartono &
Weng Heng Chan & Suan Siong Yeo

Received: 19 November 2010 / Revised: 14 December 2010 / Accepted: 14 December 2010 / Published online: 7 January 2011
# Springer-Verlag 2011

Abstract The rapidly expanding market for biodiesel has Introduction

increased the supply and reduced the cost of glycerol,
making it an attractive sustainable feed stock for the fuel The rapidly expanding market for biodiesel is remarkably
and chemical industry. Glycerol-based biorefinery is the altering the cost and availability of glycerol. In general,
microbial fermentation of crude glycerol to produce fuels approximately 10 lb of crude glycerol are formed for every
and chemicals. A major challenge is to obtain microbes 100 lb of biodiesel produced (da Silva et al. 2009; Johnson
tolerant to inhibitors such as salts and organic solvents and Taconi 2007; Karinen and Krause 2006; Pagliaro et al.
present in crude glycerol. Microbial screening was attemp- 2007). Bioethanol process also generates glycerol as a
ted to isolate novel strain capable of growing on crude byproduct up to 4–10% of the sugar feedstock (Nissen et al.
glycerol as a sole carbon source. The newly isolated 2000; Temudo et al. 2008) and 17% (w/w) of ethanol
bacteria, identified as nonpathogenic Kluyvera cryocrescens produced (Brumm and Hebeda 1998; Munene et al. 2002).
S26 could convert biodiesel-derived crude glycerol to Crude glycerol has thus been widely recognized as an
ethanol with high yield and productivity. The supplemen- attractive sustainable resource for oil and chemical indus-
tation of nutrients such as yeast extract resulted in tries. There have been several attempts to explore anaerobic
distinguished enhancement in cell growth as well as ethanol microbial assimilation of glycerol using reconstructed
productivity under anaerobic condition. When glycerol microbial systems via microbial screening or metabolic
fermentation is performed under microaerobic condition, pathway engineering. As a result, fuels such as ethanol,
there is also a remarkable improvement in cell growth, hydrogen and biogas as well as chemicals including 1,3-
ethanol productivity and yield, compared with those under propanediol, organic acids and some other high-value
strict anaerobic condition. In batch fermentation under products were found to be produced by microbial fermen-
microaerobic condition, K. cryocrescens S26 produced tation of glycerol (Choi 2008; da Silva et al. 2009). The
27 g/l of ethanol from crude glycerol with high molar yield newly isolated bacteria, Zobellella denitrificans produced
of 80% and productivity of 0.61 g/l/h. poly(3-hydroxybutyrate) (Ibrahim and Steinbüchel 2010).
The yeast Yarrowia lipolytica successfully converts glycer-
Keywords Crude glycerol . Ethanol . Fermentation . ol via phosphorylation pathway into valuable products such
Kluyvera cryocrescens as single cell oil and citric acid (Papanikolaou et al. 2008;
Makri et al. 2010). The engineered Escherichia coli
exhibited efficient utilization of glycerol for the production
of D-lactic acid (Mazumdar et al. 2010) and succinic acid
(Blankschien et al. 2010). In most commercial applications,
W. J. Choi (*) : M. R. Hartono : W. H. Chan : S. S. Yeo the quality of the glycerol must be high with acceptable
Institute of Chemical and Engineering Sciences, purity that is completely different those out of biodiesel
Agency for Science, Technology and Research (A*STAR),
plants. With respect to fuels and chemicals, it is important
1 Pesek Road,
Jurong Island, Singapore to employ cheap and readily available raw material. The
e-mail: [email protected] use of crude glycerol eliminates the need to refine it into
1256 Appl Microbiol Biotechnol (2011) 89:1255–1264

pure form and thus allows fermentation process to be take protein, the overall ethanol production was enhanced by
economically feasible. Currently, biodiesel is often made 3.4-fold and reached at 2.4 g/l (Yu et al. 2010). There have
from the triglycerides, or fat, found in vegetable oils, been several attempts to produce ethanol using crude
generating crude glycerol with glycerol content from 60% glycerol as feedstock. Enterobacter aerogenes HU-101,
to 80% (w/w). This low-grade glycerol also contains water, new isolate from methanogenic sludge could anaerobically
salts, other organic materials including residual methanol convert crude glycerol obtained from biodiesel wastes to
and free fatty acids and each component varies widely in ethanol with a yield of 0.85 mol/mol glycerol (Ito et al.
contents depending on multiple feedstocks such as rape- 2005). Ethanol was produced along with hydrogen by E.
seed, canola, palm, and soybean. Special concern needs to aerogenes NBRC 12010 from biodiesel waste glycerol in a
be taken on remaining methanol as well as sodium or bioelectrochemical reactor with thionine under anaerobic
potassium salt among these impurities because they are condition (Sakai and Yagishita 2007). The high level ethanol
known to be inhibitory to cell growth for fermentation (10 g/l) was also produced by engineered E. coli from 22
(Petitdemange et al. 1995). There have been several g/l of crude glycerol with productivity of 0.09 g/l/h under
attempts to utilize biodiesel-derived crude glycerol as anaerobic condition (Yazdani and Gonzalez 2008).
fermentation feedstock for the production of 1,3-propane- Glycerol-based biorefinery is the microbial fermentation
diol (González-Pajuelo et al. 2004; Mu et al. 2006; processes using inexpensive and readily available glycerol
Papanikolaou et al. 2004), succinic acid (Scholten et al. as the raw material to produce fuels and chemicals. A major
2009), β-carotene (Mantzouridou et al. 2008), docosahex- challenge in fermentation of the low-grade crude glycerol is
aenoic acid (Pyle et al. 2008), and human erythropoietin to obtain microbial strains tolerant to undesirable inhibitory
(Çelik et al. 2008). components such as salts and organic solvents that present
In the area of ethanol production from glycerol, yeast in crude glycerol. In this study, novel microorganism was
Pachysolen tannophilus could previously convert glycerol isolated from soil samples. Based on its ability to utilize
to ethanol at 4 g/l with a yield of 0.4 mol/mol glycerol crude glycerol as a sole carbon source, it was employed as a
during aerobic growth (Maleszka et al. 1982). Ethanol and biocatalyst for the production of fuel ethanol.
formic acid were produced during anaerobic glycerol
fermentation by mixed cultures (Temudo et al. 2008). The
synthesis of ethanol and 1,2-propanediol from glycerol was Materials and methods
also reported by use of Paenibacillus macerans (Gupta et
al. 2009). This bacterial strain could produce ethanol and Crude glycerol
1,2-propanediol as major metabolites at 3 and 0.8 g/l,
respectively, during anaerobic fermentation with 10 g/l of Crude glycerol was obtained from palm oil-based biodiesel
pure glycerol. plant in Malaysia operated by Vance Bioenergy. The
Recently, there have been intensive efforts for the composition of these crude glycerols is summarized in
efficient conversion of glycerol to ethanol via metabolic Table 1, according to company’s specification. Crude
pathway engineering of E. coli by minimizing the synthesis glycerol was used as carbon source in culture medium
of byproducts. E. coli has been reconstructed in its without purification.
metabolic processes and tested to convert glycerol into
ethanol through an anaerobic fermentation process Isolation of crude glycerol-utilizing microorganism
(Dharmadi et al. 2006; Murarka et al. 2008). The from soil
engineered E. coli strain produced 21 g/l of ethanol from
60 g/l of pure glycerol with the volumetric productivity of Soil samples were collected at various locations in
0.216 g/l/h under microaerobic condition (Durnin et al. Singapore. The surface soil was removed and the layer
2009). Furthermore, E. coli strain was designed based on
elementary mode analysis for efficient conversion of Table 1 The composition of crude glycerol derived from palm-oil
glycerol to ethanol with 90% of the theoretical ethanol yield based biodiesel plant
(Trinh and Srienc 2009). The E. coli mutant strain expressing
Component Conc. % (w/v)
alcohol–acetaldehyde dehydrogenase could grow on glycerol
under oxygen-limited conditions and exhibited ethanol Glycerol 80–85
production (16 g/l) with volumetric productivity of 0.34 Nonglycerol organics 2
g/l/h (Nikel et al. 2010). Saccharomyces cerevisiae was also Methanol 0.5
metabolically engineered to improve ethanol production Salts 5–6
from glycerol. By simultaneous overexpression of glycerol Water 10
dehydrogenase, dihydroxyacetone kinase and glycerol up-
Appl Microbiol Biotechnol (2011) 89:1255–1264 1257

underneath is collected. Water was added to the sample and Culture was initiated by inoculating 1 ml cells obtained in
vortexed. The mixture then allowed to be settled for about early exponential growth phase and incubated at 30°C
10 min and liquid phase was spread on agar plate under anaerobic condition. Growth was monitored by
containing RG minimal medium supplemented with crude optical density (OD) measurement at 600 nm. The
glycerol (25 g/l) as a sole carbon source. The RG medium maximum specific growth rate (μ, in h−1) on different
contains (per liter of water) yeast extract 2.0 g, K2HPO4 concentrations of glycerol or ethanol was determined from
1.55 g, NaH2PO4 0.85 g, (NH4)2SO4 2.0 g, and 5 ml of the slope of the least square regression lines of the
trace element solution (composed of EDTA 1.0 g/l, logarithm of OD versus time.
CaCl2∙2H2O 0.1 g/l, FeSO4⋅7H2O 0.5 g/l, ZnSO4⋅7H2O
0.2 g/l, CuSO4∙5H2O 0.02 g/l, CoCl2⋅6H2O 0.04 g/l, Fermentation of crude glycerol
MnCl 2 ⋅4H 2 O 0.1 g/l, Na 2 MoO 4 ⋅2H 2 O 0.02 g/l,
MgCl2⋅6H2O 10 g/l). Agar plates were incubated at room The seed culture (50 ml) with RG medium containing crude
temperature for 3 days under anaerobic condition provided glycerol (25 g/l) was prepared by incubation at 30°C for
by anaerobic chamber (Anaerogen, Oxoid, UK). Isolated 3 days in a screw-capped bottle (100 ml size) with gas inlet
strains through agar plate culture were subsequently and outlet port for continuous supply of N2 in order to
transferred to liquid culture containing the same RG provide anaerobic condition. The batch culture was carried
medium with crude glycerol (25 g/l) for measurement of out in a 2.5 l stirring bioreactor (BIOSTAT A Plus,
growth and metabolite analysis. Liquid culture was con- Sartorious, Germany) with a working volume of 1 l
ducted at 30°C for 3 days in a screw-capped bottle (100 ml consisting of RG medium supplemented with crude
size) with gas inlet and outlet port for continuous supply of glycerol as a sole carbon source. Temperature and agitation
either CO2 or N2 in order to provide anaerobic condition. speed were maintained at 30°C and 500 rpm, respectively.
The pH was controlled at 7.0 by adding either 2 N of HCl
Screening of ethanol producing microorganism or NaOH. Samples were taken periodically from fermentor,
filtrated and analyzed by HPLC as mentioned above. For
The liquid culture broth of crude glycerol-utilizing strains anaerobic fermentation, nitrogen gas was first supplied to
was analyzed by HPLC to quantify its metabolites fermentor in order to remove oxygen present in culture
including ethanol and other organic acids. The culture was medium. Gas line was then closed and the fermentation was
centrifuged for 10 min at 10,000×g, 4°C to remove cells initiated by inoculating seed culture. The exit gas was
and the supernatant was analyzed by HPLC. HPLC was collected in tightly sealed plastic bag, followed by GC
performed on a Shimadzu Prominence system equipped analysis. Hydrogen and carbon dioxide were analyzed with
with UV and refractive index detector. Aqueous sulfuric a Shimadzu GC-2014 gas chromatograph equipped with a
acid solution (5 mM) as a mobile phase was pumped at the thermal conductivity detector. The temperatures of column
rate of 0.6 ml/min. Samples were injected (1 μl) and and detector were 80°C and 110°C, respectively. Nitrogen
metabolites were analyzed by using a Aminex® HPX-87 H was used as a carrier gas. For microaerobic culture, gas
column (7.8 mm×30 cm, Bio-Rad, US) at oven temperature mixture composed of air and nitrogen was used to obtain a
of 42°C. The commercially available glycerol, ethanol, and specified volumetric transfer coefficient (kLa in min−1) for
other organic acids were employed as a standard for each experiment. Air was mixed in advance with nitrogen
quantitative analysis. The mixture of ten standards includ- at different volume ratios (0%, 25%, 50%, 75%, and 100%)
ing L-lactic acid, succinic acid, acetic acid, formic acid, in a gas mixer and resulting gas mixture was sparged into
propionic acid, glycerol, 1,3-propanediol, 1,2-propanediol, fermentor at the fixed rate of 0.25 l/min. The agitation rate
2,3-butanediol, and ethanol was injected under this condi- was fixed to 500 rpm. The kLa was determined by using the
tion and each metabolite was successfully analyzed without nonsteady-state method (Shuler and Kargi 2002). The
peak overlapping. fermentor containing culture medium was initially sparged
with nitrogen to completely remove oxygen. Then, the
Measurement of substrate and product inhibition mixture of air and nitrogen with various volume ratios was
introduced into the fermentor through gas mixer. The
Growth inhibition by substrate glycerol and product ethanol percentage of dissolved oxygen (DO) can be monitored
was measured in anaerobic cultures performed in a screw- by using a DO probe. The increase in the DO was recorded
capped bottle (100 ml size) containing RG medium (20 ml). and the slope of a plot of ln (100%—DO) versus time (in
For substrate inhibition test, each bottle was supplemented min) was used to calculate kLa. In order to investigate the
with different amounts of crude glycerol. For product effect of oxygen on ethanol production, kLa was varied by
inhibition, crude glycerol (25 g/l) was added into each using different ratio between air and nitrogen in inlet gas.
bottle, followed by addition of different amounts of ethanol. Each kLa represents the oxygen supply at different levels.
1258 Appl Microbiol Biotechnol (2011) 89:1255–1264

Microbial identification analysis was carried out by the neighbor-joining method

using the PHYLIP program package, version 3.68 (Depart-
Genomic DNA extraction, PCR and sequencing of 16 S ment of Genome Sciences and Department of Biology,
rRNA gene were conducted by the general procedures University of Washington) as shown in Fig. 1. On the basis
described elsewhere. The 16 S rRNA gene was amplified of sequence similarity as well as phylogenetic analysis, it
from genomic DNA by PCR using the bacterial primers. was concluded that newly isolated bacterial strain S26
The sequences of the primers used for amplification were belongs to K. cryocrescens. The K. cryocrescens S26 was
5′-AGAGTTTGATCATGGCTCAG-3′ (forward) and deposited in American Type Culture Collection with
5′ AAGGAGGTGATCCAGCCGCA-3′ (reverse), accession number of PTA-10600.
corresponding to positions 8–27 and 1,544–1,525, respec-
tively, of the E. coli 16 S rRNA sequence. The PCR Anaerobic fermentation of crude glycerol by K.
product was purified using a Wizard PCR DNA purification cryocrescens S26
system (Promega, US) and sequenced using an ABI PRISM
310 genetic analyzer by first base Pte Ltd, Singapore. A time profile of the anaerobic fermentation of crude
glycerol by K. cryocrescens S26 was demonstrated in Fig. 2
with cell growth and product formation. The fermentation
Results was carried out in a RG minimal medium supplemented
with 25 g/l of crude glycerol (80%, w/v). The glycerol
Screening of microorganisms producing ethanol from crude content in the medium is equivalent to 20 g/l. The
glycerol fermentation was monitored by quantitative analysis of
samples taken at various times. Glycerol was almost
Among about 600 soil samples tested, 57 microbial strains completely consumed within 136 h of active growth,
were isolated based on its ability to grow on crude glycerol resulting in a maximum biomass concentration of 1.0 g
as a sole carbon source under anaerobic condition. Through dry cell weight/l. Ethanol was accumulated as a major
metabolite analysis for these candidates by HPLC, nine product at the rate of 0.052 g/l/h with molar yield of 84.8%
strains were found to produce ethanol as major metabolite per consumed glycerol. Formic acid was primarily gener-
under anaerobic condition (Table 2). Obviously, there is in ated as byproduct and negligible amount of lactic and
most cases no formation of 1,3-propanediol that is a succinic acid was observed. There were no other organic
common product during anaerobic fermentation of glycerol. acids such as acetic acid and no 1,2- or 1,3-propanediol
Strain S26 was finally selected for further investigation detected in the culture broth. In a gas phase, H2 and CO2
based on its highest conversion yield for ethanol formation. were generated gradually, resulting in final cumulative
The 16 S rRNA gene sequence (1,030 bases, GenBank amount of 95 and 90 mmol at 136 h, respectively. Carbon
accession number, HQ285932) was compared with the recovery at the end of bioreactor operation was calculated
sequence data in GenBank database by using the BLAST as 0.94 based on initial amount of glycerol, quantitative
algorithm. The BLAST analysis on sequence similarity analysis of each product, biomass and CO2 evolution. The
indicated that the closest relatives of the strain S26 were elementary composition of biomass was taken as C4H7O2N,
Kluyvera cryocrescens 169 (99.4%), K. cryocrescens correspond to a molecular weight of 101 g/mol (Herbert et
WAB1904 (99.4%), E. aerogenes BAB-223 (99.3%), E. al. 1971). The high carbon recovery in products along with
aerogenes T2 (99.3%), Kluyvera ascorbata 4105 (99.3%), the synthesis of reduced compound, ethanol as a major
and K. cryocrescens 4701 (99.3%). The phylogenetic product reflects fermentative nature of the culture.

Table 2 Newly isolated strains

producing ethanol from crude Strain Biomass (g/l)a Ethanol (mM) 13PDO (mM) LA (mM) AA (mM) Conv (%)b
glycerol
S26 1.2 88 0 2.1 0 94
S29 1.1 73 0 12 5 85
S39 1.4 85 15 10 7 72
S257 1.6 80 0 8 10 49
Metabolic products, 1,3-pro-
panediol (13PDO ), L-lactic acid S280 1.2 87 4.3 15 9 53
(LA), and acetic acid (AA) were S344 1.1 65 0 18 5 70
analyzed by HPLC S362 1.3 82 0 9 13 79
a
Dry cell weight S233 1.1 75 0 22 10 65
b
Ethanol produced/glycerol S76 1.2 63 0 25 18 60
consumed (mole/mole)
Appl Microbiol Biotechnol (2011) 89:1255–1264 1259

Fig. 1 Phylogenetic tree of

newly isolated S26 and related
organisms based on 16 S rRNA
sequences (GenBank accession
number, HQ285932)

37

42

100
65
62

15

41
25

55

44

10

Substrate and product inhibition

250 1.2
The effect of crude glycerol and ethanol on the growth of
K. cryocrescens S26 is presented in Fig. 3. The specific
1
growth rate was gradually decreased when glycerol was
200
Glycerol and products (mM)

added at level up to 100 g/l. The growth inhibition effect

was more evident when the medium contained more than
Dry cell weight (g/l)

0.8
150 100 g/l of glycerol, leading to significantly suppressed cell
0.6 growth. Beside inhibitory effect derived from glycerol
100
itself, at such a high level of nonrefined crude glycerol,
0.4 fermentation broth contains other impurities like sodium or
potassium salts and heavy metals in concentrations that
50 might severely interfere with cell growth. The product,
0.2
ethanol is generally known to toxic for microbial growth.
0 0 The growth of K. cryocrescens S26 was also hindered in
0 20 40 60 80 100 120 140 the presence of ethanol, resulting in approximately 80%
Time (h)
reduction in specific growth rate at ethanol level of 50 g/l.
Fig. 2 The time profile for anaerobic fermentation of crude glycerol
by K. cryocrescens S26. Symbols represents glycerol (empty square), Effect of nutrients on cell growth and ethanol synthesis
ethanol (filled square), biomass (filled triangle), formic acid (filled
diamond), L-lactic acid (empty diamond), succinic acid (empty
triangle), H2 (empty circle), and CO2 (multiplication symbol). H2 Some nutritional components present in complex nutrient
and CO2 were expressed as accumulative amounts in mmol sources can be essential for cell growth and metabolite
1260 Appl Microbiol Biotechnol (2011) 89:1255–1264

1.4
A extract, polypeptone and typtone resulted in improvement
Specific growth rate (1/h)

1.2
in cell growth and ethanol production, compared with
1
culture without nutrient addition. Among nutrients tested,
0.8
yeast extract exhibited more than twofold enhancement in
0.6
ethanol productivity as well as biomass yield. Since yeast
0.4
extract is known to contain nitrogen and carbohydrate, it
0.2
may be utilized as carbon source to synthesized ethanol
0
25 50 75 100 125 150
or biomass. Experiment in the same medium with yeast
Crude glycerol (g/l)
extract lacking glycerol resulted in biomass concentration
of 0.89 g dry cell weight/l (74% of control culture).
1.4 However, it is noteworthy that there was negligible
B
Specific growth rate (1/h)

1.2
production of ethanol (0.22 g/l), acetic acid (0.57 g/l)
1
and succinic acid (0.56 g/l). Therefore, it is concluded
0.8
that ethanol was exclusively synthesized from glycerol
0.6
rather than from yeast extract.
0.4
Effect of dissolved oxygen on ethanol fermentation by K.
0.2
cryocrescens S26
0
0 10 20 30 40 50 60 70 80 90 100
Ethanol (g/l) In general, glycerol fermentation has been conducted under
anaerobic condition (in the absence of electron acceptors).
Fig. 3 Growth inhibition of K. cryocrescens S26 by substrate (crude
glycerol, a) and product (ethanol, b). For substrate inhibition, cell The anaerobic fermentation of glycerol could be an
growth was measured under anaerobic condition with RG medium excellent platform but as shown in Fig. 4, it requires
supplemented with different amounts of initial crude glycerol (25– expensive nutrients such as yeast extract and tryptone for
150 g/l). For ethanol inhibition test, cells were cultured under biomass synthesis. Glycerol fermentation in the presence of
anaerobic condition using RG medium containing 25 g/l crude
glycerol with ethanol initially added at different levels (0–100 g/l) oxygen is considered as a means of eliminating the need for
rich nutrients. The use of low amounts of oxygen-enabled
redox balance while preserving the ability to synthesize
production. The effect of several nutrients on cell growth reduced products. Experiments were performed in on-line
and ethanol production was investigated (Fig. 4). K. controlled fermentation system. K. cryocrescens S26 was
cryocrescens S26 was cultivated under anaerobic condi- cultivated in the liquid culture medium (total volume, 1 l)
tion at 30°C for 1 day in liquid RG medium (total volume with 50 g/l of crude glycerol. To optimize ethanol
of 50 ml) containing crude glycerol (25 g/l) as a sole production, the volumetric transfer coefficient, kLa was
carbon source, supplemented with various nutrients. Yeast varied by using different ratio between air and nitrogen in
inlet gas while other conditions such as agitation speed
(500 rpm), gas flow rate (0.25 l/min) supplied to bioreactor,
12 90 temperature (30°C), and pH 7.0 were kept constant (Table 3).
Dry cell weight (g/l), EtOH (g/l)

80
10 This enabled precise control of the oxygen transfer rate
70
in a bioreactor. As the kLa value was increased up to
EtOH Yield (%)

critical level, i.e., 0.1 min−1, there were enhanced perform-

8 60
50
6 ances in ethanol production rate, yield, and cell growth as
40
4 30
shown in Table 3. When glycerol fermentation is performed
20
in the presence of limited oxygen (i.e., microaerobic
2
10
condition, kLa value between 0 and 0.1 min−1), there is a
0 0 remarkable enhancement in cell growth, ethanol produc-
Cont YE PP TR CSL tivity and yield, compared with those under strict
Fig. 4 Effect of nutrients supplementation on the cell growth (white
anaerobic condition. In case of higher oxygen concentra-
bar) and ethanol production (black bar). Each culture was performed tion (kLa value between 0.1 and 0.2 min−1), cell growth
under anaerobic condition with RG medium containing crude glycerol continued to increase but ethanol yield was decreased
(25 g/l) supplemented with yeast extract (YE), polypeptone (PP), substantially due to the formation of byproduct such as
tryptone (TR) or corn steep liquor (CSL) at 10 g/l each. Control (cont)
culture was performed using RG medium without nutrient supple-
lactic acid and acetic acid. Ethanol synthesis was also
mentation. Ethanol yield (hatched bar) was expressed as molar yield depressed less than that obtained in anaerobic condition,
of ethanol per total glycerol added resulting in lower productivity.
Appl Microbiol Biotechnol (2011) 89:1255–1264 1261

Table 3 Fermentation of crude glycerol (50 g/l) by K. cryocrescens S26 in the presence of oxygen

kLa (min−1)a Biomass (g/l)b Ethanol (mM) SA (mM) LA (mM) FA (mM) AA (mM) Y (%)c PD (g/l/h)d

0 1.33 272 6.64 0.8 14.8 0.5 58 0.066

0.05 2.98 310 6.85 9.5 15 1.1 66 0.75
0.10 3.46 398 7.2 16 16 4.2 84 0.92
0.15 4.42 225 8.9 48 14 9.5 48 0.58
0.20 4.55 217 10.5 50.7 15 11.3 46 0.55

Metabolic products, succinic acid (SA), L-lactic acid (LA), formic acid (FA), and acetic acid (AA) were analyzed by HPLC. Hydrogen and carbon
dioxide were not determined
a
The volumetric transfer coefficient (kLa) varied by using air and nitrogen mixture as inlet gas. Air was mixed in advance with nitrogen at different volume
ratios (0%, 25%, 50%, 75%, and 100%) in a gas mixer and resulting gas mixture was sparged into fermentor at the fixed rate of 0.25 l/min. For each
experiment, the fermentor operation condition was kept constant at 500 rpm, 30 °C, and pH 7.0
b
Dry cell weight
c
Molar yield of ethanol per total glycerol added initially
d
Ethanol production rate

Production of ethanol from crude glycerol by K. biodiesel plant with distinguished high production rate of
cryocrescens S26 under microaerobic condition 0.61 g/l/h (Fig. 5).

Glycerol fermentation was conducted to optimize initial

glycerol level. K. cryocrescens S26 was cultivated in on- Discussion
line controlled fermentation system with the liquid RG
medium (total volume 1 l) containing different amounts of The glycerol-based biorefinery so far has not been
crude glycerol. In order to provide microaerobic condition, practically applicable in chemical industry due to the lack
gas was supplied to bioreactor at the rate of 0.25 l/min, of efficient biocatalysts. There have been few reports on
including air and nitrogen with 50% each (i.e., kLa, microbial conversion of glycerol to ethanol by use of wild
0.1 min−1). Fermentation was carried out at 30°C until type strains. P. tannophilus, P. macerans, and E. aerogenes
ethanol concentration reached maximum level. Ethanol have only been known to ferment glycerol to produce
production was increased as the amount of glycerol ethanol as a major metabolite. Most works on ethanol
increased up to 90 g/l as described in Table 4. There was production were recently done by genetically modified E.
no significant byproduct formation in all experiments. As coli mainly using refined glycerol as raw material. The
described in Fig. 2a, cells were severely inhibited by engineered E. coli was reconstructed by addition of
glycerol at higher level of glycerol, resulting in reduced cell fermentative pathways such as propanediol pathways that
growth as well as depression in ethanol synthesis. Produc- served as electron acceptors. Glycerol has been utilized as
tion of ethanol was performed under microaerobic condi- carbon source for many microorganisms, particularly in
tion with the liquid RG medium containing crude glycerol anaerobes (Chen et al. 2003; Gonzalez et al. 2008). Under
of 90 g/l. The newly isolated, K. cryocrescens produced anaerobic condition, it can be either oxidized to dihydroxy-
27 g/l of ethanol from crude glycerol directly obtained from acetone or dehydrated to 3-hydroxypropionaldehyde. Dihy-

Table 4 Effect of initial crude

glycerol concentration on etha- Initial crude glycerol concentration (g/l)
nol production by K. cryocres-
cens S26 under microaerobic 30 60 90 120
condition
Biomass (g/l) 2.12 4.22 4.85 3.69
Microaerobic condition was
provided by setting the volu- Residual glycerol (g/l) 0 0 0 0
metric transfer coefficient (kLa) Ethanol (g/l) 9.1 18.3 27 15.2
to 0.1 min−1 Ethanol yield (%)a 84 84 80 33
a
Molar yield of ethanol per total Ethanol productivity (g/l/h) 0.89 0.92 0.61 0.27
glycerol added initially
1262 Appl Microbiol Biotechnol (2011) 89:1255–1264

80 5
generated during cell growth cannot be consumed by the
Glycerol and products (g/l)

70
ethanol synthesis because its formation is redox-balanced

Dry cell weight (g/l)

4
60
50
process. Although there was no external electron acceptor
3
40 provided in culture medium for K. cryocrescens, it may be
30
2 derived from crude glycerol containing unidentified impu-
20
1
rities such as nonglycerol organics.
10 It was reported that glycerol fermentation in E. coli
0 0 required the supplementation of the medium with rich
0 10 20 30 40 50
nutrients in the form of tryptone or yeast extract (Dharmadi
Time (h)
et al. 2006; Durnin et al. 2009; Murarka et al. 2008; Nikel
Fig. 5 Production of ethanol from crude glycerol by K. cryocrescens et al. 2010). Under anaerobic condition, these nutrients were
S26 under microaerobic condition (kLa, 0.1 min−1). Culture was found to be essential for cell growth and not served as electron
performed with the liquid RG medium containing crude glycerol of
90 g/l. Symbols represents glycerol (empty square), ethanol (filled acceptors, enabling to maintain the fermentative nature of
square), biomass (filled triangle), and succinic acid (empty triangle) glycerol metabolism for the synthesis of reduced products like
ethanol (Murarka et al. 2008). In K. cryocrescens, yeast
extract showed the highest improvement among nutrients
droxyacetone can be further metabolized into pyruvate, tested in ethanol productivity as well as biomass yield. Yeast
which is one of the main metabolic intermediate found in extract mainly contributed to cell growth but there was no
glycolysis pathway and subsequently converted to various influence on the production of metabolites such as ethanol.
fermentation products such as lactic acid, acetic acid, Apart from being a carbon and nitrogen source, yeast extract
ethanol, and succinic acid. 3-Hydroxypropionaldehyde is provides a significant amount of vitamins. For instance, it
reduced to 1,3-propanediol, which is the most common was found that D-pantothenate was mainly responsible for
product during anaerobic fermentation of glycerol. The enhanced ethanol synthesis from glycerol by E. coli because
production of 1,3-propanediol is a way to regenerate NAD, it is an important precursor in the acetyl-CoA biosynthetic
which is used for biomass production and other products pathway of E. coli (Nikel et al. 2010).
from glycerol. Glycerol can also be assimilated via glycerol It has been believed that glycerol metabolism in E. coli
3-phosphate to pyruvate under aerobic condition. Based on was limited to respiratory conditions. Recently, availability
the metabolite production observed in anaerobic culture of low level of oxygen enabled redox balance while
(Fig. 2), the newly isolated K. cryocrescens is likely to maintaining the ability to produce reduced products. The
follow proposed metabolic pathways as previously de- engineered E. coli strain produced 21 g/l of ethanol from
scribed (Dharmadi et al. 2006; Durnin et al. 2009) in which 60 g/l of pure glycerol with the volumetric productivity of
glycerol is mainly converted into biomass and reduced 0.216 g/l/h under microaerobic condition (Durnin et al.
products like hydrogen and ethanol. The generation of 2009). K. cryocrescens also metabolized glycerol to
hydrogen and carbon dioxide under anaerobic condition produce ethanol in the presence of limited oxygen. The
reveals the fermentative metabolism of glycerol in K. supply of limited amount of an electron acceptor such as
cryocrescens. The combined amount of formate (28 mmol) oxygen facilitated in glycerol metabolism, enabling redox
and hydrogen (95 mmol) is close to total ethanol balance by consuming the excess reducing equivalents
(153 mmol) synthesis, indicating that pyruvate is mainly generated in the course of biomass synthesis. Oxygen could
dissimilated via pyruvate-formate lyase. This result corre- also eliminate the use of expensive nutritional components
lated with the fact that enzyme pyruvate dehydrogenase in culture medium. Although the use of oxygen as an
activity is known to be negligible during the anaerobic electron acceptor enables redox balance and eliminates the
fermentation of glucose (de Graef et al. 1999). In other need for medium supplementation, high level oxygen
species such as Clostridium butyricum, Klebsiella pneumo- would lead to low product yields as most of the carbon is
niae, and E. aerogenes, pyruvate can also be converted into integrated into cellular mass and converted to carbon
acetyl-CoA along with the regulated formation of reducing dioxide. The use of microaerobic conditions (i.e., limited
equivalents and hydrogen via pyruvate dehydrogenase amount of oxygen), on the other hand, has the potential to
(Zeng et al. 1993). In anaerobic condition, cells can grow eliminate the requirement of rich nutrients while preserving
only if external electron acceptors are provided either in the the capacity to synthesize large amounts of products.
form of feeding substrate or metabolic pathways that can Limitation of oxygen consumption would likely lead to
consume reducing equivalent NADH. 1,2- or 1,3-Propane- higher ethanol yield. The level of oxygen available plays a
diol, the product of metabolic pathway that can serve as an key role in controlling the competition between the biomass
electron sink, was not detected in the fermentation broth of formation and ethanol production pathways. As demon-
K. cryocrescens. The excess of reducing equivalents strated in Table 3, K. cryocrescens was able to convert
Appl Microbiol Biotechnol (2011) 89:1255–1264 1263

glycerol to ethanol with an ethanol yield as high as 84% in Acknowledgments This work was supported by Science and
Engineering Research Council of Agency for Science, Technology
a defined medium under oxygen-limiting growth conditions
and Research (A*STAR), Singapore, grant number ICES/07-173A02.
in which kLa was set as 0.1 min−1. In contrast, it produced
mainly biomass and not ethanol when sufficient oxygen
was supplied at higher kLa. Oxygen supply at high level can References
enhance the generation of ATP by reducing NADH, which
is then used for biomass synthesis. Therefore, more carbon Blankschien MD, Clomburg JM, Gonzalez R (2010) Metabolic
flux is driven to biomass production but less carbon flux for engineering of Escherichia coli for the production of succinate
ethanol synthesis. On the other hand, under microaerobic from glycerol. Metab Eng 12:409–419
condition, limited amount of oxygen managed to convert Brumm PJ, Hebeda RE (1998) Glycerol production in industrial
alcohol fermentations. Biotechnol Lett 10:677–682
NADH generated during cell growth into NAD while Çelik E, Ozbay N, Oktar N, Çalik P (2008) Use of biodiesel
maintaining carbon flux into ethanol synthesis. This ethanol byproduct crude glycerol as the carbon source for fermentation
production pathway is dramatically enhanced in the strain processes by recombinant Pichia pastoris. Ind Eng Chem Res
K. cryocrescens S26. In glycerol metabolism, biomass 47:2985–2990
Chen X, Xiu Z, Wang J, Zhang D, Xu P (2003) Stoichiometric
production requires NAD for oxidation reactions, resulting analysis and experimental investigation of glycerol bioconversion
in NADH. Cells can only grow under redox-balanced to 1, 3-propanediol by Klebsiella pneumoniae under micro-
condition in which the formed NADH is regenerated to aerobic conditions. Enzyme Microb Technol 33:386–394
NAD. The redox balanced condition can be obtained by Choi WJ (2008) Glycerol-based biorefinery for fuels and chemicals.
Recent Pat Biotechnol 2:173–180
synthesizing reduced metabolites such as 1,3-propanediol. da Silva GP, Mack M, Contiero J (2009) Glycerol: a promising and
Since E. coli and K. cryocrescens do not have 1,3-propane- abundant carbon source for industrial microbiology. Biotechnol
diol pathway, they need another form of NAD regeneration Adv 27:30–39
such as oxygen utilization or even prevent the necessity of de Graef MR, Alexeeva S, Snoep JL, Teixeira de Mattos MJ (1999)
The steady-state internal redox state (NADH/NAD) reflects the
NAD regeneration by not synthesizing biomass but using external redox state and is correlated with catabolic adaptation in
complex compounds present in the medium. Escherichia coli. J Bacteriol 181:2351–2357
There have been several attempts to produce ethanol Dharmadi Y, Murarka A, Gonzalez R (2006) Anaerobic fermentation
from glycerol using microbial fermentation. In most cases, of glycerol by Escherichia coli: a new platform for metabolic
engineering. Biotechnol Bioeng 94:821–829
however, pure glycerol was utilized as substrate (Maleszka Durnin G, Clomburg J, Yeates Z, Alvarez PJJ, Zygourakis K,
et al. 1982; Dharmadi et al. 2006; Murarka et al. 2008; Campbell P, Gonzalez R (2009) Understanding and harnessing
Durnin et al. 2009; Gupta et al. 2009) due to the toxicity of the microaerobic metabolism of glycerol in Escherichia coli.
crude glycerol towards microbial strains, especially when Biotechnol Bioeng 103:148–161
Gonzalez R, Murarka A, Dharmadi Y, Yazdani SS (2008) A new
used at high level. There have been only few reports on use model for the anaerobic fermentation of glycerol in enteric
of crude glycerol as fermentation feedstock for ethanol bacteria: trunk and auxiliary pathways in Escherichia coli. Metab
production. However, the level of crude glycerol used was Eng 10:234–245
not as high as its pure form. The newly isolated K. González-Pajuelo M, Andrade JC, Vasconcelos I (2004) Production of
1, 3-propanediol by Clostridium butyricum VPI 3266 using a
cryocrescens produced 27 g/l of ethanol from biodiesel- synthetic medium and raw glycerol. J Ind Microbiol Biotechnol
derived crude glycerol with much higher production rate of 31:442–446
0.61 g/l/h. It is the first report on the fact that novel Gupta A, Murarka A, Campbell P, Gonzalez R (2009) Anaerobic
bacteria, K. cryocrescens was isolated from nature and it fermentation of glycerol in Paenibacillus macerans: metabolic
pathways and environmental determinants. Appl Environ Micro-
was applied to microbial fermentation of crude glycerol in biol 75:5871–5883
order to produce ethanol with the highest productivity. Herbert D, Phipps PJ, Strange RE (1971) Chemical analysis of
Although bioprocess development for ethanol production microbial cells. In: Norris JM, Ribbons DW (eds) Methods
has been almost saturated, there might be a niche market in microbiology, vol 5b. Academic Press, London, pp 209–
344
for glycerol-oriented ethanol as crude glycerol price is Ibrahim MHA, Steinbüchel A (2010) Zobellella denitrificans strain
getting decreased and availability is continued to MW1, a newly isolated bacterium suitable for poly(3-hydrox-
increase. Considering worldwide surplus of crude glyc- ybutyrate) production from glycerol. J Appl Microbiol 108:214–
erol, ethanol fermentation using glycerol as substrate is 225
Ito T, Nakashimada Y, Senba K, Matsui T, Nishio N (2005) Hydrogen
promising alternative process to supply fuel ethanol. In and ethanol production from glycerol-containing wastes dis-
biorefinery process for the production of fuels and charged after biodiesel manufacturing process. J Biosci Bioeng
chemicals, the product price is affected mainly by raw 100:260–265
material costs. The introduction of new technologies for Johnson DT, Taconi AT (2007) The glycerin glut: options for the
value-added conversion of crude glycerol resulting from bio-
the efficient use of such raw materials is currently being diesel production. Environ Prog 26:338–348
promoted. The utilization of crude glycerol, for example, Karinen RS, Krause AOI (2006) New biocomponents from glycerol.
provides great potential. Appl Catal Gen 306:128–133
1264 Appl Microbiol Biotechnol (2011) 89:1255–1264

Makri A, Fakas S, Aggelis G (2010) Metabolic activities of acid methyl esters) manufacturing process: production of 1,3-
biotechnological interest in Yarrowia lipolytica grown on propanediol, citric acid and single cell oil. Biomass Bioenergy
glycerol in repeated batch cultures. Bioresour Technol 32:60–71
101:2351–2358 Papanikolaou S, Fick M, Aggelis G (2004) The effect of raw glycerol
Maleszka R, Wang PY, Schneider H (1982) Ethanol production from concentration on the production of 1,3-propanediol by Clostrid-
D-galactose and glycerol by Pachysolen tannophilus. Enzyme ium butyricum. J Chem Technol Biotechnol 79:1189–1196
Microb Technol 4:349–352 Petitdemange E, Dürr C, Andaloussi SA, Raval G (1995) Fermenta-
Mantzouridou F, Naziri E, Tsimidou MZ (2008) Industrial glycerol tion of raw glycerol to 1, 3-propanediol by new strains of
as a supplementary carbon source in the production of Clostridium butyricum. J Ind Micribiol 15:498–502
β-carotene by Blakeslea trispora. J Agric Food Chem Pyle DJ, Garcia RA, Wen Z (2008) Producing docosahexaenoic acid
56:2668–2675 (DHA)-rich algae from biodiesel-derived crude glycerol: Effect
Mazumdar S, Clomburg JM, Gonzalez R (2010) Escherichia coli of impurities on DHA production and algal biomass composition.
strains engineered for homofermentative production of D-lactic J Agric Food Chem 56:3933–3939
acid from glycerol. Appl Environ Microbiol 76:4327–4336 Sakai S, Yagishita T (2007) Microbial production of hydrogen and
Mu Y, Teng H, Zhang DJ, Wang W, Xiu ZL (2006) Microbial ethanol from glycerol-containing wastes discharged from a
production of 1, 3-propanediol by Klebsiella pneumoniae using biodiesel fuel production plant in a bioelectrochemical reactor
crude glycerol from biodiesel preparations. Biotechnol Lett with thionine. Biotechnol Bioeng 98:340–348
28:1755–1759 Scholten E, Renz T, Thomas J (2009) Continuous cultivation
Munene CN, Kampen WH, Njapau H (2002) Effects of altering approach for fermentative succinic acid production from crude
fermentation parameters on glycerol and bioethanol production glycerol by Basfia succiniciproducens DD1. Biotechnol Lett
from cane molasses. J Sci Food Agric 82:309–314 31:1947–1951
Murarka A, Dharmadi Y, Yazdani SS, Gonzalex R (2008) Fermenta- Shuler ML, Kargi F (2002) Bioprocess engineering: basic concepts.
tive utilization of glycerol by Escherichia coli and its implica- Prentice Hall, Upper Saddle River
tions for the production of fuels and chemicals. Appl Environ Temudo MF, Poldermans R, Kleerebezem R, van Loosdrecht MCM
Microbiol 74:1124–1135 (2008) Glycerol fermentation by (open) mixed cultures: a
Nikel PI, Ramirez MC, Pettinari MJ, Méndez BS, Galvagno MA chemostat study. Biotechnol Bioeng 100:1088–1098
(2010) Ethanol synthesis from glycerol by Escherichia coli redox Trinh CT, Srienc F (2009) Metabolic engineering of Escherichia coli
mutants expressing adhE from Leuconostoc mesenteroides. J for efficient conversion of glycerol to ethanol. Appl Environ
Appl Microbiol 109:492–504 Microbiol 75:6696–6705
Nissen TL, Hamann CW, Kielland-Brandt MC, Nielsen J, Villadsen J Yazdani SS, Gonzalez R (2008) Engineering Escherichia coli for the
(2000) Anaerobic and aerobic batch cultivation of Saccharomy- efficient conversion of glycerol to ethanol and co-products.
ces cerevisiae mutants impaired in glycerol synthesis. Yeast Metab Eng 10:340–351
16:464–474 Yu KO, Kim SW, Han SO (2010) Engineering of glycerol utilization
Pagliaro M, Ciriminna R, Kimura H, Rossi M, Pina CD (2007) From pathway for ethanol production by Saccharomyces cerevisiae.
glycerol to value-added products. Angew Chem Int Ed 46:4434– Bioresour Technol 101:4157–4161
4440 Zeng AP, Biebl H, Schlieker H, Deckwer WD (1993) Pathway
Papanikolaou S, Fakas S, Fick M, Chevalot I, Galiotou-Panayotou M, analysis of glycerol fermentation by Klebsiella pneumoniae:
Komaitis M, Marc I, Aggelis G (2008) Biotechnological regulation of reducing equivalent balance and product formation.
valorization of raw glycerol discharged after bio-diesel (fatty Enzyme Microb Technol 15:770–779