Pre-Treatment and Utilization of Raw Glycerol From Sunflower

8/3/2019 Pre-Treatment and Utilization of Raw Glycerol From Sunflower

http://slidepdf.com/reader/full/pre-treatment-and-utilization-of-raw-glycerol-from-sunflower 1/9

Journal of Chemical Technology and Biotechnology J Chem Technol Biotechnol 83:1072– 1080 (2008)

Pre-treatment and utilization of rawglycerol from sunflower oil biodiesel forgrowth and 1,3-propanediol production

by Clostridium butyricum Asad-ur-Rehman,1 Saman Wijesekara R.G,2 Nakao Nomura,1 Seigo Sato1∗ and

Masatoshi Matsumura1,2

1Life Sciences and Bioengineering, Graduate School of Life and Environmental Sciences, University of Tsukuba, Tennoudai 1-1-1, Tsukuba

city, Ibaraki 305-8572, Japan2Sun Care Fuels Corporation, 4679-1 Kidamari, Tsuchiura City, Ibaraki 300-0026, Japan

Abstract

BACKGROUND: The objective of the present work is to report an efficient pre-treatment process for sunflower

oil biodiesel raw glycerol (SOB-RG) and its fermentation to 1,3-propanediol.

RESULTS: The growth inhibition percentages of Clostridium butyricum DSM 5431 on grade A (pH 4.0) and

grade B (pH 5.0) phosphoric acid-treated SOB-RG were similar to those of pure glycerol at 20 g glycerol L−1; i.e.,

18.5± 0.707% to 20.5± 0.7% inhibition. In grade A, growth inhibition was reduced from 85.25± 0.35% to 32± 1.4%

(a 53.25% reduction) at 40 g glycerol L−1 by washing grade A raw glycerol twice with n-hexanol (grade A-2). The

kinetic parameters for product formation and substrate consumption in anaerobic batch cultures gave almost

similar values at 20 g glycerol L−1, while at 50 g glycerol L−1 volumetric productivity (Qp) and specific rate of

1,3-propanediol formation (qp) were improved from 1.13 to 1.85 g L−1 h−1 and 1.60 to 2.65 g g−1 h−1, respectively,

by employing grade A-2 raw glycerol, while the yields were similar (0.5– 0.52 g g−1).

CONCLUSION: The results are important as the pre-treatment of SOB-RG is necessary to develop bioprocess

technologies for conversion of SOB-RG to 1,3-propanediol.

2008 Society of Chemical Industry

Keywords: biodiesel; Clostridium butyricum; free fatty acids; pre-treatment; 1,3-propanediol; raw glycerol

NOTATION

PD 1,3-Propanediol

PTT Polytrimethylene terephthalate

FFAs Free fatty acids

SOB-RG Sunflower oil biodiesel raw glycerol

Qp Volumetric productivity, grams 1,3-pro-

panediol formed per litre per hour (g

L −1 h−1)

Y p/s Yield, grams 1,3-propanediol produced pergram glycerol consumed (g g−1)

Y p/x Yield, grams 1,3-propanediol formed per

gram biomass dry weight (g g−1)

qp Specific rate of 1,3-propanediol formation,

grams 1,3-propanediol produced per gram

biomass dry weight per hour (g g−1 h−1)

Qs Glycerol uptake rate, grams glycerol con-

sumed per litre per hour (g L −1 h−1)

Y x/s Yield, grams biomass dry weight produced

per gram glycerol (g g−1)

qs Specific rate of glycerol consumption, grams

glycerol consumed per gram biomass dry

weight per hour (g g−1 h−1)

INTRODUCTION

Glycerol is usually released as a by-product of oil

and fat saponification. However, nowadays, biodiesel

production on a large commercial scale has led to theaccumulation of surplus raw glycerol in world markets.

With the production of 10 kg biodiesel, 1 kg glycerol is

produced.1 This has resulted in a very sharp decline in

raw glycerol prices over the past 2 years (from about

55 c kg−1 to about 5.5 c kg−1, a 10-fold reduction),

causing Dow Chemicals and Procter & Gamble to shut

down their glycerol production plants,2 thus attracting

researchers’ attention to convert this feedstock to

high-value added commodity chemicals. Glycerol

bioconversion to other commodity and specialty

chemicals is also essential to increase and maintain the

∗ Correspondence to: Seigo Sato, Life Sciences and Bioengineering, Graduate School of Life and Environmental Sciences, University of Tsukuba, Tennoudai

1-1-1, Tsukuba city, Ibaraki 305-8572, Japan

E-mail: [email protected]

( Received 29 November 2007; revised version received 15 January 2008; accepted 16 January 2008 )

Published online 26 March 2008; DOI: 10.1002/jctb.1917

2008 Society of Chemical Industry. J Chem Technol Biotechnol 0268–2575/2008/$30.00



Pre-treatment and utilization of raw glycerol from sunflower oil biodiesel

economic viability of the biodiesel production process

using vegetable oils.

Glycerol has been subjected to bioconversion to

propionic acid,3 succinic acid,4 butanol,5 ethanol,

formate,6 and hydrogen,7 as well as the specialty

chemical 1,3-propanediol (PD).2 However, special

attention has been paid to the microbial production

of PD from glycerol as this diol finds applicationin the production of new polymers, especially

Polytrimethylene terephthalate (PTT), with enhanced

biodegradable properties.8 Certain bacterial groups

ferment glycerol to PD. These include species of

Klebsiella, Citrobacter, Enterobacter ,9,10 Lactobacillus,11

and Clostridium.12

The biochemical pathways involved in the biosyn-

thesis of PD are shown in Fig. 1.13 The physiological

role of PD formation is the regeneration of reduc-

ing equivalents released during biochemical reactions

involved in the dehydrogenation of glycerol.14

Although the utilization of raw glycerol in fermen-

tation offers a remarkable advantage against the use

of pure glycerol owing to its lower cost and abun-

dant availability as a result of commercialization of

biodiesel, the majority of studies in microbial PD pro-

duction have been conducted using pure glycerol. In

fact, only a very few reports have recently investigated

the potential of raw glycerol issuing from biodiesel

production, for PD biosynthesis.13,15,16

Clostridium butyricum DSM 5431 could not grow at

all in the presence of raw glycerol.17 Owing to the

presence of certain impurities such as soap, methanol

and free fatty acids (FFAs) in considerable amounts,

raw glycerol issuing from biodiesel production using

sunflower oil is unfit for microbial growth andfermentation. It is therefore the purpose of this

paper to report an efficient pre-treatment process

for biodiesel raw glycerol and the utilization of pre-

treated raw glycerol for growth and PD production

by C. butyricum DSM 5431. Growth inhibition

experiments using various grades of pre-treated raw

glycerol were conducted and the data were compared

with those of pure glycerol. In addition, fermentation

experiments using various grades of pre-treated raw

glycerol were conducted in batch cultures and the

data were analyzed kinetically at 20 and 50 g L −1

of pure and pre-treated raw glycerol (grade A

and grade A-2). To our knowledge, there is no

report on the pre-treatment of raw glycerol issuing

from biodiesel production using sunflower oil for

its bioconversion to PD by C. butyricum DSM

5431.

Glycerol

3-Hydroxypropionaldehyde

Dihydroxyacetone

Biomas

GDHt

NADNADH2

ATP ADP

PDOR

3-P-Dihydroxyacetone

NAD

NADH2

ADP

ATPNADH2

NADH2

NAD

NAD

NADH2

Pyruvate

CH3COSCoA

H2 / NADH2

F D

- C O2

GDH

ADP

ATP

CH3COOH

NAD NADH2

Butyryl-SCoA

CH3CH2CH2COOH

ADP

ATP

DHAk

EMP

1,3-Propanediol

Figure 1. Biochemical pathways involved in anaerobic degradation of glycerol by Clostridium butyricum.13 GDH, glycerol dehydrogenase; GDHt,

glycerol dehydratase; PDOR, 1,3-propanediol oxidoreductase; DHAk, dihydroxyacetone kinase; FD, ferredoxine oxidoreductase.

J Chem Technol Biotechnol 83:1072– 1080 (2008) 1073

DOI: 10.1002/jctb



Asad-ur-Rehman et al .

MATERIALS AND METHODS

Biodiesel preparation

A methanol:sunflower oil molar ratio of 10:1 (366.2 g

methanol kg−1 sunflower oil) was mixed together at

400 rpm, along with 0.5% (w/w) of NaOH, at 60 ◦C,

for 1 h. These conditions were already optimized in our

laboratories18 and the process for efficient biodiesel

production from sunflower oil was scaled up to pilotplant.

Pre-treatment of raw glycerol

The biodiesel and raw glycerol phases were allowed to

separate by gravity. The raw glycerol, obtained directly

from the lower phase without any pre-treatment, was

labelled as grade E raw glycerol. It was subjected to

necessary pre-treatment by adding 85% phosphoric

acid with constant stirring at 200 rpm, at room

temperature, until the pH decreased from 12.5 to 7.0.

This acid-treated sunflower oil biodiesel raw glycerol

(SOB-RG) at pH 7.0 was labelled as grade D rawglycerol. Grade C raw glycerol was prepared by further

adding phosphoric acid to the grade D raw glycerol

until the pH reached 6.0. For grade B (pH 5.0), the

amount of phosphoric acid was further increased in

grade C (pH 6.0) raw glycerol and grade A (pH 4.0)

was prepared by further increasing the addition of

85% phosphoric acid in grade B raw glycerol (Fig. 2

and Table 1). Each acid-treated SOB-RG grade was

then left overnight in a separating funnel, followed by

decantation of the upper dark layer (FFAs and other

impurities are insoluble in the raw glycerol phase).

Phosphate salts were filtered through a 300 mm size

filter paper no. 5B with a pore size of 4µm (Advantec,

Toyo Roshi Kaisha, Tokyo, Japan). Methanol was

then removed from each grade by vacuum distillation.

This was followed by another phase separation of

glycerol from other impurities (mainly containing

FFAs and some plant pigments) by gravity settling

in a separatory funnel overnight.The grade A raw glycerol, after the above mentioned

pre-treatment, was washed three times with n-hexanol

at 300 rpm for 15min at room temperature. To

our knowledge, there is no report on the inhibitory

effect of n-hexanol on bacteria. The n-hexanol is

rather produced by Clostridium19 and certain other

bacteria,20 which is the reason for its selection for

subsequent washings of grade A raw glycerol. As

a consequence of n-hexanol washings, three further

Table 1. Characterization of various grades of sunflower oil biodiesel

raw glycerol (SOB-RG)

SOB-RG

grades

pHa

% (w/w)

Soapa

% (w/w)

Glycerolb

% (w/w) Methanolb

Ad 4.0 n.dc 90 <1.0

Bd 5.0 1.0 88 1

Cd 6.0 3.8 85 1.5

Dd 7.0 9.3 75 1.0

a Before methanol distillation.b After methanol distillation.c Not detected.d Raw glycerol after acid treatment.

Figure 2. A flow chart for the pre-treatment process of raw glycerol issuing from biodiesel production using sunflower oil.

1074 J Chem Technol Biotechnol 83:1072– 1080 (2008)

DOI: 10.1002/jctb




grades of n-hexanol-washed grade A raw glycerol,

depending upon the ratio (w/w) of n-hexanol, were

prepared (Table 2): when the obtained grade A

raw glycerol was washed once with n-hexanol (n-

hexanol:grade A raw glycerol, 1:1), it was called grade

A-1 raw glycerol. This grade A-1 raw glycerol was

again washed with n-hexanol (n-hexanol:grade A-1

raw glycerol, 1:2) and called grade A-2 raw glycerol.When grade A-2 raw glycerol was again washed with

n-hexanol (n-hexanol:grade A-2 raw glycerol, 1:4),

grade A-3 raw glycerol was obtained. The n-hexanol

in each washing was recovered by vacuum distillation

and used in subsequent washings. Only a negligible

amount of n-hexanol remained in the glycerol residue

after distillation.

Microorganism and culture maintenance

Clostridium butyricum DSM 5431, obtained from the

German collection of microorganisms (DSMZ), was

used in the present investigation. Spores of the strain

were obtained by growing vegetative cells on milk

agar slants.21 Spore suspensions of the strain were

maintained on reinforced clostridial medium (RCM,

Difco, Detroit, MI, USA) in Hungate tubes, at +4 ◦C.

The spores were heat shocked at 80 ◦C for 10min

prior to inoculation.

Growth inhibition experiments

Growth inhibition experiments were conducted in

Hungate tubes containing 9 mL RCM. The experi-

mental tubes were supplemented with either pure or

various grades of pre-treated SOB-RG at 20, 40, 60,

80 and 100 g glycerol L −1

. The control tubes weredevoid of glycerol. The initial pH in all tubes was 7.0.

The pre-culture for growth inhibition experiments was

also grown in Hungate tubes containing RCM with-

out glycerol. All experimental and control tubes were

inoculated with 1 mL exponential phase pre-culture

under anaerobic conditions and incubated at 32 ◦C

without agitation. Samples (about 200 µL) were with-

drawn from each tube after 6 h with the aid of a

syringe, for optical density measurements. The growth

inhibitory effect (IE, %) was determined from the fol-

lowing formula, proposed by Petitdemange et al .17 and

Abbad-Andaloussi et al .:22

IE =

(ODc(t =6 h) − ODc(t =0))

−(ODEi (t =6 h) − ODEi (t =0))

(ODc(t =6 h) − ODc(t =0))× 100(%)

Table 2. Characterization of various further grades of grade A

sunflower oil biodiesel raw glycerol (SOB-RG) after n-hexanol

washings

Glycerol grades

n-Hexanol/

raw glycerol

ratio

Soap

% (w/w)

Glycerol

% (w/w)

Methanol

% (w/w)

A-1 1:1 n.d.a 94 n.d.a

A-2 1:2 n.d.a 96 n.d.a

A-3 1:4 n.d.a 95.6 n.d.a

a Not detected.

where OD = optical density; c = control and Ei =

experimental (i = 1, 2, . . . , x).

Fermentation experiments

The pre-cultures for fermentation experiments were

grown in 100 mL screw-capped bottles with rubber

septa for syringe operation at 32 ◦C and without agi-

tation, overnight. The bottles were filled with 50 mL pre-boiled medium and sealed under nitrogen, fol-

lowed by autoclaving at 121 ◦C for 15min. The

medium contained (per litre deionized water): pure

glycerol (99% w/w) 20 g, K 2HPO4 3.4 g, KH2PO4 1.3

g, (NH4)2SO4 2 g, MgSO4.7H2O 0.2 g, CaCl2.2H2O

0.02 g, FeSO4.7H2O 5 mg, CaCO3 2 g, yeast extract

1 g, trace element solution23 SL 7 2 mL. The trace ele-

ment solution SL 7 had the following composition (per

litre deionized water): HCl (25%) 1mL, ZnCl2 70 mg,

MnCl2.4H2O 100mg, H3BO3 60 mg, CoCl2.6H2O

200 mg, CuCl2.2H2O 20mg, NiCl2.2H2O 20mg,

Na2MoO4.2H2O 40 mg. When this medium was usedfor pH-controlled batch cultures, the phosphate con-

centration was reduced to K 2HPO4 1 g, KH2PO4

0.5 g, and CaCO3 was omitted. All chemicals (exclud-

ing yeast extract, Difco) were purchased from Wako

Chemicals, Japan. An exponential phase pre-culture

was used as an inoculum. The concentration of pre-

treated SOB-RG or pure glycerol for batch cultures

was either 20gL −1 or 5 0 g L −1. The anaerobic fer-

mentation experiments were carried out using a 1.5

L magnetically stirred bioreactor (Biochemical Engi-

neering Marubishi, (BEM), Tokyo, Japan) with a

working volume of 1 L. The reactor was filled with

the medium, autoclaved at 121◦C for 15min andsparged with a 99.9995% pure nitrogen until the

oxidation–reduction potential (ORP) of the medium

declined to −150 mV. The pH was maintained at 7.0

throughout the course of fermentation by automatic

addition of 4 mol L −1 KOH solution. The incubation

temperature was kept at 32 ◦C. 5 mL samples were

withdrawn periodically and analysed for growth and

PD, acetic acid and butyric acid production, as well as

glycerol consumption.

Analytical methods

The optical density of both Hungate tube andbatch culture was measured at 650 nm using a UV-

visible spectrophotometer (Amersham Bioscience,

Little Chalfont, UK). The vacuum distillation of

methanol was carried out using an Eyela (Tokyo,

Japan) NE-2001 rotary evaporator operated at

165hPa. Methanol, PD, acetate, butyrate and glycerol

were analysed using a gas chromatograph (GC 14A,

Shimadzu, Kyoto, Japan) installed with a 2.1 m long

glass column (i.d. 3.2 mm) packed with Chromosorb

101 (80–100 mesh), and flame ionization detector

(FID). The injector and detector temperatures

were 200◦C and 240 ◦C, respectively. The column

temperature was 200 ◦C. Glycerol appeared at the

end as a broad peak, allowing a rough estimation

of this compound. Therefore, glycerol was further


DOI: 10.1002/jctb




0

20

40

60

80

100

120

140

20 40 60 80 100

Glycerol concentration (g L-1)

G r o w t h i n h i b i t i o n ( % )

Pure Grade A Grade B

Grade C Grade D

Figure 3. The effect of different grades of pre-treated raw glycerol on the growth inhibition of Clostridium butyricum DSM 5431 at various

concentrations. The various grades of glycerol exhibited highly significant differences (at P = 0.01) at different concentrations. Culture conditions:

incubation temperature, 32◦

C; initial pH 7.0. Determinations were carried out in duplicate.

0

20

40

60

80

100

120

20 40 60 80 100

Glycerol concentration (g L-1)

G

r o w t h i n h i b i t i o n ( % )

Pure

Grade A

Grade A-1

Grade A-2

Grade A-3

4 4 .2 %

5 3 .2 %

5 3 . 5 %

Figure 4. The effect of subsequent washings of grade A raw glycerol with n-hexanol on the growth inhibition of Clostridium butyricum DSM 5431 at

various concentrations. The n-hexanol:grade A raw glycerol (w/w) ratio was: (i) first washing, 1:1 (grade A-1 raw glycerol); (ii) second washing, 1:2

(grade A-2 raw glycerol); (iii) third washing, 1:4 (grade A-3 raw glycerol). The various grades of glycerol exhibited highly significant differences (at

P = 0.01) at different concentrations. Culture conditions as in Fig. 3.

analysed using the glycerol test kit (BoehringerMannheim, Germany) according to the instructions

of the manufacturer. Moisture content was measured

using a Karl Fischer titrator DL38 (Mettler Toledo,

Columbus, OH, USA). Soap contents were analysed

by colorimetric titration using bromophenol blue as

an indicator and 0.1 mol L −1 HCl as a titrant.

For FFA analysis, the samples were submitted to

the Environmental Research Centre (http://www.erc-

net.com), Tsukuba Science City, Japan. FFAs were

analysed by GC-17A (Shimadzu) installed with a 30 m

long DB-WAX column (i.d. 0.25 mm) and FID. The

initial column temperature was 140 ◦C (for 5 min) and

increased at the rate of 4 ◦C min−1 up to 240 ◦C (for

15 min). The injector and detector temperatures were

240 ◦C. Metals and heavy metals were analysed using

an inductively coupled argon plasma atomic emissionspectrophotometer (ICAP-757v, Nippon Jarrell-Ash,

Kyoto, Japan).

Statistical analysis

Statistical analysis was conducted using MSTAT-C

software (version 1.3). The comparison was done at

the 0.01 level. Figures 3 and 4 represent the standard

error bars and level of significance among duplicates.

RESULTS AND DISCUSSION

Pre-treatment of raw glycerol

The raw glycerol obtained during biodiesel preparation

from sunflower oil had the following chemical

composition (w/w): glycerol 30%; methanol 50%;


DOI: 10.1002/jctb




soap 13%; moisture 2%; salts (mainly sodium and

potassium) about 2– 3%; other impurities such as

non-glycerol organic matter, about 2–3%. The aim

of the pre-treatment process for SOB-RG was to

eliminate certain impurities generated during the

transesterification process, and to concentrate glycerol

to more than 80% (w/w), to reduce the concentrations

of various growth inhibitors.16 In the present studythe main growth inhibitors were methanol, soap and

FFAs. About 98% (w/w) of the soap, generated

during sunflower oil transesterification to biodiesel,

resides in the glycerol phase. This is not unexpected,

considering that glycerol is a polar compound. Similar

is the case with methanol being present in the raw

glycerol phase in the highest percentage (50% (w/w)

of the raw glycerol phase). The addition of phosphoric

acid to grade E raw glycerol resulted in a decrease

of soap content from 13% (w/w) to non-detectable

levels (grade A raw glycerol) (Table 1). However,

the acidification of SOB-RG led to the accumulationof salts of phosphoric acid and FFAs, released as

a consequence of the splitting of soap, which were

removed as mentioned under ‘Materials and methods’.

The purpose of the n-hexanol washing process for

grade A raw glycerol was to eliminate the remaining

impurities in acid-treated SOB-RG, mainly FFAs,

which inhibit bacterial growth profoundly.

Growth inhibition of Clostridium butyricum

DSM 5431

The growth inhibition results of C. butyricum DSM

5431 at various concentrations of pure and grade A andB raw glycerol were very similar, i.e., 18.5± 0.707%

to 20.5± 0.7% inhibition at 20 g glycerol L −1 (Fig. 3).

This is consistent with the results of Gonzalez-Pajuelo

et al .16 These authors used rape-seed oil biodiesel

raw glycerol. However, growth inhibition was very

profound (85.25± 0.35% to 112.5 ± 3.5% inhibition)

when the glycerol concentration was between 40

and 100gL −1 in all grades of acid-treated SOB-RG

(the growth inhibition of more than 100% is the

result of cell lysis due to the presence of inhibitory

substances, particularly FFAs). These results deviated

from those of previous workers.16

Raw glycerol oftencontains sodium and heavy metals ions which may

interfere with cell division, consequently reducing cell

viability.16 Growth of C. butyricum was inhibited at

12 g sodium ions L −1.24 Lead (Pb) interfered with cell

growth and glucose fermentation by C. butyricum at

25mg L −1 Pb.25 Nickel also inhibited cell growth.26

It should be pointed out, however, that the increasing

levels of inhibition at increasing concentrations of

various grades of pre-treated SOB-RG should not

be attributed to metal and heavy metal ions such as

sodium, lead and nickel, as the concentration of these

ions never exceeded 0.068, 1.50 and 3.0 mg L −1,

respectively, in all experiments. The other heavy metal

ions (Cd, Fe, Al, Cr, Cu etc) were either absent or

present below the detection limits of ICAP.

At low pH (grades A and B) the increased inhibitory

effect might be due to the presence of FFAs, mainly

released during soap splitting due to the addition of

85% phosphoric acid, in concentrations high enough

to inhibit bacterial growth. The various FFAs and

their relative percentages (w/w) in the raw glycerol

were: linolenic acid 0.14% < stearic acid 3.85% <

palmitic acid 5.46% < oleic acid 39.1% < linoleicacid 51.44%. The saturated FFAs (palmitic acid and

stearic acid) and linolenic acid could be removed

from raw glycerol by simple decantation; however,

oleic acid and linoleic acid could not be recovered

completely during decantation and certain proportions

of these unsaturated FFAs remained in the acid-

treated raw glycerol. Saturated FFAs do not usually

interfere with bacterial growth; however, unsaturated

FFAs (particularly oleic acid and linoleic acid) have a

profound influence on the viability of bacterial cells.27

Remarkable decreases of viability and morphological

changes such as loss of cell shape and disruption

of cell membrane were observed in bacteria exposed

to unsaturated FFAs.28–30 The growth of bacteria

used to assimilate FFAs from vegetable oils was

also restricted by unsaturated fatty acids, particularly

linoleic acid.31 At pH 6.0 (grade C) and 7.0

(grade D), the inhibitory effect might be due to

the presence of high concentrations of soap – 3.8%

and 9.3% (w/w), respectively – and FFAs, acting

simultaneously. Furusawa and Koyama reported that

the viability in bacteria was completely lost on

exposure to oleic acid and linoleic acid at about 8.5 mg

L −1, due to the instantaneous depolarization of cell

membrane potential.27 In the present investigation,however, it was observed that C. butyricum DSM

5431 could tolerate oleic acid at 18.4 mg L −1 and

linoleic acid at 32.2 mg L −1 and did not completely

lose the cell viability, i.e., when grade A acid-

treated raw glycerol was used at 40 g glycerol L −1,

presenting about 85.25% growth inhibition, which is

still less pronounced as compared to 103.5± 0.70%,

102± 1.14% and 104.5± 0.70% growth inhibition

on grades B, C and D, respectively. The mechanisms

operating behind the ability of the present strain to

retain cell viability at such a high concentration of

oleic acid and linoleic acid are unknown.The grade A raw glycerol, being most tolerated by

the bacterium, was given subsequent washings with n-

hexanol and the results of the effects of various further

grades of n-hexanol-washed grade A raw glycerol on

bacterial growth are presented in Fig. 4. A remarkable

reduction in inhibitory effect was observed between

glycerol concentrations of 40 and 100 g L −1, when

grades A-1 and A-2 raw glycerols were employed.

At 40g glycerol L −1 of grades A-1 and A-2 raw

glycerols, a 44.2% and 53.25% reduction in the

growth inhibition was observed, respectively. The

dramatic reduction in growth inhibition percentage

of the bacterium might be attributed to reduction

in the concentrations of oleic acid and linoleic acid

from 46mg and 80.5mg 100 g−1 glycerol, respectively,


DOI: 10.1002/jctb




Table 3. Fermentation times (h), PD, butyrate, acetate produced (g L−1 ), butyrate/acetate ratios and maximum optical density, OD (650 nm), in

anaerobic batch cultures of Clostridium butyricum DSM 5431

Glycerol grades Time of fermentation (h) PD produced (g L−1 ) Butyrate (g L−1 ) Acetate (g L−1 ) Butyrate/acetate ODmax (650nm)

Pure 8 10.6 1.6 1.0 1.6 2.1

Grade A 8 10.3 1.7 1.09 1.6 2.04

Grade A-2 8 10.7 1.58 0.99 1.6 2.08

Culture conditions: initial glycerol concentration, 20g L−1; incubation temperature, 32 ◦C; pH 7.0. Determinations were conducted in duplicate.

to non-detectable levels. No significant improvement

in the growth inhibition percentage was observed

when grade A-3 raw glycerol was employed (a 53.5%

reduction in growth inhibition), thus allowing the

optimization of grade A-2 raw glycerol.

A comparative kinetic study in anaerobic batch

cultures of Clostridium butyricum DSM 5431 at

20 and 50 g glycerol L−1

In order to extend the fermentation knowledge of C.

butyricum DSM 5431 grown on pre-treated SOB-RG,

anaerobic batch cultures were conducted employing

pure, grade A and grade A-2 raw glycerols, and

the data were analysed and compared kinetically.

The final concentration of PD achieved in batch

cultures containing 20 g glycerol L −1 ranged between

10.3 and 10.7 g L −1, while the fermentation times

for all types of glycerol were similar, i.e., 8 h

(Table 3). The glycerol was completely consumed.

The influence of grade A raw glycerol on Y p/s

and Qp was minimal (Table 4). These results are

consistent with those of Mu et al .15 A slight increase

in qp of grade A-2 raw glycerol (2.80 g g−1

h−1

) as

Table 4. Kinetic and stoichiometric parameters for PD formation and

consumption of pure glycerol, grade A and grade A-2 raw glycerols in


Glycerol grades

Kinetic parameters Pure Grade A Grade A-2

1,3-Propanediol formation parameters

Qp (g L−1 h−1 ) 1.32 1.28 1.33

Y p/s (g g−1 ) 0.53 0.51 0.53

Y p/x (g g−1 ) 13.25 12.87 13.20

qp (g g−1

h−1

) 2.65 2.44 2.80Glycerol consumption parameters

Qs (g L−1 h−1 ) 2.51 2.50 2.50

Y x/s (g g−1 ) 0.04 0.04 0.04

qs (g g−1 h−1 ) 5.0 4.75 5.05

Culture conditions as in Table 3.

compared to that of pure glycerol (2.65 g g−1 h−1)

can be attributed to a slightly higher growth rate

on the former glycerol. This might be due to the

presence of certain growth-promoting nutrients, such

as potassium or magnesium ions, usually present

in biodiesel raw glycerol, as well as phosphate ions

(mainly introduced during phosphoric acid addition)

in concentrations higher than those required for

fermentative glycerol metabolism; their effect became

more pronounced after the removal of FFAs in grade

A-2 raw glycerol. Butyrate/acetate ratios remainedunaltered (Table 3).

When grade A raw glycerol was used at 50g

glycerol L −1, a considerable increase in the time

of fermentation (22 h), as compared to that of

pure glycerol batch culture (14 h), was observed

(Table 5). The dramatic increase in fermentation

time is due to the slower growth, which could be

ascribed, in particular, to the inability of the strain

to tolerate oleic acid and linoleic acid concentrations

higher than 18.2 and 32.2 mg L −1, respectively, and,

consequently, a longer lag phase (2 h) as compared

to pure glycerol. However, a complete consumptionof glycerol was achieved in all cases. A drastic shift

in butyrate/acetate ratio was also observed during

fermentative growth on grade A raw glycerol batch

culture, i.e., from 3.8 for pure glycerol to 5.8 for grade

A raw glycerol. A slight increase in butyrate/acetate

ratio in some new strains of Clostridium grown on

raw glycerol was also reported previously.17 This

clearly suggests the flux of more carbon compounds

through the butyrate pathway than through the acetate

pathway. An increase in the energy requirements

of C. butyricum DSM 5431 has been pointed out

when butyrate is excreted instead of acetate, as

butyrate formation generates more ATP per mole

of the glycerol.21 The final PD concentration was

25–26gL −1.

The kinetic study revealed a diminution in Qp

value (1.13 g L −1 h−1) on grade A raw glycerol as

compared to pure glycerol (1.82 g L −1 h−1), although

Table 5. Fermentation times (h), PD, butyrate, acetate produced (g L−1 ), butyrate/acetate ratios and maximum optical density, OD (650 nm), in


Glycerol grades Time of fermentation (h) PD (g L−1 ) Butyrate (g L−1 ) Acetate (g L−1 ) Butyrate/acetate ODmax (650nm)

Pure glycerol 14 25.6 4.2 1.10 3.8 3.0

Grade A 22 25 5.96 1.03 5.8 2.65Grade A-2 14 26.0 4.3 1.18 3.6 2.9

Culture conditions: initial glycerol concentration, 50g L−1; incubation temperature, 32 ◦C; pH 7.0. Determinations were carried out in duplicate.


DOI: 10.1002/jctb




Table 6. Kinetic and stoichiometric parameters for PD formation and

consumption of pure glycerol, grade A and grade A-2 raw glycerols in


Glycerol grades

Kinetic parameters Pure Grade A Grade A-2

1,3-Propanediol formation parameters

Qp (g L−1 h−1 ) 1.82 1.13 1.85

Y p/s (g g−1 ) 0.51 0.50 0.52

Y p/x (g g−1 ) 18.28 18.50 18.60

qp (g g−1 h−1 ) 2.66 1.60 2.65

Glycerol consumption parameters

Qs (g L−1 h−1 ) 3.57 2.27 3.56

Y x/s (g g−1 ) 0.02 0.02 0.02

qs (g g−1 h−1 ) 5.21 3.68 5.10

Culture conditions as in Table 5.

Y p/s values were similar, i.e., 0.5 g g−1 (Table 6). The

lower productivity on the former glycerol is explained

by an increase in the fermentation time (Table 5).

Likewise, lower specific rates for PD formation (qp,

1.60g g−1 h−1) and glycerol consumption (qs, 3.68g

g−1 h−1) could be explained by a lower specific growth

rate, which also resulted in a longer fermentation

time. When grade A-2 raw glycerol was investigated

for its effects on the fermentation time and, hence

Qp, a decrease of 8 h in fermentation time, i.e.,

from 22 h to 14 h, and a consequent increase in Qp

from 1.13 to 1.85 g L −1 h−1, suggests the efficacy of

the solvent washing process (Tables 5 and 6). The

decrease in growth inhibition and an increase in

the fermentation efficiency of C. butyricum DSM5431 might again be attributed to the mitigation of

the inhibitory effects of oleic acid and linoleic acid,

acting simultaneously, during the solvent washing

process.

SUMMARY AND CONCLUSION

The production of PD by C. butyricum DSM 5431

using pre-treated SOB-RG as a sole carbon and energy

source was feasible due to the lower cost and abundant

availability as a consequence of biodiesel production

on a large commercial scale. The significant increasein growth inhibition (85.25± 0.35% to 112.5± 3.5%)

at 40–100g glycerol L −1 of various grades of

pre-treated SOB-RG was overcome by using n-

hexanol-washed (grade A-2) raw glycerol. The kinetic

analysis of the batch culture data reveals that the

fermentation results of pure and grade A-2 raw

glycerol were comparable at 20 as well as 50g

glycerol L −1. However, in general, 50 g glycerol L −1

was more effectively converted to PD as compared

to 20 g glycerol L −1 in terms of kinetic parameters

such as volumetric productivity of PD (Qp being

1.85gL −1 h−1 and 1.33 g L −1 h−1 at 50 and 20gL −1

grade A-2 raw glycerol, respectively), glycerol uptake

rate (Qs being 3.56g L −1 h−1 and 2.50 g L −1 h−1 at 50

and 20gL −1 of grade A-2 raw glycerol, respectively)

etc. The critical problem in the pre-treatment of

SOB-RG, i.e., the inability of C. butyricum DSM

5431 to withstand the inhibitory effects of high

concentrations of oleic acid and linoleic acid, was

lucratively solved by two washings of grade A raw

glycerol with a suitable alcohol, i.e., n-hexanol (grade

A-2 raw glycerol). Although the use of n-hexanol

during the solvent washing process could be feasibleon a semi-pilot scale due to its recovery by vacuum

distillation, it should be pointed out that the economic

viability of the pre-treatment process for SOB-RG

on an industrial scale will depend mainly upon the

cost of n-hexanol and its distillation, which are

relatively high. Therefore, we propose that some

potential alternative methods to purge SOB-RG of

its FFAs, such as in situ assimilation of FFAs by other

microorganisms, e.g., Pseudomonas sp.32 or enzymatic

de-acidification, as used for vegetable oils,33 should be

investigated to envisage an industrial process for the

pre-treatment and utilization of SOB-RG for efficient

1,3-propanediol production.

REFERENCES1 Meesters PAEP, Huijberts GNM and Eggink G, High cell

density cultivation of the lipid accumulating yeast Cryptococcus

curvatus using glycerol as a carbon source. Appl Microbiol

Biotechnol 45:575–579 (1996).

2 Yazdani SS and Gonzalez R, Anaerobic fermentation of glyc-

erol: a path to economic viability for the biofuels industry.

Curr Opin Biotechnol 18:213–219 (2007).

3 Boris A, Himmi E, Jauregui JJA, Pelayo-Ortiz C and Gonza-

les VA, Glycerol fermentation with Propionibacteria and opti-

mization of the production of propionic acid. Sci Aliments24:121–135 (2004).

4 Lee PC, Lee WY, Lee SY and Chang HN, Succinic acid

production with reduced by-product formation in the

fermentation of Anaerobiospirillum succiniciproducens using

glycerol as a carbon source. Biotechnol Bioeng 72:41–48

(2001).

5 Biebl H, Fermentation of glycerol by Clostridium pasteurianum:

batch and continuous culture studies. J Ind Microbiol

Biotechnol 27:18–26 (2001).

6 Jarvis GN, Moore ERB and Thiele JH, Formate and ethanol

are the major products of glycerol fermentation produced by

a Klebsiella planticola strain isolated from red deer. J Appl

Microbiol 83:166–174 (1997).

7 Ito T, Nakashimada Y, Senba K, Matsui T and Nishio N,

Hydrogen and ethanol production from glycerol-containingwastes discharged after bio-diesel manufacturing process. J

Biosci Bioeng 100:260–265 (2005).

8 Zeng A-P and Biebl H, Bulk chemicals from biotechnology: the

case of 1,3-propanediol production and the new trends. Adv

Biochem Eng Biotechnol 74:239–259 (2002).

9 Cheng K-K, Zhang J-A, Liu D-H, Sun Y, Liu H-J, Yang M-D,

et al , Pilot-scale production of 1,3-propanediol using Klebsiella

pneumoniae. Process Biochem 42:740–744 (2007).

10 Barbirato F, Himmi EH, Conte T and Bories A, 1,3-Pro-

panediol production by fermentation: an interesting way to

valorize glycerin from the ester and ethanol industries. Ind

Crop Prod 7:281–289 (1998).

11 Shutz H and Radler F, Anaerobic reduction of glycerol to 1,3-

propanediol by Lactobacillus brevis and Lactobacillus buchneri .

Sys Appl Microbiol 5:169–178 (1984).12 Reimann A, Abbad-Andaloussi S, Biebl Hand Petitdemange H,

1,3-Propanediol formation with product tolerant mutants

of Clostridium butyricum DSM 5431 in continuous culture:


DOI: 10.1002/jctb




productivity, carbon and electron flow. J Appl Microbiol

84:1125–1130 (1998).

13 Papanikolaou S, Fakas S, Fick M, Chevalot I, Goliotou-

Panayotou M, Komaitis M, et al , Biotechnological valoriza-

tion of raw glycerol discharged after bio-diesel (fatty acid

methyl esters) manufacturing process: production of 1,3-

propanediol, citric acid and single cell oil. Biomass Bioenerg

32:60–71 (2008).

14 Zeng A-P, Biebl H and Deckwer W-D, Microbial conversion

of glycerol to 1,3-propanediol: recent progress, in Fuels and

Chemicals from Biomass, ed. by Saha BC and Woodward J.

ACS Symposium Series 666, American Chemical Society,

Washington DC, pp. 264–179 (1997).

15 Mu Y, Teng H, Zhang D-J, Wang W and Xiu Z-L, Microbial

production of 1,3-propanediol by Klebsiella pneumoniae using

crude glycerol from biodiesel preparations. Biotechnol Lett

28:1755–1759 (2006).

16 Gonzalez-Pajuelo M, Andrade JC and Vasconcelos I, Produc-

tion of 1,3-propanediol by Clostridium butyricum VPI 3266

using a synthetic medium and raw glycerol. J Ind Microbiol

Biotechnol 31:442–446 (2004).

17 Petitdemange E, Durr C, Abbad-Andaloussi S and Raval G,

Fermentation of raw glycerol to 1,3-propanediol by new

strains of Clostridium butyricum. J Ind Microbiol 15:498–502

(1995).

18 Bambase ME, Nakamura N, Tanaka J and Matsumura M,

Kinetics of hydroxide-catalyzed methanolysis of crude

sunflower oil for the production of fuel-grade methyl esters. J

Chem Technol Biotechnol 82:273–280 (2007).

19 Kenealy WR and Weselefsky DM, Studies on the substrate

range of Clostridium kluyveri : the use of propanol and

succinate. Arch Microbiol 141:187–194 (1985).

20 Annan NT, Poll L, Sefa-Dedeh S, Plahar WA and Jakobsen M,

Volatile compounds produced by Lactobacillus fermentum,

Saccharomyces cerevisiae and Candida krusei in single starter

culture fermentations of Ghanaian maize dough. J Appl

Microbiol 94:462–474 (2003).

21 Biebl H, Glycerol fermentation to 1,3-propanediol by Clostrid-

ium butyricum: measurement of product inhibition by use of a

pH auxostat. Appl Microbiol Biotechnol 35:701–705 (1991).

22 Abbad-Andaloussi S, Manginot-Durr C, Amine J, Petitde-

mange E and Petitdemange H, Isolation and characterization

of Clostridium butyricum DSM 5431 mutants with increased

resistance to 1,3-propanediol and altered production of acids.

Appl Environ Microbiol 61:4413–4417 (1995).

23 Biebl H and Pfenning N, Isolation of members of the family

Rhodospirillaceae, in The Prokaryotes, ed. by Starr MP,

Truper HG, Balows A and Schlegel HG. Springer, Berlin,

pp. 267–273 (1982).

24 Homann T, Tag C, Biebl H, Deckwer W-D and Schink B,

Fermentation of glycerol to 1,3-propanediol by Klebsiella

and Citrobacter strains. Appl Microbiol Biotechnol 33:121–126

(1990).

25 Francis AJ and Dodge CJ, Effects of lead oxide and iron on

glucose fermentation by Clostridium sp. Arch Environ Contam

Toxicol 16:491–497 (1987).

26 Keeling AA and Cater GLF, Toxicity of copper, lead nickel and

zinc in agar culture to aerobic diazotrophic bacteria extracted

from waste derived compost. Chemosphere 37:1073–1077

(1998).

27 Furusawa H and Koyama N, Effect of fatty acids on the

membrane potential of an alkaliphilic Bacillus. Curr Microbiol

48:196–198 (2004).

28 Hazell SL and Graham DY, Unsaturated fatty acids and

viability of Helicobacter (Campylobacter ) pylori . J Clin Microbiol

28:1060–1061 (1990).

29 Khulusi S, Ahmed HA, Patel P, Mendall MA and North-

field TC, The effect of unsaturated fatty acids on Helicobacter

pylori in vitro. J Med Microbiol 42:276–282 (1995).

30 Knapp HR and Melly HA, Bactericidal effects of polyunsatu-

rated fatty acids. J Infect Dis 154:84–94 (1986).

31 Rodrigues CEC, Goncalves CB, Batista E and Meirelles AJA,

Deacidification of vegetable oils by solvent extraction. Recent

Patents Eng 1:95–102 (2007).

32 Cho SY, Kwon TW and Yoon SH, Selective removal of free

fatty acids in oils using a microorganism. J Am Oil Chem Soc

67:558–560 (1990).

33 Bhosle BM and Subramanian R, New approaches in de-

acidification of edible oils: a review. J Food Eng 69:481–494

(2005).


DOI: 10.1002/jctb

Documents

Pre-Treatment and Utilization of Raw Glycerol From Sunflower